Oil palm plantations are highly vulnerable to outbreaks of caterpillars that damage leaf tissue, reduce photosynthetic performance, and ultimately decrease oil palm productivity. Based on this challenge, a predator-prey mathematical model is developed that describes the interaction between leaf-eating caterpillars and natural predators that prey on them to regulate their population. The model developed herein integrates multiple approaches for managing pests, including chemical pesticides, bio-pesticides, and manual mechanical control through the removal of pupae. Four equilibrium points are determined through dynamic analysis, and their existence and stability depend explicitly on several key biological and management parameters. This work investigates optimal control strategies for two ecological conditions: systems with natural predators and without them. For each condition, three control interventions are considered: chemical control alone, mechanical removal of pupae alone, and a combined approach. Cost-effectiveness analysis reveals that natural predators are an effective and inexpensive means of suppressing caterpillar populations. In the absence of biopesticides, the integrated approach of chemical treatment followed by manual removal of pupae is the most cost-effective among the methods that reduce pest pressure.
The rapid advancement and widespread adoption of Unmanned Aerial Vehicles (UAVs) have revolutionised industries such as logistics, surveillance, and disaster management, offering transformative operational capabilities. However, the growing reliance on UAVs has introduced significant cybersecurity challenges that threaten their operational integrity and reliability. Through a focused systematic review this study identifies and analyses these challenges, highlighting critical vulnerabilities associated with communication systems, data integrity, physical and remote threats, software weaknesses, and regulatory compliance gaps. Based in the classic resilience concepts of redundancy, flexibility, modularity, diversity, and tightness of feedback loops, and adapted in this study to a UAV context, insights are given into how such vulnerabilities may be countered, thereby improving the potential of UAVs in dealing with unforeseen occurrences. The findings demonstrate the urgency of developing a robust cyber-resilience framework that integrates advanced security measures, standardised regulations, and collaborative efforts among stakeholders to address both current and emerging threats. By linking key cybersecurity challenges to core resilience attributes, this review establishes an essential foundation for strengthening the design and engineering of UAVs and subsequently their secure deployment across various domains.
This paper proposes a compensator design methodology based on the Relative Gain Array (RGA) for systematically modifying the steady-state interaction structure of linear multivariable systems. Unlike the conventional use of the RGA as an analysis tool for input–output pairing selection, the proposed approach exploits its structural properties to reshape the interaction pattern toward a prescribed target RGA while preserving the plant dynamics. The method is formulated from the steady-state gain matrix and provides explicit conditions for obtaining feasible interaction configurations in systems with nonsingular steady-state gains matrices. The proposed framework is validated through simulation and experimental studies on a two-degree-of-freedom laboratory helicopter. Different interaction structures are implemented and evaluated under proportional and LQR control schemes. Performance is assessed using tracking accuracy, settling time, overshoot, control effort, and cross-coupling indicators. The results demonstrate that the proposed compensator effectively modifies the interaction structure and produces predictable changes in closed-loop behavior. Furthermore, the experimental analysis reveals practical trade-offs between interaction reduction, control performance, and actuator requirements. These findings show that the proposed RGA-based methodology provides a systematic and experimentally validated framework for interaction management in multivariable control systems.
Recent work in computer vision and deep learning has improved pavement damage detection and classification, but many models still work poorly in complex pavement scenes where damage areas are small, defect shapes are irregular, features appear at different scales, and background noise affects recognition. To solve these problems, this study proposes a Noise Reduction and Attention Enhanced Transformer (NRAET) for pavement damage classification. The NRAET is based on a Swin Transformer and has two main parts. The first is the Spatial Noise Reduction (SNR) module, which removes unnecessary information and reduces background noise. The second part is the Multi-Attention Enhancement Block (MAEB), which combines window, multi-scale, and fine attention to help the model better detect irregular cracks at different scales and learn global features. We conducted several experiments on several public benchmark datasets with different pavement types and damage patterns. For the CQU-BPMDD dataset, NRAET performed better than the baseline model. Compared with the baseline Swin Transformer, NRAET achieved improvements of 1.64% in Accuracy, 1.20% in Precision, 1.47% in F1-score, and 3.23% in AUC on the CQU-BPMDD dataset. The results show that NRAET under different data distributions has a good effect, very strong robustness, and generalization ability.
Accurate simulation of electromechanical systems requires the coherent integration of electrical, magnetic, and mechanical dynamics within a digital environment. However, conventional integrators often lose fidelity when the model enters stiff regions or undergoes abrupt transients. These limitations generate numerical distortions and unstable responses that compromise the realistic representation of the system’s dynamic behavior. This work examines the theta method as a dynamic numerical regulator rather than solely as a fixed-parameter integrator, formulating it as an adjustable scheme applied to a nonlinear electromechanical model with external algebraic coupling, whose formulation naturally leads to a differential–algebraic representation. This framework enables the evaluation of the integrator’s performance under conditions of strong coupling and structural constraints inherent to stiff systems. Through the modulation of a single parameter of the theta method, the proposed scheme modifies its balance between explicit and implicit formulations and regulates its numerical dissipation in order to improve the stability of the integration process without reducing the time step. Compared with the explicit Euler, implicit Euler, symplectic Euler schemes, and the Runge–Kutta integrator, the theta method provides greater dynamic coherence both under steady-state operating conditions and during closed-loop transients. As a result, it is presented as a versatile and robust alternative for digital simulation and control-oriented design of modern electromechanical systems with stiff behavior.
This article is concerned with studying the solvability and optimal feedback control for stochastic switched non-autonomous differential inclusion with deviated arguments and fractional Brownian motion in Hilbert spaces. Initially, the solvability results are established using fractional calculus, stochastic analysis, resolvent operators and the Bohnenblust-Karlin fixed-point theorem without imposing Lipschitz conditions. The proposed framework uses switching signals to model transitions among a finite set of subsystems and time-varying structural changes in the dynamics, where each signal uniquely determines the active subsystem at any given time. Then, the Filippov theorem and the Cesari property are employed to demonstrate the existence of a feasible pair. Additionally, an existence result for optimal pairs of the Lagrange problem is established. Furthermore, an example is given to support the theoretical results.
Smoking remains a major global health concern and a leading cause of chronic illness. This study develops a fractional six-compartment smoking model using Caputo derivatives, incorporating hospitalized smokers. Theoretical analysis establishes key model properties, including positivity, boundedness, and the existence and uniqueness of solutions. Equilibrium points are determined, and their stability is examined through the Jacobian matrix and Lyapunov functions. The model is calibrated using India’s smoking data, and the fractional-order formulation provides a better fit than the integer-order model. Sensitivity analysis using the Normalized Forward Sensitivity Index and Partial Rank Correlation Coefficients (PRCC) identifies minimizing transmission and hospitalization transitions, while improving recovery and quit rates, can effectively reduce smoking prevalence. Computationally, the spectral collocation method with Chebyshev polynomials is applied, and accuracy is validated via residual errors and exponential convergence. Parameter effects and their interplay are analyzed through figures and contour plots. An optimal control framework incorporating public awareness and nicotine replacement therapy demonstrates effective reduction in smoking prevalence, with nicotine replacement therapy alone identified as the most cost-efficient strategy for smoking cessation.
Touchless fingerprint recognition provides hygienic and user-friendly biometric acquisition; however, image quality is often degraded by non-uniform illumination, motion blur, low ridge-to-valley contrast, and background interference. These factors disrupt ridge continuity and minutiae extraction, leading to poor recognition accuracy. To address these challenges, this paper proposes a novel Neutrosophic Set Fractional Sobel Enhancer ( NS-FSE ) for touchless fingerprint images. The proposed framework transforms fingerprint images into the neutrosophic domain using three structurally independent components: truth, indeterminacy, and falsity, derived respectively from pixel intensity, local entropy and variance, and gradient magnitude. These components model ridge information, ambiguity, and background uncertainty independently without mutual constraints. An adaptive indeterminacy reduction mechanism suppresses ambiguous regions while preserving ridge structures. Subsequently, a fractional-order Sobel operator enhances ridge edges while minimizing noise amplification. The enhanced image is reconstructed using a nonlinear defuzzification process. Experiments were conducted on 2,976 touchless fingerprint images from two independent acquisition sessions of the PolyU database. Recognition performance was evaluated using a unified cross-session dataset containing 160 genuine and 25,440 impostor pairs with the SourceAFIS matcher. The proposed method was compared with seven enhancement techniques using conventional image quality and fingerprint-specific structural metrics. Results show significant improvements in ridge clarity, minutiae quality, ridge continuity, and overall clarity score. The proposed approach achieved an Equal Error Rate of 7.3% compared to 18.2% for unenhanced images. Ablation studies and paired t-tests (p < 0.001) further confirm the effectiveness and statistical significance of the proposed framework.
This is an expository paper that discusses an approach to the linear quadratic Gaussian/loop transfer recovery (LQG/LTR) design problem for finite-dimensional single-variable (single-input/single-output, SISO) control systems. The approach is based on the utilization of weighting augmentation for incorporating design specifications into the framework of the LTR technique for LQG compensator design. The LQG compensator is to simultaneously meet given analytical low- and high-frequency design specifications expressed in terms of desirable sensitivity and controller noise sensitivity functions. The paper is aimed at non-specialists and, in particular, practitioners in finite-dimensional LQG theory interested in the design of feedback compensators for closed-loop performance and robustness shaping of SISO control systems in realistic situations. The proposed approach is illustrated by a detailed design example: the torque control of a geared DC motor with an elastically mounted output shaft.
This paper presents the development and stability analysis of a block hybrid collocation method for the direct numerical solution of third-order quasi-linear ordinary differential equations subject to standard initial conditions. The proposed method is formulated without reducing the original problem to an equivalent system of first-order equations, thereby avoiding the associated increase in computational cost and implementation complexity. A continuous hybrid scheme is constructed using three interpolation points and four off-step collocation points, allowing the simultaneous approximation of the solution at multiple grid points within a block. Collocation conditions are imposed on the third derivative of the approximating polynomial to derive a unified block formulation. Theoretical properties of the method, including consistency, order of accuracy, zero-stability, and convergence, are rigorously analyzed. Numerical experiments involving several benchmark problems with known exact solutions are conducted to assess the method’s performance. The results demonstrate that higher accuracy is achieved with significantly fewer computational steps than those required by existing methods. These findings confirm that the block hybrid collocation method provides an efficient, stable, and reliable numerical framework for solving third-order quasi-linear ordinary differential equations.
Autonomous vehicle technology demands advanced lane detection (LD) techniques adaptable to complex road scenarios. Traditional LD systems relying on static inputs often fail in real time. The proposed research introduces a deep learning model integrated with a sensor fusion strategy in an IoT- based vehicular perception framework to address these limitations. The proposed end-to-end uncertainty estimation with continual learning network (E2E-UCNet) utilizes data collected through cameras from diverse real-time environments in Andhra Pradesh and Tamil Nadu, India. In the absence of physical range sensors, depth cues similar to LiDAR and Radar data are generated from monocular RGB images using MiDaS monocular depth estimation and incorporated into the E2E-UCNet model to improve feature representation for dynamic lane prediction. The Multi-Sensor Fusion with Uncertainty Estimation (MSF-UE) enhances E2E-DLN reliability through Bayesian deep learning. Further enhanced by the continual learning semantic segmentation network (CL-SSN), the framework achieves 99% pixel-wise accuracy, 0.51 IoU, and a Dice coefficient of 0.67 under varying road and weather conditions. The dynamic continual learning for long-term adaption (DyCLA) framework progressively updates the model with new data, ensuring consistent lane detection under changing traffic conditions. These findings aim to improve self-driving safety and reliability for autonomous vehicle applications.
This article addresses the chattering-free finite-time control problem of a planar two-link flexible manipulator in the presence of uncertainties and external disturbances. First, a barrier-function-based adaptive second-order non-singular terminal sliding mode controller is developed. Unlike conventional approaches, the proposed method simultaneously ensures tip-trajectory tracking and tip-deflection suppression without requiring prior knowledge of disturbance bounds. Second, a higher-order sliding framework is established by combining a tracking-error-based first-order sliding surface with a second-order non-singular terminal sliding manifold, thereby guaranteeing singularity-free finite-time convergence. Then, adaptive barrier functions are employed to restrict the tracking errors within a prescribed neighborhood and to prevent controller gain overestimation. Furthermore, modal analysis is incorporated to identify the natural frequencies and modal characteristics of the flexible manipulator under no-payload, nominal-payload, and maximum-payload conditions. Lyapunov stability analysis confirms the closed-loop stability and error convergence. Finally, comparative results verify that the proposed controller achieves superior tracking performance, smoother control action, reduced control effort, and enhanced vibration suppression compared with state-of-the-art methods.
Clothing style recognition is challenging due to large variations in global appearance, local design, and fine-grained textures, and existing studies often rely on relatively homogeneous datasets that limit generalization in practical applications. To address these issues, we introduce ShanghaiFashionStyle17, an expert-curated clothing style dataset designed to reflect modern fashion diversity, comprising 17 style categories and 33,184 images. We further propose Fashion-YOLO, an enhanced YOLOv8-based model for clothing style recognition. Specifically, Fashion-YOLO integrates the SENetV2 channel attention mechanism to strengthen global and local detail feature extraction, employs Programmable Gradient Information (PGI) to optimize gradient flow and improve the learning of fine-grained style cues, and incorporates depthwise separable convolutions (DWConv) to reduce parameter count while maintaining effectiveness. Experiments on ShanghaiFashionStyle17 and FashionStyle14 show that Fashion-YOLO achieves 81.7% and 78.7% Top-1 accuracy, respectively, outperforming the baseline by 4.5% and 3.8%. In addition, we report practical runtime indicators under a unified setting, conduct error analysis using a normalized confusion matrix and representative failure cases, and perform cross-dataset transfer evaluation on semantically aligned overlapping categories between the two datasets to assess generalization. Overall, Fashion-YOLO demonstrates a favourable accuracy-efficiency trade-off and ShanghaiFashionStyle17 supports learning transferable style representations.
Deceptively realistic deep synthetic images are widely disseminated across the internet, raising significant concerns about misinformation and content authenticity. From the perspective of responsible model publishers and credible regulators, this study introduces a novel robust generative image watermarking method that significantly advances detection and tracing of deep synthetic images. Unlike existing approaches that focus primarily on watermark embedding, our method uniquely combines adversarial training with specially designed cropping and resizing manipulations to substantially enhance watermark robustness. Furthermore, we introduce a residual signal prediction mechanism through the encoder, coupled with a discriminator-based quality enhancement approach, which significantly improves the quality of generated images while maintaining watermark integrity. With minimal impact on the original performance of generative models, this approach embeds robust generative image watermarks that demonstrate superior resilience against common disturbances such as compression, cropping, and resizing encountered on social media platforms, achieving over 75% bit-wise accuracy even under extreme distortions, thereby facilitating effective tracing and detection of synthetic images at the model level.
This paper reviews the application of adaptive dynamic programming (ADP) in controlling constrained nonlinear systems, with an emphasis on integrating ADP to address optimization and constraint issues in complex systems. First, the conventional ADP algorithm for tackling the optimal control problems of unconstrained nonlinear systems is introduced. Next, the general solutions and recent advances of ADP for controlling nonlinear systems with various constraints, mainly including input constraints, state constraints, output constraints and cost constraints, are elaborated. Moreover, several typical real control applications for constrained nonlinear systems with respect to ADP are summarized, particularly in the fields of aerospace systems, robots, autonomous systems and energy systems. Finally, some possible future prospects are explored, and the conclusions of this paper are presented. Overall, ADP plays a significant role in guaranteeing system safety and optimizing performance with broad application prospects, and the comprehensive investigation demonstrates the tremendous potential of ADP for controlling constrained nonlinear systems in the current eras of artificial intelligence, control science and engineering, and systems science.
Multi-objective Reinforcement Learning (MORL) generalizes traditional reinforcement learning to scenarios involving multiple conflicting objectives that require explicit trade-off optimization. This survey provides a comprehensive synthesis of the theoretical foundations, algorithmic frameworks, evaluation methodologies, and practical applications of MORL. It begins by formalizing the multi-objective Markov decision process and reviewing fundamental optimality notions. Then, the MORL algorithms are categorized along three dimensions: model availability, policy scope, and preference representation, encompassing model-based planning, model-free learning, and interactive preference elicitation methods. Furthermore, it examines widely adopted evaluation metrics and benchmarks, offering guidance for empirical assessment and comparison. Finally, applications and emerging challenges are discussed, highlighting open issues related to scalability, safety, and human alignment. Overall, a structured and up-to-date overview of MORL research is provided, aiming to bridge theoretical insights with practical implementations across diverse decision-making domains.
Hunger Games Search (HGS) is a novel metaheuristic algorithm favoured for its exploration ability and flexibility. Following the PRISMA 2020 guidelines, this study provides a systematic review of 115 articles published from 2022 to 2025. The review is structured around a multi-dimensional taxonomy of HGS variants and a comprehensive survey of seven major application domains, identifying engineering, machine learning, and medical optimization as the most prevalent areas. To enrich the depth of the review, this research incorporates a new quantitative performance evaluation using the CEC 2017 benchmark suite, where HGS is compared against nine state-of-the-art algorithms. The results demonstrate that HGS possesses robust global exploration capabilities, particularly when addressing complex composition landscapes, while also identifying specific areas for potential improvement in exploitation precision. Crucially, this study summarizes major improvement methods and establishes a prescriptive mapping that links inherent algorithmic limitations directly to specialized strategic enhancements such as chaotic mechanisim, opposition-based learning, and structural hybridizations. The survey findings confirm that HGS exhibits remarkable competitiveness and adaptability in solving diverse real-world optimization tasks. Despite its promising performance, HGS research remains relatively nascent. This review serves as a comprehensive resource for researchers by outlining potential research directions and providing insights into the properties of HGS variants and their suitability for various application domains.
This paper introduces a novel H-infinity robust control framework for nonlinear fractional-order systems (FOS) affected by external disturbances, input saturation, and time-varying uncertainties. A dynamic output feedback (DOF) controller is designed to ensure robust asymptotic stability while maintaining high performance under input constraints. The method utilizes Caputo fractional derivatives and advanced time-domain analysis to derive new theorems for stability verification and controller synthesis. By leveraging Linear Matrix Inequalities (LMI), the approach remains computationally efficient while guaranteeing stability and effective disturbance rejection. Key contributions include the characterization of the region of attraction (ROA) and the stable region (B-& varepsilon;), extending classical robust control results to fractional-order nonlinear and chaotic systems. Simulation results validate the proposed strategy on chaotic supply chain dynamics, demonstrating enhanced robustness, resilience, and efficiency compared to existing techniques. This study advances the theoretical foundation of nonlinear FOS control and offers a practical, powerful framework for addressing modern control challenges.
This paper proposes a novel population initialization algorithm, termed Memory and Prediction-based Population Initialization (MPPI), for addressing dynamic multi-objective optimization problems (DMOPs). MPPI monitors environmental vectors to detect changes in the environment. Upon detecting a change, the algorithm employs three mechanisms to initialize the population: First, it establishes a memory bank based on center points, knee points, and boundary points associated with the environmental vector, and generates a portion of the population based on the memory bank. Second, it predicts a subset of solutions based on feedforward center points. Finally, it generates a number of random solutions to maintain population diversity. Since the memory-based initialization mechanism does not rely on recent solutions, the proposed algorithm is capable of solving dynamic multi-objective optimization problems with dramatically and irregularly changing Pareto optimal sets (POS). The MPPI algorithm is validated on a suite of benchmark functions and a random DTLZ switching experiment. We compare the experimental results against those of other state-of-the-art population initialization methods. The comparison demonstrates that MPPI can quickly track the POS in response to environmental changes, exhibiting strong performance on both periodic and non-periodic problems.
