Keywords: Networks optimization, Robust control and Hinfty control, Optimal control
Abstract: This study focuses on robust control design for uncertain Cyber-Physical Systems (uCPS)subjected to DoS attacks via Linear Matrix Inequality (LMI) approach. The CPS system under attacks, due to ambiguities in its nature, may be modeled as a switching system and designed following the modes of the system under attack, restricting the malicious agent’s ability to execute continuous DoS attacks. Various methods have been developed to minimize the effects of violence on the system. A numerical example is provided to demonstrate the effectiveness in stabilizing uCPS systems under Denial of Service (DoS) attacks

Keywords: Networks optimization, Optimization, Power systems
Abstract: In this paper, the multi-objective distribution network reconfiguration (DNR) problem was addressed using a multi-objective Ant Lion Optimizer (MOALO) algorithm. The proposed approach was tested in the IEEE 33-bus system in order to simultaneously minimize active power losses and enhance reliability while taking into account a set of operational and topological constraints. To calculate power losses, the Backward/Forward algorithm was applied. Moreover, the union-find with path compression approach was used to keep radiality of each network configuration. To evaluate the performance of MOALO, it was benchmarked against the weighted-sum method (using Genetic Algorithm (GA)), Non-dominated Sorting Genetic Algorithm II (NSGA-II), multi-objective Particle Swarm Optimization (MOPSO) and multi-objective Grey Wolf Optimizer (MOGWO). The research findings revealed that the proposed MOALO algorithm effectively reduces total power losses by up to 31.14% and the total ENS index by 33.35%, while also providing a well-balanced compromise solution with power losses around 141.92 kW and improved reliability with a total ENS index of approximately 5037 kWh/year.

Keywords: Observer design, Robust control and Hinfty control, Linear and nonlinear systems
Abstract: This paper presents the design of a low-order estimator for predicting structural loads, such as the wing root bending moment, for maneuver loads alleviation purposes, considering a structurally and aerodynamically non-linear model under multiple flight conditions. This was accomplished utilizing a multi-model structured Hinfty-H2 approach, which involved the use of several linear reduced-order models. The estimator's performance was assessed in a non-linear simulation framework at three different flight conditions, both with and without a simple maneuver load alleviation function, showing satisfactory results. Finally, a noise sensitivity study of the estimator was conducted.

Keywords: Robust control and Hinfty control, Fuzzy and neural systems, Control applications
Abstract: By furthering previous research on linear and nonlinear systems, this study focuses on addressing continuous time-delay nonlinear systems robust dependent delay stability and stabilization conditions through the utilization of a smart decision sensor-based technique. The paper introduces new conditions for relaxed robust dependent delay stability and stabilization,which are discussed within the framework of Linear Matrix Inequalities to tackle uncertainties in model parameters. Robust stability criteria are formulated to ensure system performance in the presence of uncertainties and delays. State feedback stabilization criteria are presented for practical application, aimed at preserving system stability. Additionally, observer-based stabilization feedback is designed to estimate unmeasured states and ensure robust control. Robust state feedback stabilization criteria are suggested to manage uncertainties effectively. The study conducts numerical simulations to validate the robustness and effectiveness of the proposed control methodologies across various scenarios. The outcomes demonstrate notable enhancements in system stability and performance, affirming the viability of the proposed approach.

Keywords: Renewable Energy, Fault detection and Diagnostics, Intelligent and AI based control
Abstract: Continuous monitoring is necessary to guarantee the safety and dependability of photovoltaic plants, but many flaws go unnoticed in spite of current regulations. In order to improve fault detection and diagnosis, this study integrates artificial intelligence (AI) and the Internet of Things (IoT). We systematically assess sophisticated algorithms, such as machine learning and deep learning methods, with an emphasis on their accuracy, cost, complexity, and viability for real-time implementation. Our strategy shows that integrating AI and IoT solutions into reasonably priced hardware platforms provides a scalable and efficient way to monitor distant photovoltaic plants, greatly lowering maintenance expenses and enhancing fault management.

Keywords: Renewable Energy, Control applications, Estimations and identification
Abstract: Accurate speed feedback is vital for effective electric motor drive control. Due to the drawbacks of mechanical sensors, software speed sensors based on observer algorithms have gained prominence for their simplicity, speed, and robustness. This paper investigates and compares two new hybrid structures for robust speed estimation across all operating ranges. The first combines Model Reference Adaptive System (MRAS) and Sliding Mode Observer (SMO) into one coherent SMRAS estimator by replacing the MRAS reference model with an SMO. The second introduces a speed-based switching scheme between SMO and the hybrid SMRAS estimator continuously driven by speed instruction signal.

Keywords: Renewable Energy, Modeling and simulation, Power systems
Abstract: This paper details a dynamic simulation analysis of an alkaline electrolyzer powered by a hybrid renewable energy system consisting of a solar photovoltaic array coupled with a battery storage unit. The primary of this study was to assess hydrogen production performance via water electrolysis under varying energy availabilities. The model was developed using MATLAB/Simulink and Simscape Electrical, integrating essential subsystems such as MPPT-based solar control, battery energy management, electrolyzer dynamics, and purge logic. The simulation spanned a 7-day timeframe during which the system operated in an isolated DC microgrid setup. The results indicate that the battery effectively supports the electrolyzer during periods of low irradiance, ensuring uninterrupted hydrogen generation. The current and voltage profiles validated the coordinated operation of the renewable source and energy storage system. Furthermore, the state of charge of the battery and the hydrogen mass profile illustrated the reliability and storage efficiency of the proposed system. Thus, this study validates the potential of autonomous green hydrogen production using renewable energy sources, particularly in areas with abundant solar energy. The proposed configuration can serve as a basis for future advancements in decentralized and sustainable energy system research.

Keywords: Modeling and simulation, Renewable Energy, Power systems
Abstract: The present study investigates a hybrid photovoltaic (PV) lithium-ion battery for the purpose of powering a water pumping mechanism, focusing on detailed modelling, control strategies, and performance evaluation for sustainable irrigation. The system incorporates an induction motor pump, a three-phase inverter, a bidirectional converter, a DC/DC boost converter, and a solar array. In order to maximise the amount of solar energy extracted, an Incremental Conductance (INC) Maximum Power Point (MPPT) method is applied, whilst PI controllers are employed to regulate the DC bus voltage and the battery current. Direct Torque Control (DTC) is implemented to ensure a fast and robust motor response. The complete system has been modelled in MATLAB/Simulink and tested under conditions of variable irradiance and temperature. The simulation results demonstrate effective MPPT tracking, balanced power sharing between the photovoltaic (PV) and battery systems, and reliable water pumping with smooth torque and speed dynamics. The findings of the present inquiry lend substantiation to the efficacy of the proposed method in delivering a unsoiled, autonomous and dependable agricultural irrigation resolution.

Keywords: Motion control, Modeling and simulation, Transportation systems
Abstract: This paper presents a double-loop based sliding mode control approach to improve the tracking performance in autonomous surface vessels (ASVs) within prescribed constraints. The error system is constructed on a prescribed performance function meant to define constraints on the error states. A double-loop control structure is used - wherein a virtual velocity command designed to stabilize the outer loop based on the kinematic model, is fed as a reference into the inner loop based on the dynamic model. Both inner and outer loops are stabilized by distinct sliding surfaces, the stability of which is proved using Lyapunov theory. The stability result reveals uniform ultimate boundedness of all error signals in the system in finite time, with rejection of uncertainties, disturbances and input saturation, while ensuring prescribed performance guarantees. Numerical simulations on a 3-DOF ASV validate the proposed approach.

Keywords: Motion control, Control applications, Linear and nonlinear systems
Abstract: This paper proposes a wheel slip control-based Reference Offset Accumulation Torque Vectoring (ROA-TV) control method for electric vehicles (EVs). Existing torque vectoring control methods have primarily been studied in configurations with independent left/right control, such as in-wheel motors, which often limits their practical applicability to commercial EVs. Furthermore, these methods frequently suffered from issues like excessive wheel slip or insufficient resolution in wheel slip calculation. Additionally, to generate the desired longitudinal force and lateral moment, torque allocation for all four wheels was necessary, typically requiring optimization methods such as Quadratic Programming (QP). To overcome these limitations, the proposed ROA method introduces a novel approach for reference generation by creating and accumulating longitudinal and lateral offsets for each wheel. This eliminates the need for QP optimization, thereby enabling the proposed method to run on in-vehicle microprocessors. The proposed method was implemented in a real vehicle and tested on a simulated low-friction road, demonstrating clear advantages in yaw moment tracking and acceleration performance compared to other control methods.

Keywords: Modeling and simulation, Linear and nonlinear systems, Fuzzy and neural systems
Abstract: This study presents a robust control strategy for Diesel engine injection systems in ship applications, focusing on the regulation of pump pressure through a state feedback control approach with an H_infty performance criterion. The objective is to maintain the pump pressure at a desired reference by adjusting the engine speed. To accurately capture the nonlinear dynamics of the system, a Takagi--Sugeno (TS) fuzzy model is employed. Based on this model and a descriptor framework, a state feedback controller is designed to guarantee precise pressure regulation while satisfying H_infty performance requirements expressed as linear matrix inequalities (LMIs). The approach specifically addresses the rail injection system, which enables precise control of fuel delivery and timing. Simulation results demonstrate the effectiveness and robustness of the proposed strategy in maintaining accurate pump pressure despite bounded external disturbances.

Keywords: Intelligent and AI based control, Power systems, Modeling and simulation
Abstract: Solar flares and coronal mass ejections can lead to geomagnetic storms that disrupt power transmission grids. We present MagnoStromNet, a dual-stream attention-based deep- learning framework that combines dynamically varying space weather parameters with static solar event features to provide three risk levels to grid disruption — Low, Medium or High. The deep-learning framework, developed with a specialized dataset of solar wind and geomagnetic indices and flare characteristics, demonstrates to be highly accurate with a test accuracy of 98% and macro-precision, recall and F1-score of 0.97 values that are higher than LightGBM, XGBoost, Random Forest and CNN baselines. The use of cross-attention provides easy to understand feature attributions that are confirmable with LIME. Advanced interpretations provide accurate and actionable early warnings, enhancing global resilience-building capabilities of electrical infrastructure against storm hazards.

Keywords: Intelligent and AI based control
Abstract: This study proposes a hybrid machine learning model, PCA-LSTM, aimed at enhancing indoor temperature forecasting using a set of meteorological inputs. The model

combines Principal Component Analysis (PCA) for reducing input dimensionality and avoid data redundancy with Long Short-Term Memory (LSTM) network capable of learning temporal dependencies within temperature time series. It is

trained on hourly observations comprising seven environmental variables: outdoor temperature, relative humidity, dew point temperature, wind speed, visibility, atmospheric pressure and electric heater consumption. The results highlight the model’s ability to deliver accurate predictions, outperforming conventional methods. With this high level of accuracy, the

PCA-LSTM model is well suited for integration into an Energy Management System (EMS), supporting intelligent control strategies to optimize energy consumption while preserving indoor thermal comfort.

Keywords: Fuzzy and neural systems, Modeling and simulation, Power systems
Abstract: This paper presents a robust fuzzy logic-based control strategy for a single-phase full-bridge inverter synchronized to the electrical grid using a Second-Order Generalized Integrator Phase-Locked Loop (SOGI-PLL). The control objective is to achieve accurate voltage tracking, precise phase synchronization, and minimal harmonic distortion at the inverter output. The fuzzy logic controller (FLC) is constructed with triangular membership functions and a comprehensive rule base, enabling real-time adaptive decision-making under dynamic grid conditions. The SOGI-PLL is employed for extracting the phase and frequency of the grid voltage, ensuring stable synchronization even in the presence of voltage sags or waveform distortions. MATLAB/Simulink simulations validate the proposed system’s performance in terms of current tracking, voltage quality, and harmonic mitigation. Comparative results against a classical Proportional-Integral (PI) controller highlight the FLC’s superior dynamic response, reduced Total Harmonic Distortion (THD), and enhanced phase alignment. The proposed FLC-SOGI-PLL scheme proves effective for grid-tied inverter systems, especially in smart grid and renewable energy applications.

Keywords: Intelligent and AI based control, Control education
Abstract: This paper presents a problem-based learning (PBL) strategy for teaching Artificial Intelligence (AI) using Orange Data Mining, a visual and interactive tool designed to lower the barrier for students with limited programming experience. The method centers on real-world problems, allowing students to explore AI concepts such as classification, clustering, and model evaluation through hands-on experimentation. Implemented in an undergraduate course, the approach engaged students in collaborative problem solving using real datasets. Evaluation through pre- and post-tests, project work, and surveys indicated improved understanding, increased motivation, and practical modeling competence. Student feedback highlighted the accessibility of Orange and the effectiveness of PBL in making AI concepts more approachable. The results suggest that integrating visual tools with problem-based pedagogy can offer an inclusive and impactful pathway for AI education across diverse academic backgrounds

Keywords: Observer design
Abstract: This paper addresses the problem of simultaneously estimating the state variables and the time-delay of a class of time-delay nonlinear systems with sampled outputs involving an unknown time-delay. This observer is composed by a suitable high gain adaptive observer cascaded with an appropriate cascade of predictors. The former performs an admissible estimation of the delayed state variables,while the latter provides an admissible estimate of the actual state variables. The observer convergence is investigated thanks to an adequate Lyapunov approach with respect to the considered sampling. The proposed observer performances are evaluated by probing simulation results involving a coronary artery system.

Keywords: Observer design
Abstract: This paper deals with the trajectory analysis of crystallization process. The considered system is CSD ( crystal size distribution) of a batch crystallization process. It is modeled as a (time-varying) hyperbolic partial differential equations. It is mainly shown that the trajectory exists on the whole (nonnegative real) time axis for all strictly positive integrable growth rate. Both the control and the observation operators are unbounded but admissible. Thus, from the theory of time-invariant and/or time-varying control systems, existence and uniqueness of the state trajectory (CSD), the exponential (or internal) stability and the observability of crystallization process are analyzed. The analysis essentially uses the notion of the evolution family of operators involved in the dynamics and the concept of the (regular) time-invariant and/or time-varying well-posed systems. Finally, a time-varying Luenberger type observer was designed to estimate the CSD in batch crystallization processes. The observer is based on PDE models, in contrast to the discrete approximation method, which results in finite-dimensional models (cf. Bakir et al. (2006)).

Keywords: Observer design
Abstract: Accurate trajectory tracking in rehabilitation exoskeletons remains challenging due to measurement noise, nonlinear dynamics, and unpredictable disturbances arising from human–robot interaction. This paper presents a novel control architecture that combines a flatness-based control law with cascaded filtered observers to enhance robustness and estimation accuracy. A Filtered High-Gain Observer (FHGO) is developed to estimate joint angular velocities while attenuating the effect of measurement noise. In parallel, a Disturbance Observer (DO) is designed to estimate abrupt external torques resulting from the interaction between the exoskeleton and the user. The estimated states and disturbances are then integrated into the flatness-based control law to ensure precise and robust trajectory tracking. The effectiveness of the proposed strategy is validated through numerical simulations under a realistic scenario. Simulation results demonstrate excellent tracking performance and strong disturbance rejection capability, highlighting the potential of this control structure for real-time rehabilitation assistance.

Keywords: Observer design
Abstract: Unmanned aerial vehicles (UAVs) rely heavily on the proper functioning of their rotors to maintain stable and safe flight. This paper addresses the crucial issue of rotor fault detection and diagnosis in UAVs by integrating fault observation algorithms with identification technique. We propose an integrated approach that enables real-time detection and classification of common rotor faults. The effectiveness of the proposed method is validated through Matlab simulations and compared with existing state-of-the-art methods. The results demonstrate improved detection accuracy and faster fault iso lation, emphasizing the effectiveness of the proposed algorithms in improving UAV operational reliability and safety. Index Terms—Fault estimation, Fault identification, Fault tol erent control, Disturbance Observer, UAV’s

Keywords: Observer design
Abstract: This work presents a fault-tolerant control strategy for the lateral dynamics of autonomous vehicles subject to steering actuator faults. A Saturated Filtered High Gain Observer (SFHGO) is proposed to estimate both the full state and the fault signal in a nonlinear vehicle model. The proposed nonlinear observer provides accurate and fast estimation under unknown steering actuator failure. The estimated fault is actively compensated through a nonlinear feedback control law, ensuring robust trajectory tracking performance even in the presence of faults. The effectiveness and robustness of the proposed Active Fault-Tolerant Control strategy are demonstrated through simulations and validated using real data on the SHERPA simulator, confirming its potential for real-world autonomous applications.

Keywords: Observer design
Abstract: This paper presents an implementation of two high-gain observers, the Filtered High-Gain Observer (FHGO) and the Saturated Filtered High-Gain Observer (SAT-FHGO), for trajectory tracking of a quadrotor unmanned aerial vehicle (UAV) subjected to disturbances and measurement noise. The quadrotor’s dynamics are modeled as a nonlinear system that incorporates aerodynamic drag, control inputs, and external perturbations. We develop a robust control framework by integrating disturbance estimates from the SAT-FHGO and FHGO with a proportional-derivative (PD) controller and a supertwisting algorithm, thereby enhancing disturbance rejection and noise attenuation. Each observer estimates unmeasured states and time-varying disturbances, enabling accurate tracking even in the presence of large initialization errors and high-frequency noise. The proposed approach achieves noise reduction proportional to θ^(r+1), where θ is the chosen high-gain parameter and r denotes the order of the filter, offering a substantial improvement compared to the standard high-gain observer. Under appropriate conditions on the selection of θ and r, this can lead to up to 90% noise reduction. Simulation results demonstrate that the SAT-FHGO outperforms both the Standard High- Gain Observer (SHGO) and the FHGO by significantly reducing peaking phenomena and improving steady-state stability. The approach guarantees bounded estimation and robustness against model uncertainties, making it well suited for real-world UAV applications.

Keywords: Signal processing
Abstract: Financial time series prediction is an important problem in economy. These signals are non-stationary. Even though there are several methods in the literature, the search for performance improvement is a current concern for researchers. This paper considers preprocessing these signals through decomposition methods such as Empirical Mode Decomposition (EMD) and Multi Resolution Analysis (MRA), which improves forecasting, and proposes a new modification to the EMD which is able to improve performance. Classical ARIMA and deep learning methods such as Long Short Term Memory (LSTM) and Gated Recurrent Unit (GRU) are used as forecasting methods.

Keywords: Robotics, Autonomous Systems, Control algorithms implementation
Abstract: This communication on collision avoidance with unexpected obstacles is motivated by some critical appraisals on reinforcement learning (RL) which ``requires ridiculously large numbers of trials to learn any new task'' (Yann LeCun). We use the classic Dubins' car in order to replace RL with flatness-based control, combined with the HEOL feedback setting, and the latest model-free predictive control approach. The two approaches lead to convincing computer experiments where the results with the model-based one are only slightly better. They exhibit a satisfactory robustness with respect to randomly generated mismatches/disturbances, which become excellent in the model-free case. Those properties would have been perhaps difficult to obtain with today's popular machine learning techniques in AI. Finally, we should emphasize that our two methods require a low computational burden.

Keywords: Signal processing, Control applications
Abstract: Muscle fatigue alters muscle activation patterns, resulting in a significant shift in the spectral contents of the surface electromyogram (sEMG) signals. This poses a challenge for the task discrimination algorithms to correctly interpret the movement intent for the control of prosthetic limbs. In this study, we investigated the effect of muscle fatigue on the classification accuracy of two supervised machine learning (ML) algorithms, i.e., the linear discriminant analysis (LDA) and the muscle synergy-based task discrimination (MSD). The results show that muscle fatigue caused the classification accuracies to decline from (>90%) to (<50%) for the LDA and the MSD algorithms. We observed that improved classification accuracies could be achieved by adaptively switching between the two algorithms. In particular, LDA offered superior task discrimination in normal operation and low muscle fatigue, whereas MSD offered improved results during moderate to high muscle fatigue. Further investigation may be needed to provide a more robust switching regime between the two algorithms.

Keywords: Robotics, Optimization
Abstract: The paper considers the design problem of accelerated iterative learning control (ILC) for a discrete system with input backlash that is characteristic of actuators. In order to accelerate the convergence of the learning error, a new approach to the solution is proposed, using a combination of the heavy ball method from optimization theory and the vector Lyapunov function method previously developed by the authors. The formulation of the problem is motivated by the development trends of modern robotic smart and additive manufacturing, as well as rehabilitation medical robots. An example is given confirming the effectiveness of backlash compensation by the accelerated ILC algorithm and including a comparison with known results.

Keywords: Optimization, Optimal control, Linear and nonlinear systems
Abstract: This paper proposes a new approach to derive a bang-bang-type solution using a saturating function. The proposed method is validated through a spacecraft interplanetary transfer trajectory problem. The problem is formulated as a two-point boundary-value problem (TPBVP) and is reduced to an unconstrained basis-function representation by using the Extreme Theory of Functional Connections (XTFC). It is solved by minimizing the residuals of the governing differential equations using nonlinear least squares. Saturation with hyperbolic tangent functions is introduced to represent solutions with bang-off-bang-type structures in terms of continuous basis functions. This approach allows for more robust convergence than conventional methods without requiring additional cost variables or prior knowledge of switching events. The method highlights its potential as a versatile tool to automate spacecraft on-board optimal trajectory design and interplanetary exploration mission design.