ABSTRACT. This lecture presents the fundamental role of capital goods manufacturing and advanced robotics in driving industrial development and economic growth for Latin American countries. It examines the potential of Industry 4.0 technologies, including collaborative robots (cobots), artificial intelligence, autonomous mobile robotics, and advanced vision systems, in transforming the region's manufacturing sector. A comprehensive analysis of innovative applications such as robotic additive manufacturing and predictive maintenance, and their integration into the Industry 4.0 paradigm, is conducted. Specific challenges and opportunities for Latin American manufacturing industries are identified by studying local cases. The impact on workforce dynamics, ethical considerations, and prospects of industrial robotics are also addressed. The findings reveal significant opportunities for economic diversification and increased competitiveness by adopting these advanced technologies. The conference concludes by proposing recommendations to foster the adoption of robotic technologies, emphasizing the importance of collaboration among workers, industry, academia, and government to create a sustainable industrial growth model. This presentation contributes to the importance of technological innovation and industrial policy in developing economies.

ABSTRACT. Yarrowia lipolytica has emerged as an attractive industrial platform due to its ethanol-free metabolism and efficient capacity to synthesize complex proteins. To achieve reproducible high biomass yields, model-based optimization and process control of yeast cultures are required. However, robust mathematical models and effective and scalable process control systems for Y. lipolytica cultures are scarce. In this paper, a dynamic model was developed to design and evaluate a calorimetric control system for the growth of Y. lipolytica, using only standard and low-cost instrumentation. The control system consisted of two loops. The reactor temperature was controlled with a split-range scheme that manipulated two Peltier modules. In addition, the specific biomass growth rate was controlled by manipulating the inlet substrate flow rate using exponential feeding. In this loop, the specific growth rate was estimated by an on-line energy balance. A PID controller with anti-reset windup tuned to a first-order transfer function for temperature control achieved a smooth, oscillation-free dynamic response. The specific growth rate was controlled with a PID, which responded quickly, was effective in attenuating disturbances, and showed a small offset and overshoot. With this control system we were able to achieve a biomass concentration of 114 g/L in 53 hours of fermentation.

ABSTRACT. This paper presents the design of the Practical Nonlinear Model Predictive Control (PNMPC) algorithm for liquid level control. In this research, the developed nonlinear control algorithm (PNMPC) has been implemented in an experimental laboratory module. The proposed PNMPC was programmed in a MATLAB script that sends and receives real-time data from the plant. Additionally, a PID controller with integral windup and Generalized Predictive Control (GPC) were designed to compare performance and setpoint tracking with real data. It can be observed that the PNMPC exhibits better performance, and its implementation is simpler.

ABSTRACT. Operation and control of beer fermentation is complex due to its inherently dynamic nature and the reactor scale employed for its production (> 1000 hL). Models capable of incorporating material and energy balances accounting for the volume heterogeneity are scarce. The latter limits the design and evaluation of control strategies for the temperature of the process. To address this limitation, we developed a tailored compartmentalized phenomenological model of beer fermentation describing the kinetics of biomass, limiting substrate, ethanol, and reactor temperature at an industrial scale (> 3000 hL). Using this model, we simulated two control strategies based on conventional PID controllers and an MPC framework for tracking the reference temperature for beer production. Our results showed important differences in the temperature of the compartments, supporting the incorporation of spatial effects and the implementation of advanced control strategies. In this regard, the proposed MPC strategy achieved superior results than the PID-based control system for tracking the reference temperature in the reactor in all compartments. In the near future, we expect to validate the model and control system in an industrial setting.

ABSTRACT. The development of neural network based models considering physical constraints offers some advantages over other models. In this work, an extended Kalman filter is applied to the Furuta pendulum using a neural network based model of this kind. Firstly, a neural state-space model of the Furuta pendulum is proposed, and then the extended Kalman filter is developed. A numerical example, which includes a Monte-Carlo simulation, is used to assess the proposed extended Kalman filter. As data generating system, a phenomenological model of the Furuta pendulum is used. The example includes a comparison between the proposed filter, and the linear Kalman filter

ABSTRACT. This paper compares the performance of traditional PID, fractional-order PID (FO-PID), and Tilt-Integral-Derivative (TID) controllers in a chemical process with variable delay. The results show that the FOPID controller generally performs best in minimizing cumulative error, maintaining low error values consistently, and handling disturbances and noise.

ABSTRACT. In this work, the estimation of linear parameter varying (LPV) models in Output-Error form is addressed. In particular, it is assumed that the scheduling variable is noisy, a condition often encountered in practice but not frequently considered in the literature. An instrumental variable based system identification method, recently proposed for continuoustime LPV models, is adapted to the discrete-time case. To deal with the noisy scheduling variable, it is assumed that its noisefree version is smooth, enabling us to estimate it using the local regression method. The proposed LPV model estimation approach is compared with other available methods, first through a discussion and afterwards through a numerical example

ABSTRACT. This article presents the systematic evolution of 5G and 6G wireless networks and future projections, presenting the versions of the different 3GPP reals with a deployment plan until 2028. It also presents the challenges and vision of future generations and typical applications of future 6G wireless networks, along with the above, the characteristics of the applications and international standardization bodies are indicated. Another point that was analyzed for these networks was the characteristic problems of radio wave propagation in the different deployment scenarios, whether indoor or outdoor. The problems of link blocking and service coverage for users have been analyzed in the field, so the ITU-R P.1410 recommendation is presented, which defines situations with propagation and predictive methods for networks operating between 3 and 60 GHz. Finally, the benefits and advantages of blockchain in 5G and 6G wireless networks are presented, as services dedicated to the management of digital documentation, as is the case of intelligent agriculture 5.0, intelligent health and banking entities, among others, which will revolutionize private companies in terms of reducing operating costs and also a promising future is foreseen with the Metaverse-6G, transforming communication, driving innovation in commerce and education to build a more inclusive digital world..

ABSTRACT. In this paper we develop an identification technique for the multipath wireless channel utilizing a discrete sum distribution approximation. We focus on three of the main distributions for the multi-path channel: Rayleigh, Rice and Nakagami-m. The distributions are approximated using a weighted sum of a Rayleigh and up to three distributions. The estimation procedure is based on the Maximum Likelihood approach and Akaike's Information Criterion. We show via simulations that when the channel is Rayleigh distributed, the estimation yields the correct Rayleigh distribution. When the channel distribution is Rice, the estimation yields one Rician term, whilst when the true channel distribution is Nakagami-m, the estimation yields an adequate approximation with a Rayleigh and Rice terms.

ABSTRACT. This paper is a review of the state of the art, which considers aspects related to security in the physical layer of wireless sensor networks. For such purposes, the main characteristics of this type of networks are reviewed, as well as the main interference attacks they may suffer, in addition to anti-interference strategies focusing mainly on those based on optimization algorithms and game theory. From the literature review, it is concluded that aspects related to effectiveness, efficiency and practical issues are key, and present a challenge in the design of security strategies in this kind of networks.

ABSTRACT. This work implements a Hash Message Authentication Code (HMAC) to secure Sampled Values (SV) in the IEC 61850 protocol, addressing cyber-attack vulnerabilities in digital substations. Using Python’s Scapy, algorithms were developed to inject and capture multicast SV traffic. HMAC with SHA-256 effectively protected against data tampering, allowing only authorized devices to authenticate messages. Experimental results confirmed real-time attack detection and mitigation. This research enhances SV message security in line with IEC 62351 guidelines.

ABSTRACT. Physical Layer Security (PLS) has emerged in the last years as a field of interest for the development of new methods and techniques to achieve higher security in wireless networks. Many solutions have been proposed and a lot of metrics have been used for sizing the system's confidentiality, such as the Secrecy Capacity (SC) and Secrecy Rate (SR). This paper aims to contribute to PLS development through the analysis of a theoretical model, already existing and selected by a previous analysis of several works, in order to dimension how the SC and SR can be affected by the availability and accuracy of the Channel State Information (CSI), determined through channel estimation techniques.

ABSTRACT. Precision Livestock Farming (PLF) relies on accurate animal behavior data for informed decision-making. However, validating sensor data collected from free-grazing animals (e.g., cows, sheep) can be labor-intensive due to the challenge of verifying animal movements and behaviors. This paper addresses this gap by proposing a novel location-based target technique for PTZ cameras, allowing the tracking of grazing animals in dairy farms. This technique utilizes PTZ (Pan-Tilt-Zoom) cameras to automatically focus on and target individual animals based on their real-time GPS location, enabling precise and automated data collection for behavior identification algorithms. This approach has the potential to significantly improve data quality and streamline the process of training behavior identification algorithms in PLF applications.

ABSTRACT. The article discusses a passivity-based current control (PBCC) approach for the Versatile Buck-Boost (VBB) converter topology. The digital controller design of the VBB has enabled its suitability for various medium to high-power applications. Considering the non-linear dynamics of the VBB converter, the PBCC helps to ensure the input current regulation for power variations. Since controlling the algebraic sum of stored and dissipated energy in the VBB converter is quite challenging, the PBCC plays a crucial role in helping synthesize the control objective such that the closed-loop system can achieve asymptomatic stability and remains passive in nature. Thus, the dynamic performance enhancement could possibly be achieved by injecting a small damping into the system, which helps manage the dissipated energy efficiently. Nevertheless, the digital implementation of PBCC is equally challenging given real-time digital controller design due to the limitation imposed by various factors such as the sampling time delays, choice of numerical integration methods, and the limited memory storage, which is further detailed given the digital current control feedback system. Thus, the simulations results show stable and high performance in both buck and boost modes.

ABSTRACT. Grid-forming inverters (GFI) are a key interface as they are used to supply alternating current (AC) in distributed generation systems and islanded microgrids. Finite Control Set Model Predictive Control (FCS-MPC) has been shown to be a suitable alternative for controlling inverter systems, due to advantages like a multi-variable control, fast dynamic response and intuitive design. However, the performance of MPC has a significant dependence on an accurate mathematical model. To address the challenges of modelling error and uncertainty in predictive control systems, model-free strategies have emerged as a promising solution where the ultralocal model technique is one of the most utilized. In this scenario, a model-free predictive control (MF-PC) based on the ultralocal model and extended state observer (ESO) can effectively reduce model uncertainties. This article proposes an (ESO)-based ultralocal model for the voltage control of a 2-level voltage source inverter (2L-VSI) with LCL output filter. Hardware-in-the-loop (HIL) experiments are used to verify the real-time feasibility and applicability of the proposed MF-PC.

ABSTRACT. The dual active bridge (DAB) converter is being increasingly used in DC traction systems of electric vehicles (EV). To obtain a high-performance output voltage control, the moving-discretized-control-set model-predictive control (MDCS-MPC) has been employed in combination with the single-phase-shift (SPS) modulation. This works describes an implementation of the control system using an FPGA to achieve a high-resolution phase-shift modulation. The evaluation of the controller's dynamic response is carried out using a hardware-in-the-loop (HIL) platform, thereby providing a real-time validation of the MDCS-MPC strategy. The controller's performance is evaluated through steady-state, reference change and load disturbance conditions, demonstrating the feasibility of the predictive controller.

ABSTRACT. This paper presents a double interleaved doubler isolated single-ended primary-inductor converter (DVDiSEPIC) for high voltage gain in fuel cell systems. Featuring an Input-Parallel Output-Series (IPOS) architecture, the converter achieves significant voltage gain, reduced MOSFET voltage stress, and enhanced efficiency with soft switching. Using a high-frequency transformer and Voltage Doubler (VD) cell, the DVDiSEPIC achieves a voltage gain of 21 under ideal conditions. A cascaded closed-loop control strategy with Proportional-Integral (PI) controllers ensures robust voltage regulation. Comparative analysis shows the converter’s advantages in simplicity and performance, with plans for experimental validation and further improvement.

ABSTRACT. The advancement of power electronics field today has allowed us to generate new technologies in various areas such as renewable energies, energy storage systems with batteries, electric chargers, electric vehicle traction systems, and more. New technologies are increasingly demanding in terms of accuracy of system variables; for this, advanced control methods are applied, and this requires a model that correctly represents the operation of the system. The models of power converters are not trivial; therefore, this paper proposes to model a Three-Level Flying Capacitor Buck DC-DC Converter using the bond graph technique, which is a multiphysics modeling tool based on the electrical domain that will facilitate the comprehensive modeling of the system. In addition, the obtained model from the bond graph approach can be directly linked with a Port Hamiltonian System, to obtain a physical representation of the system and based on the conservation of energy. Moreover, results are presented with a proposed non-linear closed-loop control model decoupling the voltage and current dynamics of the multilevel converter. The obtained dynamic results demonstrate the good performance of the proposed modeling methodology.

ABSTRACT. Low Frequency AC Transmission Systems have been reported as an appropriate alternative for transmitting energy over medium distances due to their cost-effectiveness compared to HVAC and HVDC systems. These transmission systems allow for the connection of two different grids. The Modular Multilevel Cascade Converter (MMCC) is presented as a promising solution for this purpose due to its scalability, modularity, and control flexibility. Among the MMCC family, there is a topology that has not been fully studied: the Hexverter, which allows for grid connection with fewer semiconductor devices than other topologies. Therefore, this paper presents a study of the Hexverter and a decoupled control scheme. Simulations and results are analyzed to demonstrate the effectiveness of the proposed strategy.

ABSTRACT. Renewable generation systems have experienced a fast development in the last decades, especially considering the continuous increase in energy demand, the need to reduce the emission of greenhouse gases, and the growing interest in distributed energy systems. In this context, photovoltaic (PV) systems have been consolidated as an attractive alternative to conventional fossil fuel generation due to their high efficiency, modularity, flexible design, ability to generate energy on-site, low maintenance, and increasingly attractive cost. At the same time, the microgrid concept has become popular. A microgrid is an electrical network of power generators and loads that can be grid-connected (connected to an available electrical grid) or isolated (as a local grid).

This talk explores two applications of control design in the context of microgrids: 1) The robust static output-feedback control of a continuous-time uncertain linear model for grid-connected converters with LCL filter based on two-stages technique, aiming to optimize a quadratic performance criterion (LQR), allowing decentralization and magnitude constraints for the gain entries (suitable for practical implementations); and 2) The gain-scheduled dynamic output-feedback control of a continuous-time linear parameter varying (LPV) model for a microinverter-based distributed power generation system, aiming to improve transient response by incorporating D-stability and H∞ criteria into the synthesis conditions.

The design conditions used in both examples are developed regarding Linear Matrix Inequalities. Furthermore, the experiments consider real-world operational characteristics, providing a comprehensive evaluation of LMI-based approaches.

ABSTRACT. The need to improve and innovate the training of new professionals, with better teaching and learning techniques, is frequently raised. The technology incorporated into new equipment for the diagnosis and treatment of patients has a great impact on people's health. And for that, professionals who master knowledge and practice are required. In the case of imaging equipment, the CAT (Computerized Axial Tomograph) stands out, which, due to its complexity, is difficult to understand in relation to the acquisition of digital medical images. The work developed consisted of the design and implementation of a system for the acquisition and reconstruction of images oriented towards teaching and learning. The platform explained offers rotation capacity, with a stationary laser emitter pair; allows you to cut back-projections, and then reconstruct the image using an FBP (Filtered Back-Projection Reconstruction) algorithm. Tests were carried out, and the results obtained are discussed, which allow us to conclude that the proposed objective for teaching purposes was achieved.

ABSTRACT. This paper explores the implementation of cascade control strategies for the Temperature Control Laboratory (TCLab) temperature device. Chemical processes in industrial applications are inherently complex and nonlinear and often involve higher-order dynamics, making them susceptible to disturbances and uncertainties. The TCLab, a widely used educational and research tool, exemplifies these challenges. In this study, we design and analyze a cascade control system to improve the regulation of temperature within the TCLab setup. The primary controller addresses the overall temperature control, while a secondary controller mitigates the effects of disturbances and enhances response time. Experimental results demonstrate the effectiveness of the cascade control approach in achieving superior performance compared to traditional single-loop control methods. This work provides valuable insights into the practical application of advanced control techniques in temperature regulation, contributing to a broader understanding of control systems in chemical process engineering.

ABSTRACT. Learning programming in higher education faces significant challenges, including high dropout rates and difficulties in understanding abstract concepts. Previous studies have explored various teaching methods, but the effectiveness of generative artificial intelligence (GenAI) in this context has not yet been widely investigated. This study compares the efficacy of GenAI with video-based active learning methods for teaching programming to university students. Through an experimental design with 40 computer engineering students, academic performance, perception of usefulness and ease of use (TAM), and satisfaction and motivation (HMSAM) were evaluated. The results showed no statistically significant differences between the groups in academic performance, perception of usefulness and ease of use, or satisfaction and motivation. Both methods proved equally effective in improving learning and maintaining student motivation. These findings suggest that GenAI can be a viable alternative to traditional methods, offering opportunities to diversify pedagogical strategies in programming education. Educators are encouraged to consider integrating GenAI to complement existing methods, ensuring implementation that maintains high levels of perceived control, immersion, and curiosity among students.

ABSTRACT. Robotic object recognition and rearrangement are critical capabilities for industrial automation, warehousing, and logistics operations. This paper presents a novel simulation-based approach towards achieving these capabilities using the widely adopted KUKA KR5 arc HW-2 robotic manipulator. A functional Unified Robot Description Format (URDF) model was developed for the manipulator and its end-effector, enabling seamless integration with the powerful MoveIt! motion planning framework. Additionally, a state-of-the-art vision system was incorporated by leveraging the Kinect 2.0 sensor for object detection and localization. The simulation environment provided a risk-free virtual testbed to develop and refine the robotic system before real-world deployment. The proposed approach demonstrates the potential of simulation-driven methodologies for advancing robotic manipulation capabilities. Comprehensive experimental results validate the system's performance in recognizing and rearranging objects with high precision and efficiency. This work establishes a solid foundation for further research into sim-to-real transfer learning and industrial applications of robotic manipulation driven by simulation and advanced perception.

ABSTRACT. The banana is a crucial crop in tropical regions, facing challenges from diseases such as black Sigatoka, which affect its production and quality due to defoliation. This research proposes the use of drones equipped with high-resolution RGB cameras to capture images of banana plantations, employing a hybrid deep learning model that combines detection and semantic segmentation to accurately identify and count banana leaves. Additionally, the metadata from the images provided the geographical coordinates of each plant, exported in shapefiles compatible with Geographic Information Systems (GIS). The results show high accuracy in detection (98.5%) and leaf counting (93.45%), surpassing previous, more costly methods. This facilitates the identification of areas affected by diseases, evidenced in the detection of potential black Sigatoka outbreaks. The ability to make informed decisions based on this data improves agricultural management, promoting sustainable practices and optimizing crop quality and productivity.

ABSTRACT. This proposal presents the experimental verification of a control scheme to coordinate the variable formation of a set of differential robots. The proposed method applies the leader-follower configuration with unidirectional communication, where the robot closest to the initial point of the trajectory is defined as the leader robot and the others are the follower robots. The leader robot will follow the specified trajectories considering the presence of fixed obstacles and will also send information on its position and orientation to the follower robots in order to maintain the required formations in accordance with the trajectories to follow and the scenario. Finally, the implementation of the control scheme with three TurtleBot3 Burger mobile robots is described.

ABSTRACT. Mining and construction environments are specialized sectors that involve complex operations, heavy machinery, and specific safety considerations. Properly managing different resources (e.g., minerals. concrete, and stones) is essential for successfully executing projects. In this regard, Deep Learning (DL) algorithms emerge as a catalyst in classifying and detecting several resources across the supply chain within hazardous mining and construction environments. This work presents a comparative analysis of different DL techniques used for detecting rocks in challenging mining environments. To this end, three models—MarsNet, YOLOv8+SAM, and SegFormer—have been designed, training and experimentally tested. In addition, we analyze the generalization of proposed algorithm according to selected dataset. Experimental findings disclose that the Segformer algorithm outperforms semantic segmentation algorithms based on other deep learning algorithms regarding main metrics such as recall, F1-score, and Intersection over Union.

ABSTRACT. The advantages that microgrids can deliver and manage within the current electrical grid are well known, and various methodologies and algorithms have been proposed. However, by focusing on the application of a real electrical system, these studies fail to consider simultaneous single-phase and three-phase generation operation, as well as an unbalanced system. Therefore, this paper proposes a distributed secondary control strategy that addresses these previously unexamined topics, ensuring reliable and safe operation. The various simulation results validate the proposed strategy and open the door to expanding the scope of this type of applied research.

ABSTRACT. Electrical fault classification is one of the most complex tasks in electrical systems. In this paper, we propose a classification model based on scalograms using the continuous wavelet transform (CWT) and feature extraction using the EfficientNetV2B3 backbone. Features are then selected using the hybrid metaheuristic algorithm GWO-WOA to maximize the multi-objective function of precision and recall for training a QDA model. The dataset was generated from a three-phase electrical model in Matlab/Simulink, with measurements of currents (Ia, Ib, Ic) and voltages (Va, Vb, Vc). CWT was used to obtain scalograms for each signal, producing a total of 6,480 RGB-type images. The results indicate that the hybrid GWO-WOA algorithm maximizes the performance of the QDA model trained with the selected features, achieving an accuracy of 94%, a precision of 94%, and a recall of 94%. The results for each class indicate an F1-score above 91%.

ABSTRACT. The integration of renewable energy sources and the increasing complexity of power systems require advanced control strategies to ensure efficient and reliable operation. Non-cooperative game theory offers a framework for modeling competitive interactions among distributed energy resources. This paper reviews recent research to highlight the key features, benefits, and challenges associated with various non-cooperative control methods. It offers insights into their distinct characteristics and use cases, helping to identify the most suitable approach for specific microgrid scenarios.

ABSTRACT. This paper presents a novel methodology for determining the optimal sizing and allocation of battery energy storage systems (BESS) in weak power systems with high levels of converter-interface generation (CIG). As renewable energies experience rapid growth in power grids, the integration of CIG poses stability challenges due to fundamental differences between CIG and conventional synchronous generators. BESS offers a highly advantageous solution for their incorporation into electrical power systems. This technology has the potential to minimize operational costs, while also significantly enhancing the reliability of power systems with high levels of CIG. Unlike traditional economic approaches, this methodology incorporates robustness constraints to allocate BESS, aiming to mitigate instability issues arising from weak grid conditions with low short-circuit levels. The proposed methodology offers valuable insights for power system engineers and planners seeking to maintain grid stability while harnessing the benefits of renewable energy integration. The methodology is validated in a model of the Main Chilean Power System (SEN). The results show that integrating BESS into a power system with high levels of CIG contributes to the secure integration of CIG.

ABSTRACT. This study analyzes the technical and economic effects of reducing the power limit for defining free customers in Chile: from five hundred kilowatts to one hundred kilowatts. First, the current regulation is reviewed to set the basis of the analysis. An international comparison with similar countries in the region is also presented as background. The analysis focuses on the impact on the electrical system, customer tariffs, and the national grid operator (Coordinador Electrico Nacional). The ´ results show that (i) distribution companies would forgo estimated revenues of around four hundred million dollars if the limit is reduced to one hundred kilowatts by two thousand twenty-eight and all eligible customers opt to enter the free energy market, (ii) free customers could reduce their monthly bills by between twenty-one point one percent and thirty-five point seven percent compared to current bills, and (iii) the national grid operator will face the challenge of managing around one million sixty thousand five hundred three new contracts, which translates into an estimated demand of approximately four point five terawatthours per year.

ABSTRACT. Sub-Synchronous Resonance is a phenomenon that affects power systems, particularly those with series compensation. This work explores the occurrence of sub-synchronous resonance in the case of Chile that, because of its geography, counts with numerous series compensations. Three thermal generator units are analyzed using the frequency sweep method to assess the risk of sub-synchronous resonance. The results indicate a risk associated with torsional interaction and torsional amplification.

ABSTRACT. This review presents the state of the art about the modeling of Inverter-Based Resources (IBR´s), in the context of bulk power systems with high penetration of variable renewable energy and the impact of these resources on the dynamics. The discussion on a precise modeling through EMT or RMS models this new dynamic - be it detailed or reduced - is still open, a consequence of large-magnitude phenomena whose real systemic response has not been able to be satisfactorily represented.

ABSTRACT. The control of underactuated mechanical systems represents one of the most active fields of research in robotics and control system engineering, with a strong practical interest. Recently, methods based on differential flatness theory have been found to be useful in nonlinear control design for marine vehicles. However, to the best of the authors' knowledge, existing works based on differential flatness theory have not considered the presence of external non-zero mean disturbances or parametric uncertainties in the controller design for non-flat underactuated marine vehicles. Within this context, this work aims to explore a scenario of trajectory tracking for a non-flat and nonlinear model of an underactuated surface ship in the presence of external disturbances and parametric uncertainties. By exploiting the Liouvillian and flat properties of the linearized system, we propose a tracking controller with integral action. Numerical simulation results show the capability of the proposed controller to overcome common problems of real-world systems.

ABSTRACT. This paper introduces an improved approach to characterize the global stability of positive linear discrete-time systems subject to actuator saturation. The ramp function is used to represent saturation, allowing a precise representation of this nonlinearity. Then, the copositivity of quadratic forms and piecewise quadratic Lyapunov functions depending on both the states and ramps are used to establish sufficient conditions for global stability. Numerical comparisons with other techniques from the literature are performed to illustrate the effectiveness of the proposed conditions when compared with the existing results.

ABSTRACT. In this paper, we study the effect of estimating the states of a nonlinear system when the number of states that are present on the output varies. Specifically, we examine the case of a micro-channel water tank, where we simulate scenarios of getting information from all tank levels or only some of them. Our goal is to achieve precise estimations of multiple states corresponding to the tank levels, using, in the worst case, only information from one tank. To this end, we simulate using three state estimators: the Extended Kalman Filter, the Unscented Kalman Filter, and the Particle Filter, to determine which is more accurate in scenarios with limited information. In this way, we aim to explore the possibility of reducing the number of information required in the implementation of such systems.

ABSTRACT. Most electric vehicles are powered by lithium-ion batteries due to their characteristics such as high power and energy density, life cycle and reliability. However, these characteristics imply complex mechanisms that require robust control to ensure safe operation. Thus, it is imperative to construct models that accurately represent the complex dynamics of these batteries. Linear Parameter Varying (LPV) models are advantageous as they preserve a linear interrelation among system states while accommodating their temporal variability. This paper presents a novel LPV model of lithium-ion batteries that uses the state of charge as a scheduling parameter. Experimental results are obtained using a downscaled prototype composed of a charger and a battery cell. These results are utilized to validate the effectiveness of the proposed LPV model.

ABSTRACT. This paper introduces an extention of the existing two port ”Dynamic Energy Router” (DER) applied to a three port system in the context of an electric vehicle integrating a hydrogen fuel cell and a supercapacitor. The control strategy implemented is feedback linearization to manage the nonlinear dynamics of the system and the efficient energy distribution between the multiports. By incorporating voltage control mechanisms in one port, this method addresses the challenges of the DER energy decay and enhances the overall performance of this device. MATLAB/Simulink simulation results demonstrate the effectiveness and robustness of the proposed control strategy in maintaining constant DC link voltage and achieving precise current tracking of the currents given by the DER. This work contributes to extend the two port DER to a three port DER by using a Feedback Linearization control technique, providing a versatile solution for multiport system where each port is capable of generate, store and consume energy.

ABSTRACT. This study focuses on addressing the sampled-data admissibility issue for a singular system. The primary objective is to develop a sampled-data controller that ensures the admissibility of such systems, utilizing an appropriate Lyapunov–Krasovskii functional (LKF) to achieve less conservative outcomes. To achieve these goals, the system is first transformed into a time-delay system using the input delay approach. Subsequently, suitable admissibility criteria are formulated by introducing an appropriate LKF. To mitigate conservatism during the estimation of the LKF derivative, additional slack variables are introduced using the Finsler lemma. Finally, two numerical examples are provided to illustrate that the designed sampled-data controller effectively guarantees the admissibility of the considered systems.

ABSTRACT. This paper addresses the reference tracking control problem for medical cyber-physical systems (MCPS). The control theory is employed to guarantee the suitable concentration of drugs in the body of patients to guarantee a safe treatment. The MCPS is modeled as a switched system, and the modes consider the different scenarios for the problem. A discrete-time model is utilized for the pharmacokinetic process, and the zero input control strategy is employed to design state-feedback controllers with a guaranteed exponential convergence rate. A numerical experiment is presented to illustrate the validity and effectiveness of our method.

ABSTRACT. Hoy en día, ingenieros, investigadores y científicos manejan grandes cantidades de datos provenientes de diversas fuentes, como sensores, imágenes, videos, telemetría y bases de datos. El Machine Learning es clave para descubrir patrones ocultos en esos datos y construir modelos que predigan resultados basados en información histórica.

En este seminario, exploraremos conceptos fundamentales de Machine Learning utilizando MATLAB, una herramienta poderosa y versátil. A través de MATLAB, podrás explorar, analizar y evaluar tus datos de manera eficiente, aplicando técnicas avanzadas para resolver tus problemas de forma rápida y precisa.

¿Qué cubriremos?

- Entrenamiento y evaluación de modelos: entrenaremos, evaluaremos y compararemos distintos modelos de Machine Learning.

-Optimización y reducción dimensional: veremos cómo refinar modelos usando técnicas de reducción para maximizar su capacidad predictiva.

- Procesamiento en paralelo: aceleraremos los resultados ejecutando modelos predictivos en paralelo con múltiples procesadores.

-Implementación en producción: conoce cómo desplegar tus modelos en sistemas reales y embebidos para aplicaciones del mundo real.

-Dónde conseguir más información, incluyendo cursos en línea.