Effective path planning is vital for Autonomous Underwater Vehicles (AUVs) for performing diverse roles, including rescue efforts and logistics operations. In recent times, dynamic path planning has emerged as a prominent research area, enabling AUVs to navigate without the need for previous static information. Researchers are leveraging deep reinforcement learning (DRL), as another trending field of study, to tackle the challenges associated with dynamic path planning. In this study a dynamic sensing and collision avoidance approach is proposed to navigate in an entirely unknown environment with unstructured obstacles. Firstly, we employ Deep Q-Networks (DQN), a distinct domain within DRL, to address the challenge of path planning in a dynamic environment. Initially, established methodologies were implemented such as Double Deep Q-Networks (DDQN) and Dueling Double Deep Q-Networks (D3QN) to develop a model capable of navigating an AUV through environments containing both static and dynamic obstacles. Nevertheless, the previously mentioned techniques exhibit restricted generalization abilities in complex, high-dimensional state environments, frequently leading to inadequate performance in dynamic situations, thus making the more adaptable I-DQN method essential. Subsequently, Improved DQN (I-DQN) is introduced, an enhancement of the original DQN, aimed at further refining the DRL model's performance. Test within a real-time setting is carried out to evaluate the effectiveness of the DRL models against randomly generated starting and destination points. The results obtained indicate that the I-DQN approach outperforms DQN, DDQN, and D3QN regarding the path efficiency and the travel duration to reach the destination while avoiding collisions. The stability of the I-DQN algorithm is improved by Prioritized Experience Replay, which carefully selects significant transitions according to their temporal-difference errors, facilitating effective and reliable learning in intricate AUV navigation situations. The proposed strategy can be utilized to improve underwater search and rescue missions in inundated urban regions or areas affected by disasters. The AUVs can maneuver through intricate and unfamiliar underwater settings, skilfully dodging debris and obstructions, to swiftly and efficiently find survivors or evaluate structural damage.
The growing use of optimization models to help decision-making has created a demand for such tools that allow solving more real models of processes related to human activity in which hypotheses are not verified in a way specific for classical optimization. And just interval-valued optimization problems were developed for formulating such real-world problems which are usually not well defined and they contain uncertain data. In this paper, an interval-valued multitime control problem with first-order PDE constraints is considered. Then, the minimax exact penalty function method is used for solving the aforesaid interval-valued extremum problem. The most important property of all exact penalty function methods, that is, exactness of the penalization, is generalized to the case when one of such methods is used for solving an interval-valued multitime control problem. Namely, under appropriate invexity hypotheses, it is examined in the case of the minimax exact penalty function method which is used for solving the aforesaid interval-valued multitime control problem with the first-order PDE constraints.
In this research work, a new hyperchaotic two-scroll system with two unstable equilibrium points has been derived and the qualitative properties of the new hyperchaotic system are detailed. A bifurcation analysis has been carried out for the proposed hyperchaotic two-scroll system by varying the system parameters. Multistability property of the new hyperchaotic two-scroll system has been also illustrated by showing the coexistence of hyperchaotic attractors for the proposed hyperchaotic two-scroll system for the same set of parameter values but different initial conditions. For engineering applications, we provide an electronic circuit simulation of the proposed hyperchaotic two-scroll system using MultiSim 14.0. As a control application, we derive new results for the complete synchronization for a pair of new hyperchaotic two-scroll systems taken as the master and slave systems. We have used integral sliding mode control and Lyapunov stability theory for the derivation of the synchronizing control law for the complete synchronization design for a pair of new hyperchaotic two-scroll systems. We have provided MATLAB simulations to illustrate and highlight the main results of this work.
This study presents and evaluates three steering angle selection methods formulated for path following tasks. The methods determine the steering angle input through numerical integration of the vehicle equations of motion over a finite prediction horizon, divided into small time subintervals. The first approach introduces classical Nelder-Mead optimization algorithm when the steering angle is defined by minimizing a cost functional which is trajectory-tracking error. In the second and third cases the minimization of the cost functional is obtained by applying the following numerical methods: Newton iteration and bisection. The methods are applicable to vehicle dynamics models of varying complexity. Presented methods have been evaluated using planar (3 degrees of freedom — DoF) and spatial (10-DoF) vehicle models. Simulation results demonstrate that proposed approaches achieve trajectory-tracking accuracy comparable to those reported in literature. The bisection-based method provides balance between accuracy and computational efficiency, enabling reliable steering angle computation for trajectories with varying curvature and non-uniform speed profiles. A key advantage of the proposed numerical methods is their ability to incorporate nonlinear vehicle dynamics while significantly reducing the computational effort compared to conventional optimization-based approaches
This paper presents a novel optimization strategy for photovoltaic (PV) power systems based on a synergetic control (SC) theory integrated with state observation and adaptive estimation schemes. The main objective is to enhance system reliability and reduce implementation costs through the elimination of selected physical sensors. A state observer is designed to ensure stable system operation and avoid control-law singularities, while adaptive laws are employed to accurately estimate the load and the photovoltaic output voltage. These estimated states are provided to the synergetic controller, which drives the system to operate at the maximum power point (MPP). The proposed strategy is validated through numerical simulations under dynamic conditions, including variations in solar irradiance, temperature, and load. Furthermore, experimental tests corroborate the simulation results, confirming the robustness and high efficiency under partial shading conditions and demonstrating the strategy's practicality for real-world PV systems.
This paper presents a unified framework for linear sub-model selection based on the principle of entropy maximization. By maximizing entropy under suitable constraints, we derive generalized normal distributions that define both the error model and the prior on regression weights, linking distributional assumptions directly to regularization penalties. The resulting quasi-likelihood formulation integrates likelihood maximization with regularization and allows flexible control of robustness and sparsity through two shape parameters. We develop a convex optimization approach for parameter estimation and propose a heuristic binary optimization algorithm for efficient sub-model selection guided by Bayesian Information Criteria (BIC). The framework is experimentally evaluated on simulated regression tasks with both binary and continuous weights under varying noise conditions. The experiments systematically examine the influence of the distributional parameters and on the accuracy and stability of sub-model identification. The results show that for low and moderate noise levels, the proposed method reliably recovers the true sub-model and achieves information criterion values comparable to those of established regularization techniques such as ridge and lasso regression.
ime-delay systems, which are widely encountered in industrial processes, pose significant challenges to controller design due to their complex dynamics and their limited robustness margins. This paper presents introduces an explicit design methodology for a robust Smith predictor-based control structure employing a fully isolated dual-loop configuration. The proposed controller combines integer and fractional order dynamic operators to enhance robustness against plant uncertainties while maintaining favorable dynamic performance. The effectiveness of the proposed approach is demonstrated through simulation studies on various integrative process models, including comparative analysis with several recent control techniques. Furthermore, experimental validation on a liquid level control system representative of FOPTD industrial processes confirms the simulation findings, showing excellent reference tracking, effective disturbance rejection, and smooth control effort even under increased time delays. The results demonstrate the superior robustness and flexibility of the proposed design compared to existing methods, highlighting the practical benefits of the isolated dual-loop Smith predictor structure for real industrial applications.
This paper presents a new fractional-order power system stabilizer, PSS3B_FRAC, developed as a modification of the standard PSS3B stabilizer. Optimization-based procedures are applied to determine stabilizer parameters in single-machine and multi-machine systems. The results demonstrate that introducing fractional-order differentiation improves damping of electromechanical swings while maintaining acceptable voltage regulation. The study is conducted in three stages: (1) single-objective optimization in single-machine systems, (2) single-objective optimization in a 7-machine CIGRE system, and (3) poly-optimization in the same system.
This article addresses modeling systems using fractional order derivatives, highlighting three basic approaches: differential equations, operator methods, and state space representations. Each approach carries different advantages and limitations in the context of time invariant linear systems (LTI) of fractional order. The focus of the article is on Fractional Order Transfer Functions, which represent a special subject of interest because they offer practical utility with available simulation libraries (e.g., CRONE, FOMCON, NiNteger) and approximation techniques (e.g., Oustaloup, CFE, Thiele, Padé). This paper describes a transition between fractional order transfer functions (FOTF) and pseudo-rational representations of such systems. While the existing literature contains the basics of fractional differential equations and operator theory, it often omits explicit formulas for conversion between different representations. The paper fills a significant gap by proposing a novel algorithm for converting FOTF models to fractional LTI representations. Unlike previous works that implicitly assume such transitions exist, this paper provides formulas for coefficient transformations and demonstrates its effectiveness by minimizing the degree of the system.
In this paper, we provide some sufficient conditions for the exponential stability of solutions of nonlinear impulsive differential systems by using some inequality of Gronwall-Bellman type. Practical exponential stability is also investigated for a class of perturbed impulsive systems. Several numerical examples are provided to demonstrate the effectiveness of the theoretical results. Furthermore, Hopfield neural networks system is discussed as an application.
As cyberattacks become more advanced, advanced AI-based big data visualization is now needed for effective threat detection. Yet, choosing the best visualization tools is some multicriteria decision-making (MCDM) task that involves considering many criteria containing both positive and negative aspects. WhileMCDMmethods that address the selection and classification of AI-driven big data visualization tools focus on the positive aspects of the evaluation criteria and ignore the negative aspects of the criteria, resulting in incomplete evaluations. Further, although Einstein operators have shown strong results in uncertain and imprecise situations and MCDM approaches, they have not yet been used in bipolar fuzzy frameworks, which leaves a major gap in decision-making methods. To overcome these problems, this article interprets a bipolar fuzzy MCDM methodology based on Einstein prioritized operators to systematically evaluate and classify AI-driven big data visualization tools for cybersecurity threat detection. For this method, Einstein prioritized operators within a bipolar fuzzy framework devised in this article, which can aggregate both positive and negative aspects of the criteria.Acomprehensive case study is shown to assess and classify the prominent AI-driven big data visualization tools for cybersecurity threat detection, considering critical criteria with dual aspects. The proposed methodology is meticulously compared with the prevailing MCDM methods to validate its dominance in handling uncertainty and the bipolarity of the criteria. This article helps security professionals choose the right AI-powered visualization tools which, in turn improve the cybersecurity of their organizations and make it easier to detect threats.
The emergence of Game theory (GT) enabled with demand side management (DSM) has the applications in the field of smart grid applications. A mathematical method called game theory uses desirable rules to identify the circumstances under which all actors can win. Various agents can be used to optimize their gains. In terms of customer utility, demand response algorithms are categorized as agents in terms of customer utility. A centralized demand response (DR) scheduling algorithm that meets the varied energy consumption needs of a community can be difficult owing to the differences among residents. A non-cooperative DR-GT model is proposed to improve individual benefits in the energy consumption scheduling algorithm. The appliance information comprises different power levels to categorize the residents, which reduces the scheduling traffic between the residents and aggregator. There is a 23% reduction in the peak-to average ratio and increase in renewable energy usage by 13–25%, as better scheduling based on the flexibility of consumer loads and pricing schemes. Smart grid efficiency is improved by 23–30%, owing to reduced energy losses, fewer system imbalances, and lower wears on grid infrastructure.
The aim of this paper is to analyze the development of algorithms for Fast Matrix Multiplication (FMM) in both historical and technical contexts, as well as to compare available solutions on consumer-grade computer hardware. We review advancements in estimating the theoretical computational complexity of FMM and optimization techniques that are used in widely adopted algorithms, with a particular focus on optimal cache memory usage and leveraging Graphics Processing Units (GPU). The methodology of tests and their analysis highlight the performance differences of the considered algorithms depending on the matrix size and the nature of the data stored in them. Results indicate the significant role of tailoring the chosen algorithm to the available hardware and the specific application in which the algorithm is being performed. Also, we emphasize that the FMM algorithms can be applied not only to linear algebra problems but also to current problems in science and engineering, such as artificial intelligence, databases, parallel computations, computational biology, pattern recognition, and compiler construction, to mention just a few examples.
This work presents a reinforcement learning framework for controlling a planar threesection continuum robot in environments with static obstacles. Assuming constant curvature for each section, the robot is trained to navigate toward a fixed goal while avoiding collisions with multiple static objects. A custom simulation environment was developed to support three levels of scenario difficulty, easy, medium and hard, each with varying obstacle density and placement. The learning process is driven by the Deep Deterministic Policy Gradient (DDPG) algorithm, which enables smooth and continuous curvature control. Careful attention was paid to the design of the reward function and the network architecture, both of which were critical to achieving stable and reliable policy learning. Performance was evaluated across multiple runs, revealing that the agent successfully generalized its behavior across scenarios of increasing complexity. The proposed framework demonstrates the potential of reinforcement learning as a viable approach to safe and adaptive control in continuum robotic systems, with promising implications for applications such as medical navigation, search and rescue, and inspection in confined environments.
In this research work, we first obtain a new 3-D chaotic Lü system by modifying the dynamics of the classical Lü chaotic system (2002). Next, by introducing a state feedback to the new 3-D modified Lü chaotic system, we obtain a new 4-D hyperchaotic Lü system with a curve equilibrium.We carry out a detailed bifurcation analysis of the new4-D hyperchaotic system with a curve equilibrium and describe the bifurcation transition diagrams and Lyapunov exponents diagrams.We also derive new multistability results of the new 4-D hyperchaotic Lü system with a curve equilibrium. For engineering applications, we provide an electronic circuit simulation of the proposed hyperchaotic Lü system using MultiSim 14.0. As a control application, we derive new results for the complete synchronization for a pair of new hyperchaotic Lü systems taken as the master and slave systems. We have used integral sliding mode control for the derivation of the synchronizing control law for the complete synchronization design for the new hyperchaotic Lü system. MATLAB simulations are provided to illustrate the main results of this research work.
This paper presents a novel computational framework for assessing cardiovascular disease (CVD) risk by integrating unsupervised clustering techniques with survival analysis. The proposed method enables dynamic and individualized risk prediction by organizing patient data into structured clusters based on shared cardiovascular risk factors. The framework begins with competitive learning, an unsupervised clustering method, to group patients into clusters that reflect distinct risk profiles. Each cluster is represented by its centroid, calculated as the mean of the 9-dimensional feature vectors of its members, ensuring that the clusters effectively summarize patient data while preserving critical risk characteristics. For each cluster, an independent Cox Proportional Hazards Model is applied to analyze survival data, capturing the unique relationships between cardiovascular risk factors and survival outcomes within that cluster. A key innovation of this study is the introduction of the Cumulative Prevalence Ratio (CPR), a new metric that aggregates hazard rates over time separately for each cluster. This approach provides a comprehensive view of cumulative cardiovascular risk, enabling precise categorization of the patient into risk groups based on cumulative exposure to evolving risk factors. By integrating cluster-specific hazard functions and temporal risk metrics, the proposed framework improves the precision and adaptability of CVD risk predictions, paving the way for personalized and data-driven healthcare interventions.
The invariant properties of the stability, reachability, observability and transfer matrices of positive linear continuous-time systems with integer and fractional orders are investigated. It is shown that the stability, reachability, observability and transfer matrix of positive linear systems are invariant under their integer and fractional orders.
We prove that the Young measure associated with a Borel function f is a probability distribution of the random variable f(U), where U has a uniform distribution on the domain of f. As an auxiliary result, the fact that Young measures associated with simple functions are weak* dense in the set of Young measures associated with measurable functions is proved. Finally some examples of specific applications of the main result are presented with comments.
This paper investigates a class of multi-objective variational control problems, named (VCP), characterized by vector-valued integral objective/cost functional and differentiable inequality constraints. More precisely, we formulate the problem in a general framework and, in order to analyze optimality criteria, we associate scalar subproblems with the multi-objective setting and establish conditions under which efficient solutions of the original vector problem correspond to optimal solutions of these scalarized problems. We derive the necessary optimality conditions in the form of Euler–Lagrange-type equations, involving multipliers and piecewise smooth functions as variables. Furthermore, we introduce the new notion of (ϵ0, ϵ1) − (c0, c1) − type − I pair of functionals, adapted to the variational control setting. The introduced generalized convexity concepts are very important in characterizing efficient solutions and in deriving enhanced optimality conditions. These results enrich the theoretical framework of multi-objective variational control and provide a solid basis for future research and applications in mathematical optimization and control theory.
In this paper, a cybersecurity problem where the security team faces the challenge of preventing the hacker from breaching the multiple security layers is studied. The cybersecurity challenge mirrors the Chakravyuh, an ancient battlefield formation described in the Indian epic Mahabharata. The multi-layered complex structure of Chakravyuh, designed to trap the enemy, serves as an analogy for modelling the cybersecurity model of the digital economy. A hybrid model that integrates chaos theory and fuzzy logic is developed to enhance the defense mechanism in digitalization process. It is shown that the chaos theory approach can tackle the non-linear dynamics and unpredictable behavior of the hacker. On the other hand, fuzzy logic provides a more structured and adaptive defense mechanism. Monte Carlo simulations are used to analyze hacker’s breaching probability across security layers. Further, the Binary search algorithm is applied to optimize security layers dynamically while maintaining computational efficiency. This study integrates ancient wisdom with modern computational techniques to effectively mitigate cybersecurity threats.
