One of the most critical security problems today concerns our data on the internet. Cybercrimes, perpetrated through malicious programs, have become commonplace, posing risks to personal and public information during data exchanges or in data storage. In recent studies, the worm propagation dynamics in cryptovirology in blockchain systems could be successfully modeled by means of three-dimensional fractional-order chaotic systems based on the SEI nonlinear epidemic representation. This paper introduces an adaptive sliding mode control design for the robust stabilization of fractional-order cryptovirology in blockchain systems in presence of disturbance. The problem of stabilizing chaos in cryptovirology systems is solved analytically following the methodology of Lyapunov's theorem. We were able to show that the application of the proposed adaptive SMC control approach allows the system states to converge towards stable values. Furthermore, this is largely validated by numerical simulations on Matlab/Simulink.
Centralised and decentralised controller design problems are considered for distributed-time-delay systems to robustly track certain references and to reject certain disturbances. Solvability conditions for these problems and the controller structures which solve each of these problems are first presented. It is then shown that both the solvability conditions and the controller structure which solves the problem under the centralised control structure are special cases of those under the decentralised control structure.
This paper investigates the (3+1)-dimensional potential Yu-Toda-Sasa-Fukuyama (YTSF) equation using the Bilinear Neural Network Method (BNNM). This novel hybrid framework integrates Hirotas bilinear formalism with neural network modeling. By reformulating the YTSF equation into its bilinear form and embedding this structure into the BNNM architecture, multiple classes of exact analytical solutions are derived, including kink-type, periodic, and rational forms. The proposed method yields closed-form solutions that preserve mathematical rigor while improving computational efficiency for high-dimensional nonlinear evolution equations. The results demonstrate the effectiveness of the BNNM in generating diverse solution structures for the YTSF equation, providing potential applications in fluid dynamics, plasma physics, and other nonlinear physical models.
Industrial Control Systems (ICS) and SCADA networks underpin operational technology, yet intrusion detection system (IDS) research keeps reporting strong laboratory numbers while sidestepping deployment blockers—strict false-alarm budgets, real-time latency, nonstationary plant behavior. We present an operational sensitivity audit of an explainable hybrid IDS pipeline, treating windowing/stride, thresholding, and SHAP-based explanation generation as coupled components. On two labeled ICS datasets (CIC Modbus and TU Wien), 10 s to 30 s stride changes rarely flip attack-window decisions (2.15% and 3.87%), but on a benign-only IEC-104 trace the false-alarm rate jumps from 1.50 to 3.55 per hour—FAR is not stride-invariant. Global SHAP explanations stay stable under stride perturbation, yet decision-critical flip-window explanations do not (median cosine distance 0.161 on TU Wien). A KS drift detector flags an injected mean shift with zero pre-drift false alarms. Inference and fusion run sub-millisecond; every claim maps to a regenerable artifact.
Many power electronic applications, such as locomotives and Hybrid Electric Vehicles (HEVs), are moving towards a Multi-Input Multi-Output (MIMO) integrated DC-DC converter for its reliability and flexibility. This work proposes a Three-Input Integrated Bidirectional DC-DC (TIIBD) Converter to combine a Photovoltaic (PV) panel, Fuel Cell (FC) stack, and battery for grid-connected Hybrid Energy Microgrids (HEMs), enabling bidirectional power flow and energy storage. The TIIBD converter is modeled using state-space analysis, yielding a Transfer Function Matrix (TFM). PV and FC are optimized for maximum power via the P&O MPPT algorithm and CRCBP-based PI control. A 3-level Neutral Point Clamped (NPC) Inverter is used for power management between the grid and the TIIBD converter. The controllers for the three inputs are designed, and the steady-state performance of the converter with an AC load of 10KW for Microgrid (MG) application in different operating conditions is analysed and verified in MATLAB/Simulink platform.
This research presents the design and implementation of a cyber-physical system based on IoT technologies and computer vision aimed at the early detection of drowsiness in land transport drivers with applications in the field of occupational e-Health. The system integrates a camera, a facial analysis model in Python using Mediapipe and OpenCV, and embedded devices such as the ESP32 for continuous monitoring of visual signals to detect eye closure or yawning. The proposed architecture includes real-time communication via the MQTT protocol, triggering local alerts (sound and light), storing visual evidence, and sending notifications to Telegram. In addition, a remote monitoring interface using Node-RED and a mobile application in Kotlin for local event visualization are implemented. Functional tests were conducted, demonstrating the viability of the system as a preventive tool for reducing fatigue-related risks in the context of monitoring ground vehicle drivers. The solution's modular and portable approach allows for adaptation to other biomedical monitoring environments and suggests a low-cost alternative for digital health applications.
Bluetooth is a technology that has passed multiple iterations to reach its current state. It all began back in the late 90s when there was a growing need to begin replacing cables with seamless wireless communication. From it’s first release back in 1999, it has already reached version 6.x. Each consecutive version brought qualitative and quantitative improvements, new protocols, use cases, higher bandwidth and better optimization. Currently, the technology is found in nearly all smart devices spanning numerous devices such as smartphones, laptops, smart wearables, audio devices and many others. The latest updates to the Bluetooth stack made it a viable option for the Internet of Things (IoT), cyber-physical systems (CPS) and robotics. This paper delivers a detailed overview of the technical development of the technology, its architecture and real-world deployment use cases. This overview paper is based on 34 recent scientific publications and goes over Bluetooth’s way of enabling smart environments, synchronized control systems, industrial sensing, smart wearable device and autonomous robotic coordination. The paper also reviews how Bluetooth is performing when there are constraints presented such as low power, limited spectrum, dynamic topologies and latency sensitivity. It is also noted how Bluetooth is used in increasingly complex real-world use-cases and in decentralized applications. The combined synthesis brings a better understanding of the role of the technology in shaping our current and future world.
In view of the increasing number of accidents and fatalities on the roads, the need for the application of automatic road safety surveillance has also increased. The paper proposes a multi-model system for improved road safety, wherein deep learning models, primarily based on Convolutional Neural Networks (CNN), are used for the enforcement of critical safety parameters. The parameters include helmet usage, seat belts, lane discipline, and face mask usage. The proposed research is based on the development of individual detection models. However, the models are made to run either individually or in parallel, thereby ensuring real-time checking for the enforcement of the aforementioned parameters in the presence of demanding vehicles and conditions. The proposed model, despite being based on real-time operation, has the advantage of ensuring ease of operation without the requirement for excessive computation, thus helping law enforcement agencies in pursuing violators for the aforementioned offenses.
Magnetic levitation systems are nonlinear and open-loop unstable, which makes accurate and robust control difficult. This paper investigates two sliding-mode-based controllers for a nonlinear MagLev system: the Super-Twisting Sliding Mode Controller (STSMC) and a Conditioned Barrier Adaptive Function Double-Integral Terminal Super-Twisting Sliding Mode Controller (CB-STSMC). Both controllers are designed to regulate the levitation gap while ensuring proper tracking of magnetic flux and momentum states. Their gains are optimized using a Genetic Algorithm with the Integral of Time-weighted Absolute Error as the fitness criterion. Closed-loop stability is established through Lyapunov analysis. MATLAB/Simulink results show that both controllers achieve stable levitation, while the proposed CB-STSMC provides faster convergence, smaller overshoot, and improved disturbance rejection compared with the conventional STSMC. These results confirm the effectiveness of the proposed strategy for robust and precise nonlinear MagLev control.
Overlapping decentralized controller design is considered for large-scale nonlinear time-delay systems (NLTDSs). Principles of inclusion, limitation, and completion are first defined for such systems. It is then shown that limitation and completion are, in fact, special cases of inclusion. Conditions for an NLTDS to be a limitation or completion of another such system are then derived. Four different stability definitions are considered for NLTDSs. It is shown that if a stable (in any one sense considered) NLTDS includes another such system, then the latter system is also stable. It is then shown that if an NLTDS is a limitation of another such system, then reduction of any controller is assured. Furthermore, this reducted controller stabilizes the former system if the original controller stabilizes the latter system. Finally, overlapping decompositions of large-scale NLTDSs are considered.
The paper considers a quadratic criterion optimal control problem for an inhomogeneous equation of vibrations of a three-layer plate. The existence and uniqueness of the solution of a boundary value problem were studied for each fixed admissible control. Then the existence theorem of optimal control was proved, and a necessary condition for optimality was derived in the form of an integral inequality.
This research study examines global asymptotic stability of Cohen–Grossberg neural networks that involve both discrete time delay terms in neuron states and neutral delay terms in the time derivatives of neurons’ states. In this context, a suitable Lyapunov functional, constructed as a linear combination of three complementary main Lyapunov functional candidates, is employed to present new alternative sufficient criteria guaranteeing global asymptotic stability of neutral-type neural system possessing discrete delay components. The obtained criteria are established through explicit algebraic inequalities that utilize key matrix properties and structural characteristics of the system functions. The proposed results are basically stated in terms of parameters associated with the considered neutral neural system. These conditions are completely independent of the delay terms and can directly be tested by checking a set of simple algebraic inequalities. In order to illustrate some efficiency aspects of the derived stability results, a numerical example is analyzed.
The growing number of satellites orbiting the Earth has necessitated sophisticated methods for monitoring and controlling their relative locations and velocities. Precise assessment of a satellite's relative orbital states is essential for a variety of uses, such as satellite formation flying, collision avoidance, and rendezvous operations. In this study, a Pseudo-ranging method was employed to determine the orbital positions of the satellites. Satellite orbital state estimates were made separately using the Multiple Scaling Kalman Filter and the Linear Kalman Filter. Then, relative inter-satellite state estimations were made by adding uncertainty disturbances, and the results of the methods were compared.
The rising awareness of green energy has boosted the use of distributed storage and renewables in networks and microgrids. As a result, the rising use of power electronic devices has significantly contributed to declining power quality (PQ) in distribution networks. The photovoltaic (PV) integrated unified power quality conditioner (UPQC) presented in this paper operating using a flexible compensatory strategy built on the adjustable leaky least mean square (ALLMS) algorithm. This soft-computing approach attains faster convergence by updating parameters iteratively within defined bounds. The ALLMS-based algorithm generates UPQC control signals without needing low-pass filters to clean distorted voltages and currents. It corrects point of common coupling (PCC) current harmonics and low power factor, while maintaining network and PV power balance via the shunt VSC’s feed-forward compensation. It also regulates the DC-link voltage, while the series converter maintains a pure sinusoidal load voltage despite grid sags, swells, or harmonics.
With increasing elderly populations and people with disabilities all over the world, there is an emerging need for accessible, autonomous home environments. This paper proposes a fully embedded, low-cost, and multimodal home automation system focused on solving interaction challenges for users with physical or sensory impairments. The proposed solution integrates eye-tracking, hand gesture recognition, and voice-based command inputs for a sustainably adaptable control towards diverse users. This system not only answers the needs of elderly and disabled people but also contributes to a global vision per the United Nations' Sustainable Development Goals. It was built using the Raspberry Pi 4 platform with open-source Python libraries, allowing the implementation of custom offline solutions with lower latency. Validation of the system was carried out by functionally deploying it at a caregiving facility, where it showed very high accuracy rates, from 90 to 96%, for the tested input modalities and robust performance under real-world conditions.
Reliable operation of modern power systems demands accurate forecasting. Errors can reduce resilience, especially in mission-critical military, defense, and strategic infrastructures. While hybrid forecasting models perform well under normal conditions, their robustness to noisy data still needs thorough evaluation. In this work, a sensitivity and robustness analysis of a novel adaptive hybrid forecasting framework is examined using two scenarios of correlated Gaussian noise. These scenarios were selected to represent measurement degradation, communication failures, and physical disturbances commonly encountered in resource-limited defense operational environments. The proposed hybrid method integrates the Multi Model Partitioning Filter, Nonlinear Autoregressive Exogenous models, and a Genetic Algorithm for Resource Allocation, with performance evaluated using real commercial data. The results highlight the successful performance of the proposed method and underline the importance of the sensitivity analysis as a validation process for forecasting methods intended for defense energy systems that need to remain highly reliable at all times.
Heart disease is a principal cause of illness and mortality worldwide. Early prediction and intervention are vital for active treatment and reducing the mortality rates. Smart healthcare systems offer a hopeful approach to attain this goal by applying machine learning (ML) and deep learning (DL) techniques for the prediction ofheart disease. The present study discovers the application of ML and DL algorithms to examine patient data and predict the heart disease risk. The Machine Learning and deep learning models with voting and stacking classifier, and XGBoost have been applied. We conducted a comprehensive analysis to identify the optimal machine learning algorithm within the considered models. It is observed that, with mean accuracies around 0.82-0.83 and significant low standard deviation, stacking classifier and LR are showing top performance.
This paper concerns an improved H∞ filter design for discrete-time linear singular uncertain systems with constant time delay and unknown input. The proposed approach yields a robust observer that reconstructs a state functional despite the presence of uncertainties that are functions of a bounded noisy signal. The uncertain quantities are considered as disturbances. H∞ filter is designed to attenuate the effect of parametric uncertainties on the estimation error and reconstructs a state functional independently of the actuator fault described by an unknown input vector. The proposed filter is based on the Krasovskii stability theory to ensure converging unbiased dynamics. The bias-free conditions are formulated into a convex optimization problem, leading to an optimal filter gain design. The results, validated on a benchmark numerical example and compared with existing methods, confirm the effectiveness of the proposed approach.
In recent years, many ships use powerful electric propulsion, which requires ensuring reliable and stable operation of the ship's power plant. One of the ways to ensure the reliability of the operation of the ship's electric propulsion is the use of multiphase drive electric motors. The article presents a non-iterative model of a ship's electrical system using a powerful electric motor drive with two stator windings. Using the created and modeled mathematical model, the operating modes of the ship's electrical power system are studied, and the results of simulations of the different operating modes are presented.
A position control of a two-degree-of-freedom planar robot modeled in bond graph is proposed. Bond graph is a methodology that allows direct modeling of mechatronic systems, so robots represented in bond graph are interesting case studies to apply the properties in this graphic work environment. In this way, a control called calculated torque to remove the nonlinearities characteristic of robot modeling is applied in bond graph. To know the shape of the controller, the mathematical model of the robot is necessary, so this is obtained through the junction structure then the controller is designed in bond graph. To know the effectiveness of the closed-loop system. Simulation results are shown where it is verified that the position of each link of the planar robot reaches the desired reference input.
