ABSTRACT. Nowadays IT education is very important, so a lot of effort is put into the development of tools for helping students in learning, or teachers in lecturing. This paper describes a prototype of the program segment assembling system. These segments can be used for examining in the field of algorithmic complexity. The system is under development, so this paper presents developed parts and projections for future work. New program segment assembling model, which uses rules and templates is introduced. The segment is built with the set of the templates, which are connected with rules. The template is a simple program segment. The rule defines combining method and data dependencies if they exist. Finally, an example of program segment assembling by the proposed system is given.

ABSTRACT. U radu je predstavljen sistem za testiranje funkcionalnosti i robusnosti programskog koda. Testiranje se vrši konvencionalnim metodama Equivalence Partitioning (EP) i Boundary Value Analysis (BVA), koje su nezavisne od platforme, tehnologije, namene ili jezika u kome je pisan programski kod. Kao alat za ekstrakciju parametara funkcija i integraciju odgovarajućeg kompajlera i debugger-a korišćeno je okruženje za testiranje pretežno embedded softvera napisanog u C-u ili C++ - u.

ABSTRACT. Korisnički režim emulatora QEMU se bavi emuliranjem funkcionalnosti jezgra Linuksa jednog procesora koristeći jezgro Linuksa drugog procesora, omogućavajući time izvršavanje jednostavnih Linuks aplikacija pisanih za jedan procesor na sistemu sa drugim procesorom. Iako je taj režim QEMU-a razvijan više od decenije od strane zajednice otvorenog koda, njegova emulacija nije sasvim verna, najviše zbog nepostojanja podrške za neke složenije funkcionalnosti jezgra Linuksa. U ovom radu je izloženo rešenje koje upotpunjuje funkcionalnost korisničkog režima QEMU-a u vidu dodavanja podrške za specijalne operacije nad deskriptorima datoteka. U sadašnjoj fazi tog rešenja, implementirana je podrška za manipulaciju deskriptorima datoteka koji su u vezi sa podsistemom jezgra Linuksa za notifikaciju fanotify. Međutim, rešenje je od početka zamišljeno da može da bude lako prošireno i na preostale slučajeve specijalnih operacija nad deskriptorima datoteka.

ABSTRACT. U ovom radu predstavljeno je proširenje IaaS Cloud simulatora CloudSim. Prvo je modelovan zadatak u vidu transakcije nad transakcionom memorijom i komunikacija između centara za obradu podataka upotrebom protokola ažuriranja u dve faze. Zatim je implementiran model prototipa distribuirane STM. Na kraju je urađena evaluacija i dobijeni rezultati poređeni su sa prethodno izvedenim teorijskim rezultatima. Prezentovani rezultati su pozitivni i stimulišu budući rad u razvoju distribuirane STM.

ABSTRACT. Algoritmi mašinskog učenja i pretrage podataka često se primenjuju u rešavanju problema u različitim oblastima. U ovom radu prikazano je nekoliko algoritama za pretprocesiranje i klasifikaciju teksta u mejl porukama sa ciljem izbora najboljeg modela i njegove primene unutar postojećeg Postfix mejl servera. U radu su obrazloženi algoritmi koji su korišćeni, pojašnjena njihova implementacija, prikazan postupak uporedne analize, koji je rezultovao najboljim modelom za klasfikaciju i prikazana pozicija modela kao komponente unutar mejl servera.

ABSTRACT. U radu je opisan razvoj virtuelne stvarnosti sa naglaskom na moderne uređaje i alate za razvoj aplikacija. Prikazano je poređenje popularnih VR uređaja za PC, desktop i mobilne uređaje. Nakon toga biće opisana okruženja za razvoj nativnih, višeplatformskih i veb aplikacija. Kroz razvoj VR aplikacije biće opisane najbolje prakse i standardi za razvoj softvera. Na kraju, biće predstavljena VR demo aplikacija i detalji razvoja.

ABSTRACT. The paper discusses designing of the frame filtering tables in distributed computing or telecommunication systems. The proposed method of filtering tables design can reduce the time of frame processing by network bridges and switches and provide a low probability of filtering table overflowing. We have studied the new method and determined optimal distributing of memory amount for table allocating. Also we discuss the energy efficient of devices in dependence on filtering table design. Hash table is a common approach to build associative arrays, database indexes and various kinds of program-defined caches. Our approach allows to design ASIC to perform function of fast search in associative arrays. It leads to significant decreasing power consumption. Moreover, hashing techniques are suffered from large probability of collision in the case of hash size acceptable for mobile devices. This makes it necessary to perform additional energy-inefficient memory access operations to resolve these collisions. We propose hashing technique with lower probability of collision for the hash of the same size. We show that unlike existing collision free approaches our hashing method has a much broader area of applicability. To support these claims both theoretical and experimental studies are presented. Experimental comparison with existing approaches has shown significant improvement of energy-efficiency for common applications.

ABSTRACT. In this paper, the new method and the methodology are presented for the education of students of electrical engineering at low-cost computers and free software as computer algebra system (CAS). These manner and method contribute to students for solving electrical circuits in transient regimes, especially if they have a poor knowledge in the field of mathematics. CAS helps in the formation of equations relating to a circuit and it is not necessary to know the field of differential equations. The obtained solutions are symbolic expressions, which indicate the transient response in closed-form expressions. Such expressions can be used further for manipulation and processing. In this way, students are exempted from long-term learning of mathematics.

ABSTRACT. SchematicSolver is an application for symbolic-numeric simulation, analysis and implementation of the linear continuous-time systems and linear or nonlinear digital systems and algorithms. In order to provide for this application to be used in the same environment (the computer algebra system Mathematica), as other built-in functions, the strict architecture of folders and files was developed. Instead of using development platforms that are usually used in other software environments, the whole methodology is based on the Wolfram language. The proposed methodology is applicable for small and medium sized application.

ABSTRACT. SchematicSolver application and Wolfram language are used for the analysis and design of filters that consist of LC sections and quartz crystals. The filter is modelled by LC equivalent sections and represented as rational function. In this paper we present a template notebook for modeling, analyzing, designing of known filter structure. From schematic model the filter transfer function can be derived as closed-form expression so that other analysis are possible, such as parameter optimization in order to satisfy desired filter specifications.

ABSTRACT. As powerful image editing tools are widely used, the demand for identifying the authenticity of an image is much increased. In a Copy-Move forgery, a part of the image itself is copied and pasted into another part of the same image, and possible postprocessing. In recent years, the detection of Copy-Move forgery has become one of the most attractive method in image forensics. In this paper, several examples of images obtained by copy- move method are shown, as well as their multifractal spectrum. Analysis of the obtained multifractal spectrums shows that the method of copy move forgery affects on the brightness, and thus on multifractality of images, inserting the object which does not correspond to environment of existing image. It is expected to be incurred additional "unnaturalness" depending on the statistical nature of local parts of image.

ABSTRACT. Memory intensive problems are a group of problems which require a lot more access and work with the memory in regard to computing problems. One example of such is a multidimensional (2D and 3D) FDTD algorithm which served as a motivation for this paper. To cope with the specifics of memory intensive problems, the most common solutions include invoking software parallelism and/or making improvements to the current hardware or creating a completely new one. The principle chosen by the authors was to use the programmability of an FPGA in order to achieve maximal memory bandwidth with a lower price and power consumption. To realize such a specific architecture an appropriate chip is required. The goal of this paper is to describe a script used to find an optimal FPGA based on the search criteria – highest achievable memory bandwidth per price.

ABSTRACT. Our main goal is to design and realize the computer based instrument, with additional microcontroller hardware for analogue signal processing, for modified digital stochastic measurement (MDSM) of biomedical signals. In order to achieve this goal, and design the computer based instrument correctly, an experimental hardware solution for MDSM of non-stationary biomedical signals has been developed. The MDSM method is based on stochastic analog-to-digital (A/D) conversion, with a low-resolution A/D converters, digital multipliers and accumulation. An electrooculography (EOG) signal is taken as an example. Tests were carried out for set of 100 measurements. Various number of Fourier coefficients per set were measured: first 15 Fourier coefficients were measured, and after that only 7. The results were compared. Comparison was also made with results obtained with classical digital measurement. The obtained results are presented, analyzed and discussed.

ABSTRACT. This paper describes proposed system for real-time data acquisition based on System on Chip with an FPGA connected with a PC over PCI Express bus. Results show that this architecture enables high speed data acquisition for a reasonable price which makes it commercially viable.

ABSTRACT. This paper analyzes a comparison of two methods for determining a measurement uncertainty: as recommended by ISO - "International Organization for Standardization" and using Monte Carlo method. The first concept is accepted and required in modern metrology. However, there are special cases when it doesn't provide good results. In those cases, Monte Carlo method is used.

ABSTRACT. Measurement uncertainty of an ideal one-bit analog-to-digital converter is presented. A low resolution of such converters is overcome by adding a specific random uniform noise, dither, to the measured signal, and by averaging the digital output on a time interval. High resolution of measurement result, and accuracy too, can be achieved when the measured quantity is defined as integral over the certain time interval, like effective value, active/reactive electrical power or energy, etc. For example, choosing time interval of approx. 1 s and sampling frequency of approx. 1 MHz, it is shown that (standard) measurement uncertainty due to limited resolution of the converter can be less than 0.05 %. The effective resolution for this type of converters is introduced in order to compare theirs metrological characteristic with other converters.

ABSTRACT. This article is a detailed presentation of the VMPCalc ver. 2.1 software, which is used to support the single phase power analyzer VMP20. Also, it briefly describes the measuring method used in VMP20. Two very significant cases from the end of 2016 have demonstrated the effectiveness of programs VMPCalc ver. 2.1 and VMP20 in recognition, qualification and quantification of events in the electricity grid (EG). The software VMPCalc ver. 2.1 is the BSC thesis of the first author.

ABSTRACT. U ovom radu je predložena hardverska struktura uređaja i optimalni postupak merenja ekstremno malih prostoperiodičnih napona korišćenjem stohastičke digitalne merne metode (SDMM). Pokazano je teorijski i simulaciono da je sa relativnom mernom nesigurnošću manjom od 1 % moguće meriti predloženim uređajem prostoperiodične napone reda nV. Problem je rigorozno matematički analiziran. Nađena je jednostavna formula koja definiše teorijsku granicu detekcije AC nule u zavisnosti od parametara SDMM – kvanta dvobitnog fleš A/D konvertora, frekvencije semplovanja, vremenskog intervala trajanja merenja, veličine pojačanja merenog signala i zahtevanog intervala poverenja, odnosno faktora prekrivanja. Opsežne simulacije rada uređaja, navedene u ovom radu su, bez izuzetka, potvrdile korektnost formule i time i kompletne teorijske analize. Za granični, realan, izbor parametara SDMM teorijska granica detekcije AC nule predloženim uređajem je reda pV. Ta činjenica govori da bi predloženi merni uređaj mogao da bude i veoma dobar detektor AC nule.

ABSTRACT. U ovom radu su istražene mogućnosti Hexiwear multisenzorske platforme, sa akcentom na njenu upotrebu u merenjima u konceptu Internet of Things. U II poglavlju predstavljen je hardver Hexiwear platforme i dat detaljan opis mikrokontroleta i senzora od kojih je Hexiwear sačinjen. III poglavlje sadrži kratak opis firmvera koji izvršavaju mikrokontroleri ove platforme. Izgled aplikacije kojom korisnik upravlja izložena je u IV poglavlju sa svim segmentima i podsegmentima. Mogućnosti povezivanja ove platforme sa pametnim uređajima i skladištenje podataka na cloud-u opisano je poglavljem V.

ABSTRACT. Izmenom opšteprihvaćenih definicija efektivne (RMS) i srednje vrednosti periodičnog signala, objašnjeno je poboljšanje određivanja efektivne i srednje vrednosti periodičnog signala metodom asinhronog uzorkovanja. Simulacionom metodom je pokazano da početni momenat odabiranja signala značajno utiče na rezultat merenja. Objašnjeni su principi merenja signala asinhronim uzorkovanjem, kao i uslovi za minimalno rasipanje rezultata merenja. Pored objašnjenja merenja tradicionalnim metodama, analizirana je i objašnjena predložena metoda merenja koja se pokazala kao efikasnija. Nakon uvoda u problematiku tradicionalne metode i objašnjena rasipanja rezultata, grafički su prikazani rezultati simulacije koja daje prednost predloženoj metodi. U zaključku su razmatrane mogućnosti unapređenja tradicionalne metode merenja u praksi.

ABSTRACT. This paper describes fabrication technique for building three-dimensional (3D) microstructures in polydimethyl siloxsane (PDMS) elastomer using sacrificial microstructure of acrylonitrile butadiene styrene (ABS). 3D microstructure of ABS was made by Ultimaker 2 printer. ABS microstructure was immersed in liquid PDMS and after curing, ABS was dissolved using solvent acetone. Also we performed well-known smoothing techniques for ABS macrostructures to achieve better optical visibility.

ABSTRACT. In order to improve the efficiency of the pressure sensors characterization process at the Center of Microelectronic Technologies (CMT), a new measurement setup is developed that enables fully automatic characterization of multiple silicon piezoresistive MEMS pressure sensors. The whole experiment, including high-performance measurement of sensors output signals at several temperature and pressure values, and storage of the measured results in a computer file, requires only a minimum of operator action. The setup consists of both hardware and software modules. The hardware is partly based on equipment designed and built at CMT for this purpose, and the software is entirely developed at CMT. This paper is focused on the software aspects of the developed experimental setup. The main objective of the presented work is to make the small series production of industrial pressure transmitters at CMT more efficient. The results are also very useful for research and development of pressure sensors and instrumentation.

ABSTRACT. This paper presents results of the modeling of the thermoelectric wireless sensor network node realized with the aluminum core PCBs using SPICE. Static performances of the node under different thermal gradients are investigated considering three commercial thermoelectric modules used inside the node as thermoelectric generators. Model parameters are derived from the datasheets, and thermal and geometrical properties of the other node building elements. Simulation results are compared with those obtained by full numerical analysis and with experimental ones. An emphasis is put on the influence of the Peltier effect and Joule heating inside the thermoelectric generator on the model accuracy. It is demonstrated that the equivalent electro-thermal model of the node, in conjunction with electrical properties of the load, can be used for the estimation of harvesting efficiency of the nodes realized with different thermoelectric generators.

ABSTRACT. Ovaj rad opisuje realizaciju interakcije između troosnog akcelerometra i RGB LED diode na PSoC (Programable System on Chip) razvojnom okruženju. Opisana je konfiguracija i princip rada realizovanog sistema. Predstavljene su karakteristike korišćenog akcelerometra i razvojnog okruženja neophodne za izradu projekta. Detaljno je opisana konfiguracija korišćenih komponenata PSoC okruženja, kao i algoritam na osnovu koga je napisan odgovarajući programski kod za realizaciju projekta. Ispravnost funkcionisanja sistema je proverena očitavanjem izmerenih vrednosti ubrzanja pri poznatim radnim uslovima. Za slikovit prikaz rezultata iskorišćena je tehnika „pisanje svetlom“.

ABSTRACT. Pojačavači slike (Image Intensifier) zauzimaju značajno mesto u sistemima za osmatranje zbog svojih dobrih optičkih osobina i velike pouzdanosti. Konstruisani kao kompaktne i robusne jedinice sa baterijskim napajanjem, oni poseduju sposobnost prilagođavanja svog pojačanja uslovima osvetljenost koju obezbedjuje elekstronski sistem za napajanje. Ova osobina je ključni element za karakterizaciju celog pojačavača slike i njegova realizacija zavisi od konstruisanja napajanja i njegove funkcionalnosti. U radu će biti prikazana jedna realizacija elektronskog napajanja za pojačavač slike koji je u stanju da prati ključne karakteristike mikro kanalne ploče pojačavača i time prilagodi parametre rada karakteristikama modula pojačavača.

ABSTRACT. U ovom radu je prikazan prototip uređaja koji vizuelno prikazuje rad punotalasnog ispravljača (Grecovog spoja). Cilj izrade ovakvog uređaja je da se, nakon konačne realizacije, upotrebi kao pomagalo za razumevanje rada punotalasnog ispravljača na časovima gde se studenti sreću sa ovakvom materijom. Za generisanje sinusnih signala je upotrebljen mikrokontroler PIC18F45K22 koji je implementiran na štampanu ploču zajedno sa pratećom elektronikom. Projektovanje i simulacija električne šeme je izvršena pomoću softvera Proteus, dok je za programiranje mikrokontrolera iskorišćen softverski paket Proton IDE.

ABSTRACT. The aim of this study was to examine the correlation between the measures based on the Drawing Test (DT) in horizontal plane and the score on the shoulder-elbow tasks of the Wolf Motor Function Test (WMFT). Fourteen stroke survivors who participated in the study performed the two tests before and after the four-week rehabilitation program. The DT task included drawing a square in the predefined template and was performed using a digitizing board and a mechanical manipulandum. Two outcome measures of the DT were evaluated: movement duration and the distance of the movement endpoint from the target corner. Significant correlations were found between WMFT time and DT movement duration in contra lateral proximal to distal direction (before: r = 0.61, p = 0.019; after: r = 0.57, p = 0.034) and WMFT score and DT distance from distal ipsilateral corner (before: r = -0.69, p = 0.006; after: r = -0.58, p = 0.031).

ABSTRACT. Foot drop is common gait impairment after stroke. Functional electrical stimulation (FES) of the ankle dorsiflexor muscles during the swing phase of gait can help correcting foot drop. The aim of this study was to determine whether FESIA WALK system combined with conventional therapy is an effective intervention for improving gait speed in both acute and chronic patients with stroke related drop foot. The results show that use of FESIA WALK combined with conventional treatment for 4 weeks may improve gait speed both in acute and chronic stroke patients with stroke related drop foot and provide shifting to a higher class of ambulation.

ABSTRACT. In recent years, there exists a trend of adding sensing and communication capabilities to different objects in our environment. The area where this makes most sense is healthcare, where remote monitoring of patients can improve their quality of life and potentially allow doctors to detect symptoms that otherwise may not be perceived in ambulatory environment. In this paper, we propose an architecture of remote patient monitoring system whose main part is Android application TremorSense that is used to collect data from wireless inertial sensors and convey them to the remote server for centralized storage and analysis. The application has been tested in clinical trials in Neurology clinic with patients that suffer different movement disorders. The trials have shown that such application can be of great help to doctors that can use it in everyday procedures for quantification of patient symptoms and evaluating the disease evolution over time.

ABSTRACT. For humans with central nervous system lesion the spasticity is an important indicator of the impairment and the course of the recovery. The pendulum test was accepted as the quantification method of spasticity. We present a new, easy to use inexpensive pendulum apparatus for estimation of spasticity. The new system uses smart phone camera and markers positioned at the leg segments for measuring the knee joint movement. We compared results obtained with the new camera-based system with the system that uses knee joint angle encoder and inertial measurement units. The differences (errors) between the two systems are within the acceptable margin for clinical applications (5%). The first tests in the clinical environment suggest that the applicability of the system and the overall acceptance are appropriate (donning, doffing, setup time, precision, repeatability, ease of results interpretation).

ABSTRACT. A parameter mapping sonification method can be used to convert pathological EEG data into a sound. Sound equivalents of different EEG changes could help clinicians during EEG recording, monitoring and analysis. This method could also be used in home surveillance systems in people with epilepsy. We present an example of using a parametric based algorithm for epileptic discharge detection in two patients with different EEG abnormalities. Results include inter observer analysis of performed sonification algorithm.

ABSTRACT. Non-contact detection of heartbeat and breathing rate has great potential for various applications (health systems, sleep studies, rescue, motion detection/correction etc.). One of the most promising ways for non-contact physiological measurement is using Doppler radar technology. This paper presents the design of the system for evaluation of Doppler radar sensor capabilities used for detection of heartbeat and respiration rate. Developed environment consists of: 1) radar sensor data acquisition board, 2) software for data acquisition and logging and 3) mechanical emulator which simulates human’s movements due to breathing and cardiac activity in normal physiological range. Finally, we used the designed environment to test the custom-made Doppler radar sensor.

ABSTRACT. In this paper we present a Laplacian pyramid construction method that uses four coefficient kernel. This approach is motivated by the possible processing time reduction compared to the standard five coefficient kernel approach. The method was evaluated on clinical medical images, as a part of a radiography image processing framework. Results were compared to the ones obtained with the standard Laplacian pyramid. Experimental evaluation shows that comparable results can be obtained without introducing processing artifacts. It was shown that this approach uses 50 % less operations than the original one. Even when compared to an advanced binomial processing scheme, it was shown that operation count reduction of 12.5 % can be achieved.

ABSTRACT. Radiography image merging is used to present anatomies which are larger than the available x-ray detector. Merging is utilized to obtain long bone anatomy images composed of two or more individual images. A homomorphic approach is introduced in order to equalize statistical properties of composing images. Alpha blending is employed to perform merging and achieve gradual transition between the composites. Clinical medical images were used for proposed algorithm evaluation. Experimental results show successful merging which can be used as a part of long bone image acquisition systems.

ABSTRACT. The two basic design parameters for polynomial-based digital interpolation filters are number of polynomial-segments defining the finite length of impulse response, and order of polynomials in each polynomial segment. The complexity of the implementation structure and frequency domain performance depend on these two parameters. This contribution compares two types of estimation formulae for length and polynomial order of polynomial-based filters for various types of requirements including attenuation in stopband, width of transitions band, deviation in passband, weighting in passband/stopband.

ABSTRACT. In this paper an efficient FPGA implementation of the multistage FIR decimator is presented. The proposed implementation is based on the cascaded connection of the m decimation by 2 stages, obtaining overall decimation factor of 2^m. Each stage is designed as a half-band FIR filter decimator. The FPGA realization is verified by the simulation and by analysis of the round-off quantization noise at the output of the decimator.

ABSTRACT. High dynamic range images tend to provide more details in both dark and light areas. By introducing a high range of luminance, high quality detailed images may bring more realistic scenes. On the other hand, the change in the experience of the images may be found in comparison to traditional images. It is still not clear how high dynamic range may affect human visual attention, and that information can be relevant in different applications including compression. Fixation Density Map images can be helpful in understanding the human visual attention. In this paper, a comparison experiment with the map images is performed based on large deviation analysis for high dynamic range images.

ABSTRACT. Amplification of HER2 gene occurs in more than 25% of breast cancer cases. A drug named Herceptin is being used in breast cancer treatment for patients with confirmed amplification of this gene, which makes it necessary to determine HER2 status before the start of the treatment. This paper presents a program for automated analysis of HER2 fluorescent in situ hybridization images. Interactive graphical interface allows the user to have an effect on the determined result by manually altering different stages of evaluation process.

ABSTRACT. Attractiveness of Python program language is partially based on availability of large number of open-source libraries. In this paper we demonstrate application of one of Python libraries (BioSPPy) on a set of physiological signals from the Physionet CEBS database. Electrocardiogram and respiratory signals are analyzed, both before and after listening of classical music (states B and P, respectively). Heart rate variability analysis is performed using functions from the toolbox and by inspection of spectral aspects of the signals. Several metrics are used for comparison, like instantaneous heart rate and ratio between the spectrum segments. The obtained results show further possibilities for developing graphical interfaces for efficient vital signal analysis, and implementations based on open-hardware platforms.

ABSTRACT. U ovom radu prikazana je metodologija razvoja algoritma za procenu parametara tokom procesa hemodijalize i predikciju vremena njenog trajanja. Razvoj algoritama izvršen je na osnovu realnih merenih vrednosti bioimpedanse i indeksa telesne težine koji se u poslednje vreme veoma često koriste u dijagnostici. Na osnovu njih korišćeni su interpolacioni polinomi Lagranžov i Njutnov koji su implementirani na mikroračunaru Raspberry Pi.

ABSTRACT. Termovizijski sistemi imaju široku primenu u velikom broju vojnih i civilnih aplikacija. Primena termovizijskih sistema podrazumeva poznavanje performansi samog sistema koji se koristi, njegovih ograničenja i mogućnosti. Radi tačne predikcije, odnosno estimacije stanja u svim sistemima, modelovanje i analiza šuma koji se javlja, je od primarnog značaja. Zbog toga je predmet ovog rada analiza šuma koji se javlja u termalnoj slici. Postoji više faktora, odnosno uzroka šuma u termalnoj slici i samo razlaganje tog šuma na komponente doprinosi poboljšanju kvaliteta same slike, tako što se neke komponente mogu smanjiti ili potpuno eliminisati. Dakle, osnovna ideja jeste razlaganje šuma koji se javlja na šum koji potiče od samog detektora i na šum koji potiče od fluktuacije zračenja u sceni, što se postiže obradom podataka koji su dobijeni eksperimentalno. Eksperimenti su pokazali da je dominantan šum scene na termovizijskim slikama. Vizuelno je uočljiva razlika na objektu posmatranja slika razlike bez i sa fluktuacijom šuma. Objektivne mera, na primer standardna devijacija nivoa sivog na termovizijskoj slici razlike dva susedna frejma, za posmatrani objekat, je oko 4 puta manja od standardne devijacije slike razlike dva frejma snimljena u vremenskom intervalu 1 sat.

ABSTRACT. Mathematical model of transmission line is formulated using fractional calculus, in order to include the memory effects of capacitive and inductive phenomena, which are not taken into account in the classical model. Topological generalization of the elementary circuit is performed by introducing a capacitor parallely connected with resistor in a serial branch of elementary circuit, including the effects of charge accumulation along the transmission line. The aim of this paper is the frequency analysis of transfer function corresponding to this model.

ABSTRACT. Plasmonics and optical metamaterials offer numerous applications in different fields, from transformation optics and chemical sensing to merging the beneficial properties of electronic and optical circuits. Crucial for their function are interfaces between materials of which one has to exhibit negative value of relative dielectric permittivity due to the existence of free electron plasma. However, the choice of convenient materials is rather limited while the performance of different applications is severely impaired by strong absorption losses. This is the reason why alternative plasmonic media are currently of an increasing interest. In this contribution we consider one such media, the heterometallic multilayer consisting of copper and nickel. Copper is an excellent plasmonic material, but needs protection against surface oxidation, a role fulfilled by nickel layers which simultaneously form interfaces supporting surface waves. We describe our proposed heterometallics and consider their electromagnetic properties and experimental fabrication. Ab initio numerical simulations were done using the finite element method for Cu-Ni multilayers on a copper substrate. Laminate composite structures of alternately electrodeposited nanocrystalline Ni and Cu films on cold-rolled polycrystalline copper substrates were fabricated. Highly-densified parallel interfaces can be obtained by depositing layers at a very narrow spacing. Our results show that Cu-Ni pairs are a viable alternative to conventional plasmonic media, while the electrodeposition approach offers acceptable structural and electromagnetic parameters with large area and good uniformity at a low cost.

ABSTRACT. Plasmonic nanocomposites are a new class of materials that offers unprecedented opportunities to tailor the optical response, including the possibility to design their spectral and spatial dispersion at will. This includes the optical parameters rarely or never met in nature, which opens a path toward plasmonic metamaterials and the wide new area of transformation optics. Responsible for such a unique behavior are bound surface modes propagating along interfaces between materials with different signs of relative dielectric permittivity known as surface plasmon polaritons (SPP). Most metals possess negative relative permittivity in optical range due to the existence of free electron plasma. However, they also exhibit large absorption losses and are bound to a given spectral range defined by the metal itself, which is the reason why alternative plasmonic materials are being actively sought upon. One possible way to extend the toolbox of available materials is to use alternating metal-dielectric or metal-metal layers – the one-dimensional plasmonic crystals. Typically gold and silver are used for the metal part due to their large conductance and generally favorable properties. In this contribution we perform an analysis of the suitability of the use of copper for plasmonic nanocomposites. Its oxidation, the main barricade towards its more widespread use in plasmonics, is avoided by combining it with nickel. We utilize ab initio analysis by 2D finite element modeling and realistic material parameters to assess different electromagnetic modes. Tailorability of the response is attained by simple changing of the Cu to Ni fill factor. The analyzed Cu-Ni plasmonic crystals are convenient for simple, low cost biochemical sensors and superabsorbers.

ABSTRACT. An investigation about tunneling times in complex potentials is reported. Analyzed potential structure consists out of complex double Dirac delta potential barriers and analytical expressions for dwell time, self-interference time and group delay are obtained. It is stated that appropriate values for the set of parameters of the potential can be acquired in order for the tunneling times to achieve very large values and even approach infinity for the case of resonance. The conditions for the occurrence of tunneling times singularities, which will be given, are satisfied for only one particular positive value of the imaginary part of the potential, if all other parameters are known.

ABSTRACT. We investigate an electronic properties of square HgTe quantum dots. The semi-empirical tight binding model in sp3d5s* basis of orbitals is used for calculations. The hopping links were limited to the first nearest neighbors and spin-orbit coupling is included in the model. We demonstrate appearance of topologically protected edge states in the energy gap of square HgTe quantum dots. The results provide the basis for future calculations of optical and transport properties of edge states in HgTe quantum dots.

ABSTRACT. We theoretically investigate properties of hexagonal phosphorene nanorings with normal and skewed zig-zag and armchair boundaries. The electronic structure, density of states and density distributions of electronic states are calculated numerically by the tight-binding method. We found that edge states, with energy in the band gap close to the Fermi level, appear in both types of ring. These boundary states are found only along normal zig-zag and skewed armchair boundaries. Furthermore, there is a higher dispersion of the edge states around Fermi level in the case of ring with zig-zag boundaries, than in the case of ring with armchair edges.

ABSTRACT. Silicene is the buckled honeycomb monolayer of silicon atoms, which is topological insulator due its intrinsic spin-orbit interaction (SOI). At the edges of silicene nanoribons emerges the quantum spin Hall (QSH) effect, i.e. helical edge states appear. Employing the Kane-Mele model, we calculated the electronic structre of hexagonal silicene nanoring with zigzag edges, subjected to a perpendicular magnetic field. The results include the magnetic field dependence of energy spectrum, distributions of the spin-up and the spin-down electron eigenstates, and probability current densities of these states. We show that QSH state exist at the outer and inner edges of the nanoring. The helical edge states can be qualitatively explained using classical model of the magnetic dipole in magnetic field.

ABSTRACT. U ovom radu smo razmatrali elektronska i transportna svojstva dvoslojnih nanotraka od fosforena sa običnim cik-cak ivicama. Elektronska struktura je računata pomoću modela jake veze. Transportne osobine su razmatrane u smislu promene konduktanse, koja je računata za režim balističkog transporta prema modelu neravnotežne Grinove funkcije u okviru formalizma metoda jake veze. Ustanovljeno je da male promene eksternog električnog polja dovode do željene promene transportnih svojstava traka, što nije bio slučaj kod jednoslojnih struktura.

ABSTRACT. The state of the art research and development of the powered prosthetic devices controlled via neural interfaces are aiming at the problems that are preventing the natural-like use of an artificial limb. Although in wide use, myoelectric prosthetic hands, interfaced via stumps and controlled using superficial EMG electrodes are known for their poor functionality, controllability and sensory feedback, mainly due to the use of surface electrodes. Furthermore, the interface with the user is established through the soft interface that deteriorates over time. In this paper, we will discuss developing of a novel prosthetic hand with improved functionality, smart mechatronic devices for safe implantable technology, and improved paradigms for natural control (action) and sensory feedback (perception) of the prosthesis through the Osseointegrated implant.

ABSTRACT. We present the system for evaluation of myoelectric control with electrotactile feedback in the closed-loop environment. This test bench was implemented in MATLAB Simulink, and relies on Mujoco HAPTIX virtual hand simulator and the Maxsens system for electrotactile stimulation. We selected a set of dynamic stimulation patterns to provide the user proprioceptive feedback from the multi-DOF virtual prosthesis regarding hand aperture and wrist rotation. The patterns were spatially coded and delivered to the user through a custom designed array electrode. Mujoco HAPTIX virtual reality hand simulator provided visualization of the prosthesis and task information. The results from the pilot test on one healthy subject suggest that the proposed dynamic patterns can be recognized with a high success rate and can be successfully exploited for controlling the extent of aperture and rotation of the multi-DOF myoelectric prosthesis.

ABSTRACT. Functional electrical stimulation (FES) cycling can be used as a rehabilitation tool and also a recreational activity for spinal cord injured (SCI) patients. It provides cardiovascular exercise, increases muscle mass and therefore lowers the incidence of secondary diseases associated with paralysis. However, FES cycling is limited by the power that can be produced and the rapid muscle fatigue that occurs. This study proposes a systematic method to determine customised cycling stimulation patterns, by assessing the contribution of specific muscle groups, in order to optimize muscular synergies and maximize recruitment. We measured the tangential forces exerted by the pilot’s feet on the pedals during the cycling process, and calculated the resulting power output for a quadriplegic patient on a commercial cycling device. We found that our pattern is more efficient than the recommended pattern from the literature. Thus, direct pedal force measurement appears to be a relevant method to optimize the muscle stimulation pattern during FES cycling.

ABSTRACT. We developed a method based on an artificial neural network for the estimation of gait parameters from data recorded by inertial measurement units (IMU) mounted on the shank and foot. The input for the training of the neural network are the signals recorded by inertial measurement units (Yost Labs, Portsmouth, Ohio, USA) positioned at the foot and the shank, and the output of the network are the fuzzified data from the sensors recorded by the force transducers built into the insoles. The communication between the sensors and the computer was realized wirelessly. The input and output data were captured synchronously on a Windows platform during the gait. The differences between the gait parameters (relative errors) estimated from the reference data and the IMU system were below 4%. The system is now being used in a small clinical trial in stroke patients.

ABSTRACT. Izvori jednosmerne struje predstavljaju krucijalnu funkcionalnu komponentu uređaja za jontoforezu. Dostava potentnih ili osetljivih naelektrisanih formulacija jontoforetskim pumpanjem zahteva kontrolabilnu primenu električnih struja niskog intenziteta. U ovom radu je prikazana jednostavna i jeftina realizacija programabilnog izvora konstantne jednosmerne struje niskog intenziteta zasnovana na Arduino Uno platformi. Uređaj se sastoji iz hardvera za kontinualnu generaciju i monitoring struje, dok implementirani softver omogućava dinamičku regulaciju vrednosti struje i jednostavan korisnički interfejs. Strujni izvor je pokazao zadovoljavajuće performanse u opsegu intenziteta struja od 10 do 230 μA i demonstrirana je njegova upotrebljivost u rutinskim in vitro laboratorijskim testovima jontoforeze. Konstruisani strujni izvor može biti koristan alat za istraživače jontoforeze različitih profila.

ABSTRACT. U ovom radu je predstavljena inovativna realizacija jeftinog programabilnog pulsnog strujnog izvora namenjenog laboratorijskim in vitro ispitivanjima procesa jontoforeze. Uređaj je zasnovan na Arduino UNO platformi i LabVIEW korisničkom interfejsu čime je omogućena potpuna kontrola parametara pravouganih strujnih impulsa kao što su faktor ispunjenosti, amplituda koju je moguće varirati u opsegu od oko 100 µA do oko 4.5 mA i učestanost čiji opseg seže od 1 Hz do 10 kHz bez značajne degradacije oblika signala. Digitalna kontrola strujnih impulsa je ostvarena primenom JFET tranzistora kao naponski kontrolisanih otpornika. Izvršena je karakterizacija performansi pulsnog strujnog izvora pri različitim rezistivnim opterećenjima. Strujni izvor je testiran i u realnim eksperimentalnim uslovima u laboratoriji čime su demonstrirana njegova odlična svojstva, efikasnost i upotrebljivost u istraživačkoj praksi.

ABSTRACT. Within the last ten years, we have seen an enormous expansion of European robotics moving from manufacturing and automation into many other industrial and professional service sectors, mostly in agriculture, health, logistics, mining, and public services. Improved technologies like sensor-based collaboration with humans and growing autonomy on one side, and, on the other side, steadily decreasing investment costs have made this possible – together with a European mix of first-class research, the ability for interdisciplinary work, and public funding from the European Commission. In order to improve the transfer from laboratories to applications, the Commission engaged in 2013 into the Public Private Partnership “SPARC” with the European robotics association euRobotics. Under the umbrella of euRobotics, more than hundred experts participate in over 20 Topic Groups in shaping the European Roadmap for Robotics which served the European Commission as a strategy for the robotics programme under Horizon2020. So far there is evidence that H2020 has stimulated a very active European robotics community with substantial growth rates observed by both users and producers of robots. For the remaining years under H2020, the strategy of the Commission focusses much more on (i) the integration of robotics into the DEI initiative (Digitizing European Industry), and (ii) regional Digital Innovation Hubs (DIH) as a consequence of the very unbalanced geographical situation of robotics in Europe. DIHs are to stimulate innovation, entrepreneurship, the creation of jobs in regions which have the potential, but suffer from non-technical barriers to evolve. Robotics is seen as having high potentials for the regional development to the benefit of its economy and society. This talk will present an overview on the recent developments of robotics in Europe and present the strategies currently discussed in the Commission which may have high potentials for Serbia.

ABSTRACT. Development of reliable and efficient material transport system is one of the basic requirements for creating an intelligent manufacturing environment. Nowadays, intelligent mobile robots have been widely used as one of the components to satisfy this requirement. In this paper, a methodology based on Grey Wolf Optimization (GWO) algorithm is proposed in order to find the optimal solution of the nondeterministic polynomial-hard (NP-hard) single mobile robot scheduling problem. The performance criterion is to minimize total transportation time of the mobile robot while it performs internal transport of raw materials, goods, and parts in manufacturing system. The scheduling plans are obtained in Matlab environment and tested by Khepera II mobile robot system within a static laboratory model of manufacturing environment. Experimental results show the applicability and effectiveness of the developed intelligent approach in real world conditions.

ABSTRACT. This paper presents the realization of a manufacturing cell developed at the University of Belgrade, School of Electrical Engineering. The manufacturing cell is intended for practical education and research in the field of flexible automation and industrial robotics. It consists of two industrial robots with shared workspace, two conveyor belts, various sensors for object detection, automated storages for workpieces with pneumatic actuators and controller with a user interface. The industrial robots can be used independently for different tasks, or in cooperative action in order to perform assembly, dual-arm manipulation with workpieces and other complex tasks. The versatility of the robotic cell introduces additional unpredictable potential problems, therefore a standalone safety system was introduced to significantly decrease the probability of hazard occurrence. After a hazard identification and risk assessment, the safety system is realized with separate controller and user interface, electromagnetic, electromechanical sensors and laser curtains in order to achieve highest standards in the field of safety.

ABSTRACT. Multi-robot systems can be considered as a suitable platform for dangerous mission in accident situations. In this paper we propose a multi-robot system with distributed intelligence that can perform complex mission faster and more reliable. Robots are connected to the cloud and can benefit from the powerful computational, storage and communication resources of the cloud, which processes and shares information. Our robot team consists of one robot scout and two robot manipulators with different sensory capabilities. The mission goal is to, in conditions of limited resources and insufficient information, collect enough data about accident situation, identify the most critical points and calculate appropriate path of robots in known or unknown environment where accident occurred.

ABSTRACT. This paper investigates distributed approach to control of redundant robotic manipulator. As opposed to usual control based on knowing the whole model of the kinematic chain, idea of distributing the error information to local controllers is used. Standard approach is not as flexible to modifications and changes to kinematic chains, as distributed is. Also with multi agent approach even though no individual robotic joint knows the whole model, its communication with other agents allows them to take part in minimization of tool to goal error. By combining of reactive and cognitive agents we allow for localized adjustments to environment as well as global goal achievement. Simulation results and conclusions from a planar 4 degrees of freedom model have been presented.

ABSTRACT. This paper presents an approach to identification of parameters for modeling in robotics using Denavit-Hartenberg notation. It describes an automated procedure for obtaining parameters of industrial robots, with attention to implementation aspects. The procedure uses partial pose measurements, and it is adaptable to various configurations of manipulators, with different numbers and types of joints. Potential for industrial and practical application of the presented approach is considered together with its advantages and disadvantages.

ABSTRACT. Paper presents one simple control algorithm for specific mobile robot system. The hardware architecture of the buggy robot and the use of certain types of sensors have been reviewed. The implementation of the movement of a mobile robot on a flat surface is presented and certain defects of applied robotic platforms have been discussed. A software library for sensors applied in this particular case has been established. Improvements of control structures by using intelligent systems and the data obtained from the sensors have been recommended. At the end of the paper there is a presentation of the results and potential further development suggestions.

ABSTRACT. Radioactive waste has been produced in the past during the operation of nuclear power plants. This waste has to be stored partly for a very long time in suitable storages. To reduce the waste volume and to condition the waste for the long time storage waste treatment centres are taken into operation. During the earlier time the waste was dumped into sub ground vaults inside or outside of buildings. Because the dumping has been made without any volume reduction and because the vaults are nowadays estimate as a non safe storage in respect to the environment it is of high importance to retrieve this waste and send it for further treatment to a waste treatment centre. The first step is to erect the Solid Waste Retrieval Facilities (SWRF), placed close to the waste vaults or to the buildings erected above the vaults. The purpose of the SWRF is to extract waste from its present storage location, pre-sort it, segregate adequate material for Landfill disposal, and package non Landfill material for transport to the Solid Waste Treatment Facility (SWTF). The SWRF consists mainly of the so called retrieval units and a control building. The retrieval units are cells within which the recovery, pre-sorting, segregation for Landfill, and packaging for transfer to SWTF will take place. The control building will house all common facilities including changing room, sanitary facilities and the SWRF control room. The advantage of using different retrieval units is that it will give operational flexibility to the facility. It will be possible to modulate the output of each unit to suit the SWTF. Initial retrieval will be undertaken by a dedicated crane and grab which will deposit the recovered items into a “pre-sorting” skip. The operation of a typical retrieval cell starts with the sand pre-sorting by the skip which has a vibrating grid to allow sand to fall into a catchment tray underneath where it will be collected by vacuum extraction and placed into “big-bags”. The waste items dedicated for treatment in the SWTF will be loaded into a container docked in a special loading bay. This arrangement allows the external surfaces of the container to remain contamination free during loading. After monitoring (and registering by the tracking system) the container is moved to the SWTF under the control of the main SWMSF control room. Another solution is the use of a recovery unit which is mobile and located on top of the building. The benefit of this arrangement is that it is much safer than “tunnelling” into the side of the deeper piles of waste present in this building. First all accessible waste is recovered by the crane, after this a Remote Operated Vehicle (ROV) will be lowered into the compartment to load the remaining waste into a skip operated by the crane. To access the vault holes have to be cut into the roof of the buildings. If the content of the vault is very high active waste, then special precaution has to be made to avoid that the cores do not fall into the vault. The technique developed for this problem consists of three steps. In the first step a small hole is made where in the second step an inverted tripod system is introduced. With the tripod the roof of the vault is supported from the lower side to be stable during the cutting of the large hole as needed to introduce later a ROV. The coring tools are internally fitted with “one way dogs”, which allow the core to go in but prevent it from falling when it is cut. When all the peripheral holes are cut, the remaining part can be recovered. In the paper the techniques and gathered experience including their advantage will be presented in details.

ABSTRACT. Prior to performing calibrations in the Secondary Standard Dosimetry Laboratory (SSDL), standard radiation qualities must be determined in order to accurately assess the absorbed dose of occupationally exposed personnel and to evaluate the performance of radiation measurement instruments. Due to the energy dependence of the calibration factor and the air kerma to the personal dose equivalent conversion factor, it is of great importance to properly establish standard radiation qualities. Radiation qualities can be characterized by determining their photon energy distributions, or by using the X-ray tube voltage and the first and second half-value layers. In this paper, the half-value layers were determined by varying the thicknesses of the additional filtration absorbers. The calculated values show a deviation from the ISO standard values not greater than 6%.

ABSTRACT. In the era of the energy saving policy increases the importance of the measurement of radon exhalation from building materials. We present a new method of the radon exhalation measurement using only aHPGe detector or any other gamma spectrometer. This method provides the measurement of the emanation coefficient, the radon diffusion length and the radon exhalation rate, all within the same measurement, which additionally defines material's radon protective properties. It does not necessitate additional equipment for radon or radon exhalation measurement, which simplifies measurement technique, and thus potentially facilitates introduction of legal obligation for radon exhalation determination in building materials.

ABSTRACT. This paper presents actual occupational external and internal exposure monitoring techniques and methods being used in the Public Company “Nuclear Facilities of Serbia”, the only nuclear operator in Serbia. This monitoring system consists of a set of stationary multifunctional Geiger-Mueller monitors installed on site, hand-held instruments, electronic and thermo-luminescent personal dosimeters for external, alongside with whole body counter and gamma spectrometry system for internal exposure control. Additionally, the paper shows representative results of the occupational exposure monitoring for 2016, under the Radiation Protection Programme.

ABSTRACT. Determination of the angular dependence of thermoluminescent (TL) and optically stimulated luminescent (OSL) dosimeters is of great importance for their application in poly-energetic multidirectional photon radiation fields. In order to determine the angular dependence of the dosimeter responses for different photon energies, the dosimeters were mounted on an ISO water slab phantom, irradiated in two narrow-beam X-ray qualities (N-40 and N-80) and in one gamma ray quality (S-Cs), defined in IAEA SRS 16. The angles of incidence ranged from 0° to 60°, with an increment of 20°. Two types of TL dosimeters were used: MCP-N (LiF:Mg,Cu,P) and MTS-N (LiF:Mg,Ti) and one type of OSL dosimeters, InLight (Al2O3:C). The two types of TL dosimeters have shown similar deviations from their 0° responses for all the used radiation qualities, while the OSL dosimeters have shown a greater deviation from their reference values for the gamma radiation quality.

ABSTRACT. During exploitation of the RA heavy water research rector, a high average fuel burnup (7000 MWd/t) was achieved in low enriched (2% of U-235) metal uranium fuel elements. As a result of high fuel burnup and complex fuel element construction (cylindrical shell with aluminium clad), a high leakage of Cs-137 in the spent fuel storage pools was detected. According to measurements (from 2012), total activity of Cs-137 in spent fuel storage is ascertained to 2 TBq. Until to 2008, ventilation system of spent fuel storage did not have HEPA filters. There was a necessity to define the state of environmental contamination before beginning of the RA nuclear reactor decommissioning projects, and the essential part of this evaluation is determination of the activity of Cs-137 and other artificial radionuclides deposited in the soil around this reactor. This paper presents the results of radionuclide activity measurement in 44 soil samples taken in the vicinity of the RA and RB reactor buildings, and summarises the comparison with results obtained for control group comprised of 20 samples of uncultivated soil taken from various locations in Serbia. Statistical analysis of these results shown that there is a significant increase of Cs-137 activity in the soil around the RA and RB reactor buildings (in average 0.1 Bq/g), though in the order of clearance level magnitude.

ABSTRACT. Садржај – Јануара месеца 2015 године у САД је прихваћено покретање поступка могуће замене тренутно важећег LNT модела у заштити од зрачења, HORMESIS моделом. Разлика између ових модела је што према првом, свако додатно излагање зрачењу, без обзира на ниво, је опасно, док према другом је супротно – не само да није опасно већ може бити и корисно, ако је довољно ниско. Истовремено у свету се све чешће помиње и нуклеарно оружје са нагласком на његовом усавршавању. При чему се под усавршавањем подраузмева израда прецизнијег оружја мање снаге односно, мањег радијуса накнадне опасности по становништво погођене локације. Поставља се логично питање да ли се ове две акције, које су покренуте приближно у исто време, могу међусобно повезати. Одговор би могао бити пре позитиван него негативан, Јер, усвајањем HORMEZIS модела несумњиво би се олакшало прихватање, од стране становништва, употребе нуклеарног оружја у специфичним хипотетичким околностима. У раду су приказане основне карактеристике и научна заснованост поменутих модела заштите од зрачења, као и неки доступни подаци о расположивости и распрострањености нуклеарног оружја у свету данас

ABSTRACT. In this paper we elaborate on selection of hardware devices that are most convenient for use in active SWIR/NIR vision systems as source of illumination for generating image with SWIR/NIR sensors. Scenario of application in which active imaging is used in surveillance systems is explained and different implementation approaches are reviewed. Requirements for system architecture are specified and an estimate is given for expected systems performances comparing to passive solutions. Guidelines for selection of appropriate system components are given. We also propose our concept for control electronics implementation based on FPGA.

ABSTRACT. In this paper we explain properties of imaging in different spectral regions that are used in multi-sensor imaging systems for long range surveillance and analyze scenarios in which additional short wave infrared SWIR imaging sensor could improve its performances. Analysis of state-of-the-art SWIR components and additional functionalities of modern SWIR detectors based on InGaAs technology are given. We have implemented multi-sensor imaging system with imagers in LWIR, MWIR, VIS and SWIR spectral region and generated simultaneously captured examples of images for two scenarios of medium and long range imaging and explained the results and advantages of SWIR camera that are obtained.

ABSTRACT. In this paper we analyze integration challenges of an Electro-Optical device and a laser range finder into functional system. The laser rangefinder design and basic properties are reviewed. The key factors influencing reliability of the laser rangefinder based distance measurements are discussed using a simplified parametric performance model, fault tree analysis and failure mode error analysis we discuss possible sources of failures and their influence on field distance measurements capabilities in the multi sensor surveillance systems.Based on our practical experience, the recommendations for hardware and software integration and laser rangefinder selection are derived.In this paper we analyze integrationchallenges of an Electro-Optical device and a laser range finder into functional system. The laser rangefinder design and basic properties are reviewed. The key factors influencing reliability of the laser rangefinder based distance measurements are discussed using a simplified parametric performance model, fault tree analysis and failure mode error analysis we discuss possible sources of failures and their influence on field distance measurements capabilities in the multi sensor surveillance systems.Based on our practical experience, the recommendations for hardware and software integration and laser rangefinder selection are derived.

ABSTRACT. Video signal processing on various platforms is very common today, while the number of algorithms that use video signal as a source is still increasing. The problem occurs when the video signal from variety of cameras through various interfaces has to be imported to the platform for video processing. The usual approach is to use separate device that converts video signal from the source interface to the one that exists on the processing platform. The purpose of this paper is to describe various communication interfaces and propose the solution for their integration in the systems for video signal processing, with respect to the appropriate video standards. Detailed review of relevant standards is given. We propose the system, based on the last generation of FPGA device, including its model and implementation concept. This unified solution avoids the need for different converters and decrease the overall system price.

ABSTRACT. The dehazing problem where different atmospheric particles brings certain amount of noise to images is very important in outdoor video surveillance systems. This paper compares the most significant single image dehazing approaches, proposes additional enhancement step in dehazing algorithms, and presents test results on maritime surveillance images that represents one group of long-range images. This paper also considers usage of described method for real-time video surveillance systems.

ABSTRACT. U ovom radu je opisan postupak adaptacije stripmap SAR algoritama za formiranje radarskih slika. Realizovana je simulacija SAR sirove slike, nakon čega su primenom Doppler Beam Sharpenning (DBS) i Range Doppler (RD) algoritama formirane radarske slike generisane scene. Da bi glavne karakteristike kao i nedostaci navedenih algoritama došli do izražaja, u programu MATLAB su generisani ciljevi na jednakim udaljenostima po daljini i azimutu. Objašnjena je pojava varijacije dometa prilikom SAR snimanja, kao i nedostaci koji nastaju pri formiranju radarskih slika. Analiziran je uticaj varijacija daljine na Doppler Beam Sharpening i Range Doppler algoritme. Definisani su uslovi koji treba da budu zadovoljeni da formirana radarska slika ne bi imala određenih degradacija. Objašnjene su i primenjene adaptacije navedenih algoritama za formiranje radarskih slika u scenarijima kada dolazi do prekoračenja navedenih uslova. Na kraju je izvršeno poređenje slika dobijenih adaptacijom algoritama, sa izvorno formiranim radarskim slikama.

ABSTRACT. У раду је дата упоредна анализа објективних метода за процену квалитета слика са ниским динамичким опсегом насталих од слика високог динамичког опсега. Представљене су три мере, од којих је једна изворно намењена процени квалитета мапираних слика, док су преостале две генералне мере без референцирања. Мере су тестиране на једној јавно доступној бази слика. Перформансе објективних мера за процену квалитета слике су представљене кроз корелацију са субјективним скоровима.

ABSTRACT. U radu je analizirana mogućnost primene mapa vizuelnog značaja u objektivnoj proceni kvaliteta slika. Analiza je sprovedena na sedam javno dostupnih baza slika sa subjektivnim impresijama kvaliteta i sa reprezentativnim brojem vizuelnih signala. U okviru baza nalaze se podskupovi slika nastali primenom modifikacija karakterističnih za obradu i prenos vizuelnih signala. Pokazano je da se uvođenjem mapa vizuelnog značaja mogu poboljšati performanse objektivne procene kvaliteta zasnovane na očuvanju informacija o gradijentu.

ABSTRACT. U radu je prikazano međulaboratorijsko poređenje merenja napona smetnji na mrežnim priključcima. Međulaboratorijsko poređenje su pokrenule i organizovale dve akreditovane laboratorije s ciljem da potvrde svoju tehničku kompetentnost za merenje navedene veličine.

ABSTRACT. Predmet ove analize je etaloniranje i određivanje merne nesigurnosti etaloniranja kalibratora DC struje u Sektoru za metrologiju u Tehničkom opitnom centru. Dat je primer proračuna merne nesigurnosti etalonoranja DC struje 1 A za kalibrator Fluke 5500A na dva načina.

ABSTRACT. U radu je opisan postupak merenja koeficijenta refleksije skalarnim analizatorom mreže uz detaljan proračun merne nesigurnosti merenja. Opisana metoda koristi se u laboratoriji za mikrotalasnu tehniku Tehničkog opitnog centra, prilikom etaloniranja merne opreme iz oblasti mikrotalasne snage, slabljenja i impedanse.

ABSTRACT. Prilikom etaloniranja uređaja iz RF i oblasti mikrotalasa, javlja se potreba za merenjem apsolutne vrednosti RF/mikrotalasne snage. Postupak merenja ove snage zavisi od njenog nivoa, kao i od tipa uređaja koji se etalonira (DUT). U radu su predstavljena četiri slučaja metode merenja apsolutne vrednosti snage kontinualnog signala u frekvencijskom opsegu od 9 kHz do 26,5 GHz, u skladu sa tehničkim mogućnostima TOC-a. Takođe su dati odgovarajući proračuni merne nesigurnosti.

ABSTRACT. U ovom radu opisan je postupak etaloniranja manometara sa Burdonovom cevi na radni fluid ulje primenom metode definisane u direktivi DKD-R 6-1 Calibration of pressure gauges. Dat je proračun merne nesigurnosti za određenu klasu manometara koji se koristi u radnom procesu i ispitivanjima koja se sprovode u Tehničkom opitnom centru.

ABSTRACT. U radu je opisano ispitivanje sinhronizovanosti podataka koji su uzorkovani putem odvojenih MEMS akcelerometara. Korišćeni su bežični senzorski uređaji razvijeni na Građevinskom fakultetu u Beogradu, namenjeni merenju vibracija građevinskih objekata. Poznato je da ovi uređaji mogu da postignu tačnost vremenskog označavanja uzoraka reda veličine 10 μs za kraće merne periode (ispod minuta). Ovo znači da je greška koju elektronika unosi zanemarljiva za merenje mehaničkih veličina sa frekvencijom uzorkovanja reda veličine kHz. Analizirane su varijacije u fazi izlaznih signala različitih senzora prilikom identične pobude. Problem je ograničena mogućnost ostvarivanja identične mehaničke pobude na različitim mestima u prostoru, te je bez savršene opreme nemoguće razgraničiti koliki deo greške potiče od nepoklapanja ulaznih signala, a koliki od nesavršenosti senzorskog aparata. Pokazano je da je moguće dobiti standardnu devijaciju vremenskog ofseta između snimljenih signala reda veličine milisekunde.

ABSTRACT. U radu je prikazana procedura za eksperimentalnu verifikaciju pojačavača signala koji se može koristiti u postupku testiranja uređaja za merenje standardnih parametara kvaliteta električne energije. Realizovano rešenje pojačavača uključuje niskopropusni filter, predpojačavač i pojačavač snage, kojima se obezbeđuje odgovarajući naponski nivo od 230 V. Test signali za eksperimentalnu verifikaciju pojačavača obezbeđuju se pomoću računarski podržanog generatora signala, koji je baziran na LabVIEW softverskoj platformi i akvizicionoj kartici NI PCIe 6343. Ovaj generator omogućava generisanje naponskih signala sa različitim poremećajima kvaliteta električne energije, tipičnim za realne elektrodistributivne sisteme. Kao referentni instrument za merenje parametara kvaliteta signala na izlazu pojačavača primenjen je analizator kvaliteta električne energije Fluke 435. U ovom konkretnom slučaju generatorom signala obezbeđeno je više eksperimentalnih test signala sa različitim poremećajima kvaliteta. Neki eksperimentalni rezultati dobijeni referentnim mernim instrumentom prikazani su i analizirani u ovom radu.

ABSTRACT. Za potrebe merenja preciznih strujnih šantova u industrijskim uslovima, razvijen je kompozitni miliommetar na bazi Kelvinovog četvorožičnog merenja malih otpornosti. Sistem se sastoji od preciznog, temperaturno stabilizovanog strujnog izvora i komercijalnog multimetra srednje klase sa rezolucijom od 4 1/2 cifre, kojim se meri napon. Specifični zahtevi vezani za industrijsku proizvodnju i eksploataciju preciznih strujnih šantova, uslovljavaju da standardne metode i uređaji za merenje miliomskih otpora nije moguće primeniti. Korišćenjem kompozitnog etalonskog otpornika, obezbeđena je temperaturna stabilnost preciznog strujnog izvora baziranog na naponskoj referenci REF102. Pri merenju otpornosti strujnog šanta od 20 milioma, postignuta je tačnost od 10 mikrooma.

ABSTRACT. U ovom radu je predstavljen novi pristup razmatranja i izračunavanja kvaliteta kvantovanog ECG signala. Ovaj pristup se zasniva na podeli celog signala na intervale koji nose suštinski bitnu informaciju o zdravlju srca tj. pacijenta, i na intervale signala koji ne nose bitne informacije. Intervalima koji nose bitne informacije se daje veći značaj kako pri obradi, tako i pri izračunavanju parametara kvaliteta obrađenog signala. Povećanjem tačnosti parametara kvaliteta signala, povećava se i pouzdanost procene kardiologa pri određivanju dijagnoze pacijenta.

ABSTRACT. Inspekcija proizvodnje i uklanjanje defektnih uzoraka postaje neophodna u savremenom poslovanju, usled težnje za isporukom skoro savršenog proizvoda. U radu je predstavljena mogućnost realizacije pristupačnog laserskog skenera, koji ne prelazi petinu cene komercijalnih sistema. Ostvarena brzina skeniranja zavisi od mogućnosti same kamere, jer brzina izdvajanja profila ne prelazi jednu milisekundu, pri čemu se visinska razlika od 1 mm predstavlja sa minimalnom promenom od 30 piksela u slici.

ABSTRACT. Developing a gripper that can realize grasping of objects of different shapes and sizes, as well as objects with different stiffness, represents a challenging task. Compliant (elastic) mechanisms with embedded actuators and sensors are one way to realize a gripper which could adapt its grasping surface to objects of a different shape, thus achieving safe grasping and manipulation. This paper presents a new concept of soft/adaptive two-finger robotic gripper with embedded active elements (actuators and sensors). By using embedded actuation gripper can actively morph the shape of its grasping surface and realize different grasping patterns. By using embedded sensing gripper can detect contact with objects and via appropriate controller decide how to change the shape of the grasping surface and adapt to different and unknow shaped objects. Novel design method for obtaining adaptive grippers is also presented. It is shown that robotic gripper can achieve different complex grasping patterns via embedded actuators i.e. via combination of contracting and extending actuators that realize different stroke.

ABSTRACT. This paper presents new, inspired by nature design of soft robot arm for infrastructure inspection. The robot arm is hyper redundant, under actuated, with 20 DOFs and with gripper as end effector. Robot arm is powered by only 9 servo-motors. The power transmission from the actuators to robot links and further to the end-effector is realized intra structurally by wires. Controllability and dexterity of the soft robotic arm is verified by model simulation before implementing control algorithms to the robot controller. For the purpose of simulation the algorithms of the inverse kinematics are realized. Mechanical prototype in its’ early phase of integration is shown in this paper, too.

ABSTRACT. This paper is presents modeling approach of elastic F type Cable - Suspended Parallel Robot, eFCPR system with one mode. This topic requires studious approach and detection of specific phenomena, which needs to be defined through the process of mathematics modeling. The characteristic eFCPR structure is characterized by: shape of the work space, number of motors, number of hanging points, total number of ropes for the implementation of the CPR system, number of ropes from the camera carrier to the hanging points, type of motor, type of winch. Each of these characteristics differently affects the response of the CPR system. The relations between the motor and the camera carrier motion are highly important for the kinematic and dynamic modeling of the eFCPR system. Two scenarios have been used for the eFCPR system analysis with one mode by OGIFLEX program package.

ABSTRACT. This paper proposes a novel modal architecture of a conversational robotic agent. Two important characteristics of this architecture are that it is general (i.e., it does not depend on specification requirements for a conversational agent) and modular (i.e., it allows for introduction of modules in a general manner). It consists of a bus that serves as a backbone of the system, and an arbitrary number of modules. All modules use the same plug-in interface to the bus system, so the design of the architecture does not depend on the design of modules. The communication between modules is asynchronous and based on the principle of event bus. Finally, the architecture is illustrated for a particular robotic system.

ABSTRACT. Paper presents the development of a humanoid robot Sara and dilemmas that have accompanied the development. Robot should represent anthropomorphic mobile platform for the research of social behavior of the robot in the immediate human environment. This implies several significant dilemmas – the need to enable verbal and nonverbal communication, cognitive system in all aspects of use, problem of motion and orientation of robots in unstructured environment, manipulation and handling objects, realization of complex sensor system and the need to enable energy autonomy of the robot. Sara will be able to communicate verbally and nonverbally. To express emotion by face, the robot will be able to move the eyes and eyelids. To extend the spectrum of nonverbal communication, the robot will be able to shrug when the question is confusing or when the robot does not know what to answer. In addition, Sara will have two anthropomorphic arms with a total of 14 DOFs, neck with 3 DOFs and lumbar spine with 7 DOFs in order to increase the mobility of the upper body without moving the lower.

ABSTRACT. In this paper, we present a new approach for monitoring of patients during recovery after stroke. We propose quantitative measurements to characterize patients’ movements. The focus is on the large range upper body movements acquired using Kinect device. Finally, we develop a software application for visualization and interpretation of the collected sensor data and calculated measurements. The application is intended to support the clinical evaluations by medical doctors and to store the patients’ data over time. We record patients’ scores during their scheduled rehabilitation sessions. Based on the collected scores, we build the personal profile for each patient that gives insight into the movement performance over time.

ABSTRACT. Napredak u servisnoj robotici neizostavno je doveo do pojave konstrukcija koje osim funkcionalnih zahteva nameću i estetske. Takvi roboti, ukoliko su u neposrednom ljudskom okruženju poprimaju sve izražajnije humanoidne karakteristike. Prepoznavanje emocija preraslo je u emocionalno izražavanje. U početku napravljene grube delimično pokretne maske presvlače se veštačkom kožom. Kada se ta koža ne određenim mestima poveže sa mehanizmima mogu se postići dovoljno prepoznatljive gestikulacije koje će biti usklađene sa ponašanjem osobe sa kojom je robot u interakciji. U radu su predstavljena neka idejna rešenja mehanizama za ostvarivanje pokreta koja su bazirana na oponašanju anatomije ljudskog lica.

ABSTRACT. This paper presents some innovative antenna designs with enhanced performances for the next generation of wireless networks. High-gain printed microwave and millimeter wave antennas, both linearly and circularly polarized are discussed as well as reconfigurable circularly polarized antenna with switchable sense of polarization. Special attention has been devoted to the design of frequency scanning antennas based on matamaterial-inspired phase shifters. All presented antennas demonstrate considerable improvement of the efficiency, bandwidth of operation, axial ratio and frequency sensitivity. Also, they are very simple for manufacturing even at millimeter wave frequencies.

ABSTRACT. In this paper a closed form expression for the two-dimensional Green’s function for a truncated wedge is derived. The method of separation of variables in Laplace’s equation is used to get the function in the form of an infinite series of suitible harmonics and then the series is summed up in a closed form. Based on the obtained result the capacitance per unit length of a thin line conductor inside the wedge evaluated.

ABSTRACT. A new method for determining the direction of arrival (DOA) of received signals from two spatially independent radiation sources, using linear antenna array composed of two elements, is proposed in this paper. The method is based on a hierarchical model composed of the Radial Basis Function Neural Networks (RBFNNs), designed to detect the spatial distribution of the radiation sources and to estimate the DOA of signals, received with the linear antenna array. Analysis of the results shows that it is possible to perform the accurate DOA estimation of received signals, using the linear antenna array composed of the same number of elements as the number of the radiation sources, without affecting the quality of the obtained results, which is not possible using conventional super resolution algorithms.

ABSTRACT. Microwave tomography is an ill-posed inverse problem. It is a non-linear multimodal optimization problem. As such, microwave tomography is highly computationally intensive and needs advanced optimization methods to be solved quickly. In contrast to typical approach based on faster local optimization coupled with linearization or regularization methods, the solution presented in this paper uses adaptive global optimization methods for more precise resulting images. Also, proposed solution uses time-domain solver to obtain measurement on multiple frequencies, for easier solving of the ill-posed problem. A two-dimensional numerical prototype is implemented. Few optimization methods are compared, yielding the covariance matrix adaptive hybrid method to be most effective.

ABSTRACT. In this paper we exploit a model of combline filters with simple uncoupled transmission lines to investigate the impact of adjacent and nonadjacent couplings to stopband transmission zeros. Furthermore, we present a way to control the slope of the passband skirts by adding appropriate nonadjacent couplings. These effects are also observed in 2D and 3D simulation models and verified experimentally on a combline filter with 5 resonators.

ABSTRACT. Analysis-time requirements of a method for analysis of axially symmetric metallic antennas are considered when using 1) modified and 2) max-ortho basis functions. Several improvements in implementation of max-ortho basis functions are inspected in order to increase the efficiency of the algorithm and reduce the matrix fill-in time. Benefits of these improvements are shown on the example of a thick dipole antenna.

ABSTRACT. The corona discharge in the cylindrical geometry is numerically simulated for the negative lightning voltage impulse. The simulation is performed by solving the drift-diffusion-reaction equations for the electrons, the positive and the negative ions. The results for the concentration of charge particles and the intensity of the electric field are used to determine the corona current which is in turn necessary for the calculation of the transient overvoltages on the transmission lines.

ABSTRACT. Abstract—The paper presents a new class of printed antenna structures that have numerous advantages over conventional microstrip antennas: dipoles operating on or near the second resonance, printed pentagonal dipoles whose shape enables broadbandedness, obtaining wide range of impedances, and reducing parasitic couplings and losses to a great extent due to balanced feeding structure. Various feeding structures are also presented: symmetrical microstrip, coplanar strips (CPS), coplanar waveguide (CPW) as well as the new feeding method of slots on dielectric substrates. Moreover, the paper gives an overview of realized 3D printed antennas: printed antenna structures placed in cylindrical-parabolic or corner reflector. All presented structures are realized on a prototype level and have great number of citations. They could be designed for operations in millimeter-wave ranges which is of great significance due to rising development of 5G radio-communication systems and devices.

ABSTRACT. Static power converters and power electronics play significant role in industrial applications, power generation and transmission, home appliance, transportation, etc., etc. Today, power converters are part of our life and everything will be completely different without power converters. How we arrived here where we are now and when static power conversion started? Is it a new discipline or not really? Where we are going in next years and decades? What is future of static power converters? In the first part of the talk we will briefly go through the history of static power conversion starting from very first “power converters” such as Herz’s Oscillator and Tesla’s Transformer, then very first “Power Devices” such as mercury arc rectifiers, megatrons, thyratrons, and then finally real power devices, SCRs, BJTs, MOSFETs and IGBTs. In the second part of the talk, influence of static power converters on our everyday life will be addressed. The latest development results in the field of power converters will be discussed too. The third part of the talk will address future of power converters and some open issues that need urgent solution. We will intensively discuss: New power semiconductor devices such as SiC and GaN, “New” topologies, New material for passive devices (magnetics and capacitors) and System integration including power devices, passives and gate drivers. At the end, converters control aspects, hardware as well as strategies will be briefly addressed.

ABSTRACT. This work proposes an algorithm for sensorless detection of the rotor speed applied to induction machines. The method relies on variable magnetic reluctance, caused by the rotor slots. The analysis shows that the reluctance changes produce slot- related harmonics in the stator current. The frequency of such harmonics is related to the rotor speed. Proposed is the algorithm that tracks the slot harmonics and derives the rotor speed. The paper includes the key details of the on- line DSP implementation that runs in parallel with the remaining drive- control functions. Experimental verification and comparison to the state of the art solutions is performed on a three phase induction motor.

ABSTRACT. Paper presents an overview of the electric distribution test grids, issued by IEEE, CIGRE and others technical institutions, which includes distributed generators. It shows their qualities, quantities, the main features as well as the purpose and the type of grids. The paper also demonstrates the well-known test grids (IEEE13, IEEE34, etc.) and the new test grids which are being developed in recent years to suit changes in distribution systems. These new test grids integrate distribution generation and HVDC technology. An example of application of the IEEE 13 for distribution generation testing is presented, also.

ABSTRACT. The natural turn to renewable energy sources, from fossil fuel based electrical energy generation, has led to significant increase in number of grid connected converters. Following this, the influence of the power electronics elements on power quality characteristics of the distribution networks needed to be investigated. It can be proven that different number of control parameters, ranging from the modulation techniques to the current controller parameters, can influence the power quality, in the first place total harmonic distortion of the injected currents. This paper will present the influence of the DC-link circuit voltage oscillation influence on the grid connected converter power quality characteristics. Experimental results are acquired using advanced laboratory prototype for testing renewable energy based distributed resources, developed at the Faculty of Technical Sciences in Novi Sad.

ABSTRACT. Multi-axis drives increasingly attract the attention of the scientific community because of the benefits they provide in terms of price, compactness and energy-efficiency. In this paper, a method is proposed for reducing the effective current of DC link. The method is based on the optimal adjustment of phase difference of PWM carriers of each drive. It is shown that minimum effective current of DC link is obtained when phases of PWM carriers make arithmetic progression with step 180/n, where n is number of drives. If drive changes over from motor mode to generator mode or vice versa, the minimum current of DC link is maintained by delaying PWM carrier of that drive for 90.

ABSTRACT. simple data logging and measurements for basic academic lab experiments on electrical machines is usually performed using costly measurement equipment already available in laboratory. On the other hand, non-neglectable impact of low cost open source hardware, like Arduino or Raspberry Pi, has created an alternative for systems with low requirements. Increase in computational power of these devices and rapid decrease in sensors price now result in standalone low power electrical motor acquisition system that can be built for a price just over 100 €. This paper presents one of the solutions of that acquisition system that is intended for student’s experimentation with the hardware and software, and adopts to existing sensors of interest in the laboratory to provide the possibility of integration with various sensor and motor types and sizes. Software on the Raspberry Pi 3 is written using Python programming language, and the Arduino Due is programmed using Arduino variant of C programming language. Overall system design and some sample measurement results, along with GUI (graphical user interface) is shown to demonstrate the usefulness of the system.

ABSTRACT. The aim of this paper is to investigate experimentally the effect of different nozzles, water head and discharge on the performance of Pelton turbine system. The effect of five different nozzles with outlet diameters of (3.61,5.19, 8.87, 12, and 14.8) mm has been studied. The tip and hub diameter of the Pelton wheel of (269.89, 221.29)mm has been used with different nozzles. A water pump was used to generated the pressure head and volume flow rate. The results show for every certain nozzle that when decreasing the water head lead to reduction in water discharge and this caused reduction in the torque, brake power, efficiency and the rotational speed. In addition, the results show that when increasing in nozzle diameter lead to increasing the discharge and reduction water head. While in the case of constant head of 15 m, the results show that the nozzle with diameter 8.87mm give the best the performance than the others, where the percentage increase in torque, brake power, efficiency and the rotational speed of the turbine are (62.2%, 66.48%, 60%, and 15.35%) respectively with the second nozzle with diameter of 5.19 mm.

ABSTRACT. In this work are described in detail High Voltage Direct Current systems motifs for their application, the basic structure and types of HVDC systems, as well as the advantages and disadvantages of the High Voltage Direct Current systems in comparison with the traditional High Voltage Alternating Current systems. In the context of operation are detailed described components of the HVDC system, with a particular focus on the thyristor valves, electrically-triggered thyristors, light-triggered thyristors, thyristor inverters, the principle of operation of the 6-pulse and 12-pulse thyristor converter, comparison of their characteristics, and the effects of their use.

ABSTRACT. Recent developments in power transmission and distribution networks introduce a number of renewable power sources, electrically controller loads, smart grids, nano- and micro-grids with ac and dc voltages and also numerous static VAR (volt-ampere reactive) converters and active filters. Most of these subsystems and devices connect a number of transformerless grid-connected power converters. Due to imperfections within their topologies, sensing and control, grid-side converters introduce a number of electromagnetic disturbances into the grid, including the line harmonics and some small, parasitic dc voltages and currents known as the dc bias. This component polarizes distribution power transformers and causes higher current distortion and overheating of the transformers. This paper proposes a method for real-time, open-loop, detection of parasitic dc components in 0.4kV AC grids. The method uses a sensing reactor and DSP (digital signal processor). Method is based on low order even harmonics of the reactor current. The paper explains necessary hardware and software for testing sensing reactor and for real-time measuring.

ABSTRACT. U radu je data veza između broja štapova kaveznog rotora i egzistencije rotorovih žljebnih harmonika u spektru struje statora i parazitskih elektromagnetskih momenata trofaznog, kaveznog, mrežno napajanog asinhronog motora u stacionarnom radnom režimu. Analiza je provedena korišćenjem analitičkih izraza za magnetomotornu silu kaveznog namotaja rotora i za razvijeni elektromagnetski moment uz uračunavanje učešća viših prostornih harmonika magnetomotorne sile sa strane statora i posledično, viših vremenskih harmonika u struji kaveznog namotaja rotora. Pokazano je da egzistencija harmonijskih komponenti struje statora na visokim učestanostima, poznate kao rotorovi žljebni harmonici kao i parazitskih, pulsacionih momenta zavisi od broja štapova rotora. Prezentirana analiza podrazumijeva stator sa brojem žljebova koji rezultira cijelim brojem žljebova po polu i fazi. Rezultate proistekle iz analitičkog razmatranja u potpunosti potkrijepljuju rezultati iz numeričkog modela za različit broj štapova rotora, razičit broj žljebova na statoru i najčešće brojeve pari polova komercijalnih motora.