ABSTRACT. Differential protection is one of the most preferred schemes in power transformer protection. However, the main problem encountered while making use of differential protection is its limited ability to differentiate between inrush currents and fault currents. This not only causes incorrect operation of the differential protection, but also has a negative impact on the power system. In this paper, a differential protection algorithm for power transformers, based on the Second Central Moment, is implemented in an embedded system through Raspberry Pi 3, using the Python language. The algorithm presents considerable advantages over other methods, since it is applicable to any power transformer regardless of its power capacity, voltage level, connection and saturation curve, in addition to the fact that the fault detection time is faster than that of conventional differential protection and it can detect internal faults during transformer energization.

ABSTRACT. Radioastronomy has significantly expanded our understanding of the universe by enabling the observation of phenomena not detectable through optical astronomy. This study addresses the integration of a dipole antenna designed for the center frequency of 20.1 MHz, crucial for capturing radio emissions at the JOVE radio telescope. A data acquisition and signal processing system using Python is presented, detailing the stages of antenna construction, data capture, and analysis through digital processing techniques. This approach enables exploration of complex electromagnetic interactions in Jupiter's environment and enhances our understanding of these astronomical phenomena.

ABSTRACT. Astronomy allows us to study outer space, understand how our solar system was formed, analyze the celestial objects that exist, as well as study the effects that these objects have on our planet. Through optical telescopes we can see and study these objects by observing their radiation in the visible, infrared and ultraviolet spectrum. Through radio telescopes (radio antennas) we can study emissions outside the visible spectrum, in a radio frequency section of the electromagnetic spectrum, providing valuable information. It is known that some objects emit radiation at specific frequencies. An essential processing of a radio telescope signal is to obtain the received frequency spectrum to determine whether the frequency of interest was received. This processing can be done in software that hardly works in real time, or in hardware to try to achieve real-time processing. This work shows the software and hardware processing of signals received by the antenna and receiver of the Radio Jove project, aimed at studying the signals generated by Jupiter and its moon Io.

ABSTRACT. This paper presents the design and construction of an electric drive traction platform based on an induction machine whose capacity is around 26 kW. A visual interface is used to display the critical parameters of the system such as current, voltage, RPM, etc. In addition, a cooling system featured by a radiator, water pump and a liquid cold plate is reported as well.

ABSTRACT. This work highlights the frequency stability of a power system, focusing deeply on the oscillation equation of a rotational system found in conventional electric power generator rotors. Additionally, it analyzes the control of an inverter in a photovoltaic (PV) system and an energy storage scheme (a constant DC source emulating a battery). The focus is on ensuring control to simplify the system and conduct evaluations of virtual inertia emulation based on energy storage.

ABSTRACT. This paper presents the controller tuning required in a wind generation system using a Back-to-Back converter and a 5.5 kW doubly fed induction generator connected to a 220 V three-phase line-to-line grid. The control system contemplates the application of the classical control technique known as vector control, space vector modulation and the application of the antiwindup tracking back strategy. At the end, the simulation results are interpreted using MATLAB, obtaining through the references of the control loops and according to the generator operating conditions, the generation of 3 kW in subsynchronous mode and the generation of 5.6 kW in hypersynchronous mode, both with a voltage of 250 V in the DC bus

ABSTRACT. This paper introduces a method to minimize fluctuations in wind power generation. The aim is to decrease power variation in a wind energy conversion system using a control scheme based on ramp events. This, along with a smoothing control system that utilizes a battery energy storage system (BESS), helps to reduce power fluctuations. To test the method, a wind farm model consisting of 26 wind turbines, each with a capacity of 2 MW, situated along a coastal profile (52 MW of installed power), is used as a case study. The results illustrate the proposed method's effectiveness compared to the original system performance, demonstrating a significant improvement in the stability of power generation. The enhanced economic feasibility is further established by minimizing the costs associated with critical components, such as the BESS.

ABSTRACT. Voltage stability is a key factor in the reliable operation of distribution grid in the context of dynamic loads and changing power demand. Conventional voltage control methods depend on reactive power control and static adjustments, which make them ineffective in managing the rapid fluctuations caused by the unpredictable and variable nature of new load patterns such as electric vehicles and smart appliances. This paper focuses on a new approach to ensure voltage stability in low-voltage (LV) distribution grids using machine learning based forecasting and mixed integer linear programming (MILP) based optimization models for household load forecasting and optimization. The long short-term memory model is used on historical time series household load data, which accurately forecasts the day-ahead (short-term) household load. Then the MILP based optimization model for the home energy management system uses the forecasted load data to optimize load and ensure voltage stability in the LV distribution grid. In the base case, there are 238 instances out of 960, where voltage violations occur, but in the optimized case, the voltage violations were reduced to 36 instances in one day for 10 houses, leading to improved voltage stability in the LV distribution grid.

ABSTRACT. Detecting faults in rotating electrical machinery is necessary due to current industrial demands. This is because operations of these machines with unattended faults can lead to high electricity consumption, operational inefficiency, and unexpected breakdowns in the worst cases. One of the most common faults in induction motors tends to be stator faults, specifically the inter-turn short circuit (ITSC), which typically do not exhibit apparent symptoms in the early stages. However, these faults can be detected without physical motor intervention, as spectral signatures of faults can be identified through phase current measurements. In this way, only the sensors' installation in the motor phase is needed to diagnose the motor health correctly. The challenge relies on selecting and designing a correct methodology for these purposes. So, this study introduces a technique based on constructing images from current signals, where the fundamental frequency has been suppressed to highlight fault spectral signatures. These generated images are subsequently fed into a convolutional neural network (CNN), which develops a classification stage according to the given features of the fault conditions in previous training stages, allowing the identification of the ITSC fault. Finally, tests were conducted under three mechanical loading conditions and four short-circuit damage conditions in the stator, achieving an accuracy of 98.62 %.

ABSTRACT. Visual Impairment (VI) and blindness have serious consequences throughout the lives of people who suffer from it. Around the world, nearly 2.2 million people have a VI. The main causes of VI worldwide are refractive errors and cataracts, which generate a huge global financial burden. Vision loss affects people of all ages. This research focuses on the technological development of an electronic prototype, based on the concept of Artificial Intelligence of Things (AIoT), which supports the literacy of children who use the Braille system to read documents and the inclusion of people who already master it.

The device has a camera that is connected to a small, powerful computer, which allows it to run Deep Learning (DL) based Artificial Intelligence (AI) algorithms. This enables automatic object identification a, converting object names and characters to Braille language. The objective is to help the user recognize their browsing environment through the designed device.

ABSTRACT. An issue that is currently important worldwide is climate change and its negative consequences. In this sense, the use of renewable energies and associated technologies has been one of the strategies developed and promoted to minimize these changes experienced on the planet. Among these technologies are wind turbines, which transform wind energy into electrical energy, minimizing the carbon footprint. To extend the life span of wind turbines and maintain their efficiency, damage detection becomes a fundamental issue. Taking this into account, this work presents a simple and efficient machine learning scheme for the detection of cracks in one of the wind turbine blades. This scheme uses vibration signals, statistical indicators and neural networks, obtaining 99.75% efficiency in detection and thus demonstrating the great performance of artificial intelligence in this area

ABSTRACT. In recent years, analysis of power quality disturbances has taken on great relevance due to the greater use of power electronic equipment and the search for alternative sustainable energy sources whose use generates fluctuations in the electrical grid. In real-time applications, studying hardware platforms that allow disturbance detection and classification algorithms to run is relevant. The present work shows the implementation of an algorithm based on DWT and SVM in different single-board computers to identify and classify power quality disturbances. From the results, a success rate of 99.80% is obtained in the classification of seven power quality events and an execution time of less than 2 ms for Raspberry Pi4 and BeagleBone AI-64.

ABSTRACT. This paper presents the simulation of the loading and unloading process using the Rint equivalent model of a BIL using Simulink in order to analyze the implemented Pi control system. The importance of BILs in the transition to renewable energies is highlighted and a detailed approach on the simulation of the BMS for the charging and discharging control is presented. Equivalent circuit models, such as the Rint, RC and Thevenin model, are described and the evolution of battery management systems with the integration of artificial intelligence is highlighted. Simulation results show how the effective control of the maximum operating limit of the SOC restricts the voltage at the battery, as well as the control of the charging current given by the implemented BIL block.

ABSTRACT. In this paper, a simple method for estimating the parameters of small Permanent-Magnet DC motor from few physical measurements for Control Systems teaching purposes is presented. A total of 45 different motors were characterized, transfer functions were calculated and simulated on Simulink software and compared with measured data. Over 80% of the measurements matched a proposed criterion of 10% of error between measured and simulated data. Continuous and discrete positon P-controllers were also simulated and physically implemented.

ABSTRACT. This contribution addresses the application of a time-domain method for the assessment of harmonic distortion emissions of electronic equipment that operates at low-voltage installations and emits signals in the frequency range from 2 to 150 kHz. Their classification is standardized as a highfrequency disturbances referred to as supraharmonics. This type of phenomenon requires theoretical and/or digital analysis applications, which imply a better understanding for its analysis. The application of discrete Norton equivalent models, as well as their efficient solution through the nodal incidence matrix using the LU decomposition process, allows the understanding of this type of power quality adverse phenomena in the timedomain framework. This approach offers the applied theoretical and experimental models, as well as the comparative analysis using the digital tool Simulink®.

ABSTRACT. The Capacitive Power Transfer (CPT) is a technology that has been developed in recent years as an alternative to the Inductive Power Transfer (IPT). CPT systems have been sought to be implemented in battery charging applications where a condition to compete with the IPT systems is the need to increase the power transfer in the CPT systems without significant losses. This paper proposes a method to maintain the maximum power transfer condition in a CPT system against changes in the coupling capacitor value, which can be achieved through the output load value. An experimental prototype has been developed to validate the analysis and the results are compared with simulation results.

ABSTRACT. Technological development, where the control or manipulation of devices by command voice, plays a relevant role in our daily lives. Its application is pertinent to create intelligent voice-controlled interfaces that can be integrated into several devices such as smartphones, laptops, virtual assistants, and other command-operated systems, however, the scarcity of public models adapted to the Spanish language is limited. In this context, the proposed work explores the feasibility of creating a model for keyword spotting or command-voice detection in the Spanish language. The proposed model is based on the log-Mel Spectrograms in combination with a Dual-Layer LSTM Classifier. The experimental results show an accuracy of 90.82% in classifying 20 spoken Spanish words from the Multilingual Spoken Word Corpus (MSWC), by other way, the proposed model provides higher accuracy compared to CNN models and maintains similar results to the existing LSTM architectures

ABSTRACT. It seemed that the 4th Industrial Revolution was creating a long-term discourse as a main actor in the industrial sector, however AI (Artificial Intelligence) has left it in the background. AI is being a disruptive revolution in all sectors of activity, creating very high economic values. It is a technology highly dependent on “data”, which has currently driven other technological waves to provide it, such as the Internet of Things (IoT), 5G technology, also the Cloud and high-performance microprocessors to process it. In this work we assume that the information processing model is Edge Computing, that is, the processing is carried out in the same place where the data is produced. To do this, we designed the Lagarto architecture, an embedded microprocessor, based on the ISA RISC-V, with the capacity to execute Artificial Intelligence models. Our results show that in industrial processes where decisions have to be made in situ, using artificial intelligence models, microprocessors such as Lagarto provide the performance required for decision making, offering an IPC of 0.5 (one instructions each two clock cycle).

ABSTRACT. The state estimation of power systems is a fundamental process in the operation and control of electrical power networks. However, state estimation encounters various types of anomalies. These bad data can be generated due to faulty sensors, aging of system equipment, communication errors, and random environmental effects. Modifications in the topology of the electrical network produce sudden changes at various points in the network, which can also be considered anomalies. Existing methods to discriminate between sudden changes in the electrical network topology and gross errors cannot accurately identify the aforementioned types of anomalies. The presence of these outliers will reduce data quality, affect the analysis and control of power system monitoring, and pose a threat to the safe operation of the power system. Therefore, timely identification and treatment of outliers are essential to ensure a stable and reliable power state estimation. This work proposes three techniques for identifying bad data. The first technique is the decomposition of data into trend, seasonality, and residual using LOWESS; the second technique is the 1D-CNN Auto-Encoder; and the third technique is the LSTM Auto-Encoder. To verify the performance and compare the methods, the IEEE 14-bus network is taken as the test system, and experiments simulating different test scenarios are conducted. Numerical tests verify the efficiency of the methods in the presence of outliers, demonstrating that they can accurately identify outliers.

ABSTRACT. Breast cancer is one of the leading causes of mortality among women worldwide. In 2020, breast cancer was the primary cause of cancer-related deaths among women globally, with 685,000 fatalities. Today, healthcare professionals continue to seek and implement innovative techniques to aid in the early detection of this disease. One such technique being explored as a complementary method to traditional studies like mammography is thermography. This technique detects the infrared radiation emitted by the body. As a tumor begins to develop, blood vessels and blood flow increase, which also causes an increase in the surface temperature of the skin in that area. Artificial intelligence techniques like machine learning and deep learning have been implemented to analyze breast thermographies. This study uses machine learning models to classify breast thermographic images into two categories: without anomalies and with anomalies. Results showed that it is possible to reach an accuracy of 83.46% in the automatic classification of breast thermographies. Future work will aim to improve this result by implementing new features that could be relevant for classification.

ABSTRACT. Currently, it has been observed that patients with specific needs require limb immobilization, and the use of conventional methods does not provide optimal comfort for these individuals. These methods are often heavy and made of materials that cause skin damage and do not allow for adequate hygiene, among other inconveniences. With technological advancements, it is now possible to design custom-made orthoses tailored to the individual's physique and size, offering greater comfort for patients using this method. Additionally, due to its advantages, the material used for 3D printing significantly helps reduce the discomfort caused by conventional methods. For this, a 3D scan of the affected body part is necessary, providing a three-dimensional figure to be used when modeling the tool, ensuring it is as anatomically accurate as possible for the patient. Once the model is obtained, it is sent to the 3D printer; for the prototype of this project, PLA filament is used. Furthermore, the application of Industry 4.0 is sought so that individuals with specific needs using this 3D orthosis can be monitored with the Internet of Things (IoT) for biomedical data such as body temperature, blood oxygenation, and heart rate. In this way, utilizing Health 4.0, caregivers can monitor biomedical data, allowing them to respond quickly to any variations in signs indicating anxiety or pain crises.

ABSTRACT. Howland current sources are commonly used in the development of electrical impedance tomography (EIT) systems because they ensure control over the amplitude of the current supplied to the patient during the injection of the auxiliary signal into the tissue. However, the capability of Howland current sources is restricted up to 100 kHz, which is adequate for most biomedical EIT applications. Nevertheless, the particular case of breast cancer can greatly benefit from the use of auxiliary currents at high frequencies, exceeding the megahertz range, as the bioelectrical differences between healthy and pathological tissues present greater contrast, which may facilitate the detection of potential cancerous tumors. Various studies suggest paying special attention to the selection of resistor tolerance as well as the characteristics of the operational amplifiers used in their construction. This study presents the results obtained from mathematical modeling, simulation, and physical implementation of the effect that resistor tolerance has on the output impedance of Howland current sources, which ultimately affects their performance at high frequencies.

ABSTRACT. The Model Predictive Control (MPC) has become a widely used technique in power converters for its simplicity and intuitive algorithm. However, for topologies with a high number of operating modes, the computational cost of the controller implementation increases significantly. To face this disadvantage the controllers based on Artificial Intelligence (AI) are a suitable solution. In this paper an Artificial Neural Network Controller (ANNC) based on a MPC is proposed for a three-phase inverter that feeds a three-phase motor. Simulation results validate the proposal for reference, current demand and input voltage variations.

ABSTRACT. A boost converter increases DC voltage by storing input voltage in an inductor and releasing it through a controlled switch, typically a transistor, at high frequency. Effective current control optimizes energy efficiency, reduces power losses, and enhances voltage conversion. Lyapunov-based control in boost converters ensures stability, efficiency, and minimized voltage fluctuations, simplifying design and enhancing system resilience. This research advocates for implementing Lyapunov control in DC-DC boost converters to regulate current dynamically, ensuring a stable power supply crucial for diverse applications in advanced power conversion technologies. The proposed Lyapunov-based control, derived from a Lyapunov function, inherently ensures stability without the need for additional stability proofs. This approach leverages mathematical guarantees from Lyapunov theory, simplifying the design process by directly providing stability assurance for the control system.

ABSTRACT. This paper presents the design of an ideal MIMO antenna for applications in the Sub-6 GHz band. The antenna, with dimensions of 29×38 millimeters, is printed on an FR4 dielectric substrate with a permittivity of 4.4 and a thickness of 1.6 mm. The MIMO antenna’s patch element has rectangular cuts in the lower corners, a reduced ground plane, and a rectangular cut in the central and upper part of the ground plane. These cuts, both in the patches and the ground plane, were optimized to broaden the operating bandwidth of the patch. The 2×1 MIMO antenna is composed of two optimized patch elements. The mutual coupling response between the MIMO antenna elements and the return loss are improved through the implementation of a defective ground structure (DGS), which consists of adding a perforated dipole to the ground plane. The MIMO antenna offers an operating bandwidth of 7.64 GHz, from 1.96 GHz to 9.60 GHz, with a maximum gain of 4.85 dBi at 3.5 GHz in the theta direction equal to 133.75◦.

ABSTRACT. Planar inductors have become fundamental components in the design of integrated circuits and high-frequency electronic systems. They are widely used in different industrial applications, acting as filters, protecting electronic circuits from oscillations in the electrical network, or accumulating energy in the form of a magnetic field. This article presents a proposed calculation, design, and manufacture of planar inductors, addressing their manufacturing and design, focusing on manufacturing with 3D printers, the challenges associated with their implementation and manufacturing will be discussed in this paper

ABSTRACT. The back to back (BTB) converter of a wind energy conversion system (WGS) is always affected by perturbations. Such perturbations induce harmonic currents in the generator and/or the power filter terminals, affecting the wind to electrical energy conversion process. Thus, in this paper, it is proposed a vector super-twisting (ST) controller for the BTB converter designed to reject a particular and harmful perturbation known as dead-time in the WGS generator. The developed ST control is compared with a proportional-integral (PI) control through simulations. By replacing the PI controller for the proposed ST algorithm, the total rated current distortion (TRD) of the generator is reduced from 9.9 % to 0.75 %. The results demonstrate that the proposed ST algorithm is an option that allows reducing the harmonic current circulating in the generator of a WGS with a BTB converter

ABSTRACT. The trend in the development of microgrids consists of integrating renewable energy resources, controllable loads and energy storage systems, in a more economical and reliable way. Electric energy storage systems consist of batteries that are essential to compensate for the uncertainties that may arise in the operation of the microgrid. This paper presents a proposal with a mathematical programming approach that manages multiple sources of distributed generation that present the microgrid over different periods of time, through the economic dispatch approach to minimize operating costs. The mathematical model considers the following variables: a) conventional generators, b) renewable energy sources, c) bonus program for electricity demand response, d) application of environmental contingency; and d) a battery energy storage system, which intervenes in the operation of the system, considering the strategy of attending to demand and seeking to reduce costs. The proposed algorithm contemplates a program of benefits for customers by reducing their demand and considers the costs of regeneration of pollutants produced by conventional energy sources. In addition, the analysis includes the benefit of reducing the use of conventional sources of electrical energy, complying with a set of restrictions of the proposal. To establish the validity of the approach, a case study is implemented that includes a combination of renewable energy sources backed by the conventional electricity main grid and the use of batteries to satisfy electricity demand. The proposal is implemented in Lingo 17.0 and the simulation results show the feasibility of the proposal through the optimal configuration of the microgrid.

ABSTRACT. The numerical simulation of vertical heat exchangers (GHEX) is limited by its geometry due to its length vs diameter relationship, which brings with it divergences in its solution or discretization problems and the use of dimensionless numbers that allow the development of mathematical models. Therefore, this work develops a model in which the global heat transfer coefficient is estimated, taking into account the surrounding regions such as the slurry and soil, which are rarely taken into consideration to reduce the complexity of mathematical modeling. To validate the proposal, an experimental model of a one-pass U-type exchanger was developed to estimate thermal conductivity values of the strata and variations in speed and temperature of the working fluid. The mathematical model was carried out using an algebraic system of equations that allow estimating the global heat transfer coefficient. These results will allow improving the sizing of heat exchangers, developing new mathematical models, improving overall efficiency and reducing the implementation costs of GSHP systems.The numerical simulation of vertical heat exchangers (GHEX) is limited by its geometry due to its length vs diameter relationship, which brings with it divergences in its solution or discretization problems and the use of dimensionless numbers that allow the development of mathematical models. Therefore, this work develops a model in which the global heat transfer coefficient is estimated, taking into account the surrounding regions such as the slurry and soil, which are rarely taken into consideration to reduce the complexity of mathematical modeling. To validate the proposal, an experimental model of a one-pass U-type exchanger was developed to estimate thermal conductivity values of the strata and variations in speed and temperature of the working fluid. The mathematical model was carried out using an algebraic system of equations that allow estimating the global heat transfer coefficient. These results will allow improving the sizing of heat exchangers, developing new mathematical models, improving overall efficiency and reducing the implementation costs of GSHP systems.

ABSTRACT. This paper describes a novel statistical methodology developed to compute the ramp rate of a photovoltaic (PV) plant, which filters out the most typical and expected ramp rates of a PV plant and focuses on the more random and severe ramp rates that can have an impact in system operation, hereinafter referred to as ramp rate of interest. This methodology is applied to existing data made from measurements of the power output from the Santa Rosalía PV Plant, which is located in the Mexican state of Baja California Sur in the Sistema Interconectado Mulegé (SIM). The values obtained are considered sufficiently adequate for use in dynamic studies, analysis of operating reserves, automatic generation control settings and other processes that are affected by sudden and random changes in generation output from a PV plant.

ABSTRACT. The following paper shows an on-field-research that tries to illustrate the user how L3 rapid chargers operate in a few selection of electric vehicles (EVs). In may researches, EVs are modeled as fixed loads. However, experimental observations have demonstrated that the delivered energy is a function of the state-of-charge (SOC), and not all vehicles are charged in the same way. This issue becomes relevant when the owners or the route-planners must consider time and costs to charge, especially when the charging current drops its value, or in those cases in which EVs are charged per minute, as the case of Costa Rica. Most manufacturers only limit the charging information for the users as an approximate time to reach a certain SOC, at a certain power. Nevertheless, additional information regarding the charging process is not given, and it is not easy to find in the available literature. In this regards, this research aims to illustrate a field research that builds this information from scratch. This particular effort consists of visiting specific L3 charging stations and collect the data of different EVs when they are being charged. One of the main considerations is that not all L3 stations show all the relevant information to the users, and that the available data is subject to the EVs that visit the charging point. In this case, data for three different EVs was gathered and processed. The results show differences on the way these EVs are charged, making it clear that more data needs to be collected, at least until the manufacturers make public this information.

ABSTRACT. The application of vibration-reducing techniques in industrial machinery is pervasive, as they influence the precision and durability of the equipment. Accordingly, this study presents a methodology for the design of a trapezoidal profile trajectory in convolution acceleration with a finite impulse response filter with a Gaussian window for vibration reduction by means of smoothing the jerking profile. The methodology is validated using a test bench, the implementation in a programming card, and an accelerometer, resulting in observations of a reduction in the vibrations caused by the jerk.

ABSTRACT. Condition monitoring of induction motors (IMs) are essential in the modern industry. Researchers globally have developed various methods to monitor and detect different types of damage in IMs, such as damaged bearings, misalignment, broken rotor bars and electrical faults. Regarding to electrical faults, detecting stator windings short-circuits is particularly challenging, since it can help prevent catastrophic damage. This work investigates the potential of Hjorth parameters as indicators of stator windings short-circuits by analyzing current signals from IMs under different damage levels. The results show that Hjorth parameters are sensitive to these conditions, enabling the development of pattern recognition schemes for automatic classification. The k-means clustering method is proposed for this task due to its simplicity. The results confirm that the suggested techniques are reliable for monitoring IM conditions, achieving an accuracy of 100%

ABSTRACT. Induction Motors are critical components in industrial applications and are considered the most influential mover due to their simple and reliable construction. Induction motor condition monitoring involves continuously assessing the system's health. Recently, image-based techniques combined with artificial intelligence have garnered significant attention for their pattern recognition and data classification ability. This study introduces a Small CNN architecture designed to detect bearing faults using Infrared Thermal Images. The architecture is designed to effectively classify three types of faults with an accuracy of $1$. It performs well even with noisy thermal images with an accuracy of $0.80$ with an MSE of $20$, maintaining its robustness under varying conditions.

ABSTRACT. Abstract— Digital fringe projection profilometry (DFPP) is a well-established technique for three-dimensional surface measurement, traditionally employing vertical or horizontal fringe patterns. This study explores the use of circular fringe patterns in DFPP, aiming to leverage their unique geometric properties to potentially enhance measurement accuracy and robustness. Circular fringes, when projected onto an object, present several potential benefits over conventional linear fringes. Additionally, circular fringes can eliminate the need for a carrier to achieve satisfactory phase development, which may streamline the process and improve efficiency. The projection of circular fringes may inherently smooth the image and reduce noise, potentially reducing the necessity for low-pass filtering and preserving the integrity of high-frequency surface details. The iterative process required to achieve accurate surface reconstruction might be significantly reduced, potentially resulting in faster computation times and more efficient data processing. The inherent symmetry of circular patterns could also simplify phase unwrapping algorithms, as the phase distribution is continuous and naturally wraps around the center point, potentially reducing phase discontinuities and errors. This study presents a comparative analysis of circular versus linear stripe patterns in DFPP, investigating the potential advantages of circular fringes in terms of measurement accuracy, robustness to occlusions, and simplicity in phase unwrapping. Experimental results on various surface profiles suggest that circular fringe projection could be a promising alternative for 3D surface measurement in complex and dynamic environments.

ABSTRACT. Electrical power transformers play an important role in the electrical power system, since they are responsible for raising voltage levels in order to transmit electrical energy from the generating plants and subsequently reducing them for the load centers. When these are put into operation, the transient phenomenon called Inrush current appears. The main characteristic of this phenomenon is that the energizing current has a magnitude several times greater than the nominal current of the transformer, which could give rise to false trips of the transformers switches. The main characteristics is its waveform with asymmetric that can be positive half cycles or negative half cycles, which is damped over time and its harmonic content is mainly composed of the second harmonic component. In this sense, this paper shows the energizing of real Transformer of 60 MVA with no success and success operation.

ABSTRACT. In this paper, real time validation of active power injection by solar + wind + batteries microgrids in electrical distribution networks using software in the loop (SIL) is presented. The wind fluctuations and irradiance conditions are analyzed from the peninsular region through the Mexican Energy Atlas project. The variability of environmental conditions that impairs a continuous power production affects the voltage at the PCC. Then, this research proposes a control law for the microgrid output power interconnection into the distributed electrical networks, without voltage fluctuations and satisfying the users electrical demand. The proposed methodology functionality is validated through the simulations in MATLAB-Simulink® and tests based on the SIL methodology with the real-time simulator Opal-RT Technologies®. The results are shown an effective methodology, where the microgrid is designed to support an active power generation of up to 75 kW.

ABSTRACT. This paper introduces an energy management system for a small-scale hybrid micro-grid integrating wind, solar, and battery storage. The system encompasses wind and solar conversion systems, alongside a battery storage system, all developed with power electronics converters, control algorithms, and controllers to assess the micro-grid's performance. To maintain voltage balance, a control algorithm modulates the operation of the Dual Active Bridge, charging the battery when renewable energy production surpasses load demands and discharging it when production is insufficient. This regulation ensures a stable DC bus voltage, which in turn maintains a consistent output from the Voltage Source Converter for AC loads. The proposed small-scale renewable energy micro-grid serves as a test bench for research and algorithm testing in smart grid applications.

ABSTRACT. This paper presents a Phyton-based user-friendly graphical user interface (GUI) toolbox to synthetically generate power quality events dataset. This GUI is able to provide and export a versatile collection of tools designed specifically to create robust and realistic datasets for their utilization in the evaluation and testing of machine learning and deep learning algorithms for PQ events in power systems. A straightforward implementation is adopted to generate datasets with up to 29 different PQ events with automated label options. Numerical and graphical results demonstrate the effectiveness of the GUI.

ABSTRACT. This article proposes a tool to assist healthcare professionals in the automatic segmentation of the main heart sounds S1 and S2 present in phonocardiogram signals. To achieve this, two trained deep learning models process time-frequency images of the sounds as input. These images are generated by applying a wavelet transform to the audio signal containing the two sounds, which are semi-automatically labeled and used to train, validate, and test the deep learning models. Analysis of a subset of real-world data reveals that the tested models achieved segmentation accuracies of 99.55%, and 99.54%, using the established segmentation in the database as a baseline. This tool allows to individually identify and analyze not only the primary heart sounds but also the systolic and diastolic zones.

ABSTRACT. This paper presents an automatic monitoring and control system designed for hydroponic operations, aimed at reducing maintenance costs, increasing cilantro production, and lowering energy consumption. The system incorporates a LabVIEW™ user interface for online monitoring and control of pump on/off cycles, solar light levels, and artificial light intensity. Additionally, a PID fyzzy control algorithm is proposed to regulate light intensity, ensuring optimal conditions for crop growth. Experimental results demonstrate the system’s capability to accurately collect, process, and display data, as well as issue alarm signals, thereby enhancing efficiency and sustainability in hydroponic farming.

ABSTRACT. The ability of Convolutional Neural Networks (CNN) to identify patterns on images make them suitable for many computer vision applications. However, the convolution operations involved in their use produce high computational costs, which hinders CNN application in low-end devices or high-resolution images. To face this issue, architectures known as Fourier Networks, have been proposed, taking advantage of the convolution theorem to substitute the convolution operation with the Hadamard product, working with the frequency representation of the involved data. Unfortunately, this solution increases the number of parameters in the kernels. Here, we propose a novel architecture, called Butterworth CNN (BW-CNN), aimed at reducing the memory consumption of Fourier Networks, by replacing the commonly used definition of kernels in frequency representation by a kernel generation function, based on Butterworth filters. Our model generates kernels for the frequency representation of images parameterized by four values, independently of the size of the input image. Moreover, a novel pooling layer for frequency representation is also included. Our experiments classifying high-resolution images showed that BW-CNN used less parameters than state-of-the-art Fourier CNNs, as well as CNNs working with spatial representations. Moreover, BW-CNN performed less floating point operations than other CNNs during inference, while achieving similar accuracy.

ABSTRACT. The IoT technologies have promoted important advanced to achieve the sustainability in crop production. Precision Agriculture have been a technological paradigm which has been adopted for the innovation agricultural techniques in order to improve production and reduce environmental pollution. In this paper, we propose an IoT device which has been used to monitor the crops using a moisture and nutrient soil sensor. We use ESP32 micro-controller for signal processing of sensor. ESP32 micro-controller transmits the data signal of the sensor to Raspberry Pi Web server. Raspberry Pi 4 hosts the web application with a transactional database system and dynamic content using Flask framework. The Back-end manages and process the data which are generated by IoT device. The Back-end infrastructure includes the data storage of the nutrients and moisture of soil using MariaDB open source database. The Front-end provides the user interface which allows to the users interact, to manage the IoT device and to analyses the soil nutrients of the soil using a dashboard in real time. An analysis of the capabilities of the IoT system are presented for four types of plants.

ABSTRACT. This article presents the detailed design of an Uninterruptible Power Supply (UPS) system specifically developed to ensure the continuity of the power supply to Mechanical Ventilators in critical medical care environments. The designed UPS incorporates advanced features such as component redundancy, electromagnetic interference suppression, remote monitoring with early alerts, and strict compliance with medical standards, including IEC 60601-1, IEC 60601-1-2 and ISO 80601-2-12. The development process encompasses initial research and requirements analysis, through to integration, manufacturing, exhaustive validation testing, and obtaining the necessary regulatory certifications for commercialization. This approach ensures that the UPS not only meets the expected safety and performance standards but also guarantees critical reliability in situations where continuous power supply is vital for human life.

ABSTRACT. Breast cancer is one of the leading causes of death among women worldwide, and early detection is crucial to ensure proper treatment and, consequently, patient survival. Electrical Impedance Mammography (EIM) is a novel technology that measures electrical conductivity in the breast and associates it with healthy and malignant tissues. This article focuses on EIM image processing, which is essential to increase its effectiveness in detecting abnormalities likely related to breast cancer. EIM information allows observing the distribution of conductivity in various crosssectional images of the breast, facilitating the detection of lesions by identifying areas with abnormal conductivity. However, it is necessary to find a way to combine the information obtained from EIM images to identify relevant differences between the layers. This work proposes a comprehensive representation of the images through an in-depth quantitative estimation of the EIM layer variation that improves their visibility and analysis, increasing the precision in differentiating benign and malignant tissues. The results show that preprocessing EIM images highlights key features, significantly improving the precision and consistency in lesion classification, with logistic regression standing out as the most effective classifier, with an accuracy of 87.98 ± 0.079.

ABSTRACT. Diabetes mellitus currently affects more than 420 million people around the world according to the WHO. This disease can cause another condition known as diabetic foot syndrome. The ulcers caused by this syndrome can lead to amputations, because in many cases the biased diagnoses caused mainly by lack of information on the precise state of the wounds, lead specialist personnel in the matter to make decisions that can change the life of a patient in a negative way. The purpose of this article is to present the exploration and results through the use of available technology that provides a closer view of reality based on data, and that serves as support to specialists in this syndrome, to influence a diagnosis that could be more favorable in the care of patients with this disease. The use of devices with infrared light for both measurement and reconstruction of 3D models turns out to be a very convenient support as a reference for other techniques such as image analysis with controlled light. The article presents the use of this technology through a ToF sensor, taking measurements of wounds caused in laboratory mice in which diabetes was induced. Despite the difficulties inherent to an in vivo experiment (size of the wound, movement of the specimens) it was possible, in many cases, to obtain data that proves that the use of infrared light with devices can provide data that combined with other techniques, they would allow for a precise measurement of the volume of the wound and therefore, for specialists to have elements to give a more accurate diagnosis