This paper addresses the quasi-consensus problem of fractional-order multi-agent systems (FOMASs) governed by partial differential equations under sensing and communication constraints. Two observer-based dynamic event-triggered boundary control strategies are proposed to estimate and regulate the system states in scenarios with partial spatial sensing and boundary sensing, respectively. Both a partial domain observer and a boundary observer of the Luenberger type are constructed to reconstruct the system dynamics under limited measurements. To reduce communication burden while maintaining control performance, dynamic event-triggered controllers with adaptive thresholds are developed. Rigorous Lyapunov-based analysis shows that the proposed strategies guarantee the achievement of quasi-consensus. Moreover, the existence of a strictly positive lower bound for the inter-event times is formally established, thereby excluding Zeno behavior. Numerical simulations are provided to verify the effectiveness of the proposed methods.

This paper addresses the critical challenge of distributed fault-tolerant projective group consensus control for multi-agent systems operating under both the input constraints and persistent external disturbances. Unlike conventional approaches, this study proposes a novel framework that provides comprehensive solution for multi-fault scenarios by modeling diverse fault dynamics through time-varying polynomial formulations and polyhedral structures. This advanced representation accommodates a wide spectrum of fault modes without requiring strict matching conditions, thereby significantly enhancing both flexibility and practical applicability. To achieve accurate estimation of system faults and disturbances, we develop a distributed composite observer. This observer utilizes relative output error signals between neighboring agents while effectively decoupling the interactions between faults and disturbances. Furthermore, we introduce an innovative distributed anti-saturation fault-tolerant control strategy that synergistically integrates the observer's estimates with global output information. This approach guarantees robust performance despite simultaneous input constraints and multiple fault conditions. Through simulation studies, the proposed control framework demonstrates its effectiveness, confirming its potential for real-world implementation in complex multi-agent systems.

This paper studies an optimal innovation-based stealthy attack strategy for networked control systems (NCSs) with asymmetric information. Different from previous literature on NCSs with a single controller, in the considered NCSs consisting of two controllers where Controller 1 merely uses its own observations to design strategies and Controller 2 uses not only its own observations but also the observations of Controller 1 to perform decisions which brings forth the asymmetric information. Meanwhile, the packet dropout happens in the channel of Controller 1 and the attack occurs in the communication channel of Controller 2 where the innovation signal is tampered. In order to avoid the detector, the Kullback-Leibler divergence (KLD) is employed as a stealthy measure, enabling the attacker to execute strict or relaxed stealthy attacks. The object is to make the quadratic control cost maximized and the attack cost minimized which leads to a non-convex optimization problem. In virtue of the singular value decomposition, the optimal strict stealthy attack strategy is derived. Moreover, to gain a higher-attack cost, the optimal relaxed stealthy attack strategy is presented by transforming the non-convex optimization problem into the convex optimization problem. Numerical examples are given to show the effectiveness of the proposed algorithms.

Walking direction recognition is vital for HCI, healthcare monitoring, and navigation, but real-world data are highly imbalanced: non-straight trajectories far outnumber straight ones, biasing classifiers and reducing accuracy. We collect synchronized inertial and plantar-pressure signals from wearables and deliberately construct an imbalanced dataset to mirror practice. We then propose a multimodal framework that couples CNN-based spatiotemporal feature extractors with a DRL Q-network; a reward-driven optimization promotes balanced decisions under skewed label distributions. Experiments on the constructed dataset show consistent gains over state-of-the-art methods, with the largest improvements under severe imbalance, demonstrating robustness and suitability for deployment in wearable gait analysis systems.

This study presents an event-triggered stealthy attack scheme for networked control systems (NCSs) operating with both remote and local controllers. For NCSs without attacks, the optimal estimator is obtained in terms of asymmetric information, and the corresponding optimal control laws are characterized via Pontryagin’s maximum principle. By developing a new iterative procedure, the mutual dependence between the controller and estimator designs is effectively removed. Building upon the derived optimal control policies and under constrained attack resources, an event-based stealthy attack strategy is developed to intermittently compromise NCSs while averting the attack detector. Meanwhile, a performance index is constructed that accounts for control cost as well as attack cost. For the attacker, the goal is to maximize the quadratic control-related cost while minimizing attack expenditure. By leveraging matrix decomposition techniques, the optimal event-based strictly stealthy attack is derived. Simulations assess the proposed control scheme and the developed attack strategy.

A novel fault-tolerant control (FTC) scheme for linear systems (LS) under polynomial faults is presented in this paper, utilizing the control allocation (CA) technique. The scheme is applicable to control systems where the input matrix cannot be decomposed into two matrices with equal ranks and smaller sizes compared to the minimum number of columns and rows of the input matrix. The faults induced by the system are expressed as polynomial functions of time and are considered within the scheme. System state and fault are estimated using an observer, and the obtained virtual control signals are reassigned to fault-diagnosing controller. Numerical examples are provided to validate the results.

A cart-pendulum is a typical underactuated mechanical system with two degrees of freedom (DOF) and one control input. This paper studies the stabilisation control problem for this system. Based on the idea of energy absorption, a novel virtual friction control law is firstly designed. And the characteristics of closed-loop control system are analysed. And then, a reference motion trajectory for the cart-pendulum system is constructed. Afterwards, a fixed-time tracking controller is designed to make the system quickly track the reference trajectory. Finally, the simulation results verify the effectiveness of the presented control strategy.

This study focuses on the dynamics of rumor cross-layer propagation on multi-layer coupled networks, which is different from traditional research on rumor propagation on multiplex networks. It not only considers the heterogeneity characteristics of the population, but also classifies information, and further ponders the influence of stochastic interference on rumor propagation. A more general stochastic rumor propagation model on multi-layer coupled network incorporating video and text information is established. Firstly, the existence of the unique global positive solution for the model is manifested. Subsequently, the conditions for the rumor disappearance and persistence on multi-layer coupled networks are explored via It & ocirc;'s formula and the strong law of large numbers. Finally, numerical simulations are performed to verify the reliability of the theoretical results, and a series of targeted rumor control strategies are proposed in combination with sensitivity analysis. The research results indicate that the individual social network as the middle layer plays a crucial role in the three-layer coupled network structure. Specifically, the control measures implemented for this intermediate layer can not only effectively suppress the spread of rumors within this layer, but also have a significant control effect on the spread of rumors throughout the entire network system. This discovery provides important theoretical basis and practical guidance for rumor control in multi-layer network environments.

This paper addresses the fault-tolerant control problem for polynomial fault systems under actuator saturation. The fault is modeled as a polynomial function, while the unknown disturbance is generated by a known external system. An observer is used to simultaneously estimate the disturbance, fault, and system state. Furthermore, since saturation in real-world systems is not solely due to physical limitations of components, an anti-saturation compensator is introduced to compensate for the input and to address the coupling between the fault and saturation. An gain is used to limit the energy transfer from the saturation error to the system, ultimately achieving anti-saturation fault-tolerant control. Finally, two practical examples validate the effectiveness of the proposed method.

The bridge crane system has been widely used in industrial production. This mechanical system is a typical underactuated system that has one control input and two degrees of freedom (DOF). This paper discusses the stabilizing control problem for this 2-DOF underactuated mechanical system. A fixed-time hierarchical sliding mode control method is developed. The method first constructs two first-level sliding surfaces and a second-level sliding surface for the bridge crane system. Then, a sliding mode controller is designed to stabilize the second-level sliding surface at the origin within a fixed time. It guarantees the fixed-time convergence of two first-level sliding surfaces at the origin. This further enables the fixed-time stabilization of the bridge crane system to be achieved. Compared with existing methods, our presented control method can stabilize the bridge crane system at the desired position within a fixed time. As a result, the system's stabilizing time is independent of its initial conditions. In addition, it can be predicted in advance and can be limited by a fixed constant. Numerical simulation experiment results verify the effectiveness of the presented method.

Apples of different production origins vary in quality due to differences in natural growing conditions. The practice of passing off low-quality apples as high-quality varieties is a widespread issue in the market. To address this issue, a bionic gustatory-emotion coupling model and Hebbian-habituation learning rule (GECM-HHLR), which mimics the nerve conduction and synaptic plasticity mechanisms of the human taste and emotional systems (HTESs), is proposed and combined with an electronic tongue (e-tongue) for apple origin recognition. First, GECM-HHLR describes the transmission pathways and learning patterns of HTESs information and improves the 1/f characteristics of brain responses under the stimulation of e-tongue data, demonstrating the bionic performance of GECM-HHLR in processing e-tongue data. Second, GECM-HHLR achieves superior results in the classification of apple origins in comparison with mainstream baseline models of the e-tongue system, confirming the adaptability of GECM-HHLR for apple origin analysis. Finally, the fusion of GECM-HHLR and convolutional neural network (CNN) yields the optimal performance with an accuracy of 98.75%, a kappa coefficient of 98.57%, an F- 1 -score of 98.81%, and a precision of 98.64% in apple origin identification, outperforming multiple state-of-the-art algorithms. In conclusion, effective classification of apple origins is achieved via GECM-HHLR and CNN, providing a biological mechanism-based approach for sensor data processing in food detection.

This paper intends to address the outlier-resistant fusion filtering problem for a class of nonlinear two-dimensional systems with the energy harvesting sensors under the measurement censoring scheme. The energy harvesting technology is considered to provide the energy required for measurement signal transmission, where the energy harvested at each shifting point is characterized by a random variable, and the signal can be successfully transmitted only when the sensor accumulates sufficient energy. The censored measurements are modeled by the well-known Tobit model and further characterized by a series of Bernoulli random variables with local probability approximation. Additionally, a saturation function is introduced to mitigate the impact of abnormal innovations caused by measurement outliers. The design of the fused filter in this paper follows two main steps: firstly, the local filters are established in the presence of incomplete measurements that are caused by energy constraints and censoring, where upper bounds for the filtering error variances are derived and subsequently minimized at each shifting point by selecting appropriate parameters; and secondly, the obtained local estimates are fused by deploying a suitable fusion scheme, where the fused estimate is shown to be more accurate than the local ones. Finally, a numerical example is provided to verify the effectiveness of the proposed outlier-resistant fusion filtering algorithm.

This paper proposes an arbitrary preassigned-time sliding mode control (SMC) scheme to achieve synchronization in six-dimensional memristive cellular neural networks (6D-MCNNs). First, a novel 6D-MCNNs model is established by incorporating three flux-controlled memristors. The chaotic characteristics of the system are confirmed through Lyapunov exponent spectra and bifurcation analysis. Then, a new SMC strategy integrated with a disturbance observer is developed by introducing an effective time-varying function. A new lemma is also presented to derive the conditions for arbitrary preassigned-time synchronization (PTS), which guarantees that the error system reaches the sliding surface and maintains sliding motion within a preset time. Notably, the settling times for both phases are independent of initial conditions and other design parameters. Finally, an image encryption algorithm is designed to demonstrate the application of PTS in 6D-MCNNs. Numerical simulations and statistical evaluations are provided to validate the theoretical findings and illustrate the potential of the proposed method in enhancing communication security.

This paper investigates the leader–follower consensus problem for multi-agent systems (MASs) affected at both the network layer and the physical layer. At the network layer, denial-of-service (DoS) attacks may block information channels and render some agents completely disconnected from the leader for a certain period of time. At the physical layer, the actuators of the agents may suffer from periodic intermittent faults with abrupt jumps and may also be affected by external disturbances. To model and compensate for such periodic faults, a Fourier-series-based observer is proposed, where a rotating exosystem is introduced to transform the estimation of unknown Fourier coefficients into an initial-state estimation problem. Based on the observer, a corresponding fault-tolerant control (FTC) scheme is constructed. Furthermore, to address DoS attacks, which are common in practical networked systems but difficult to characterize accurately, an active defense strategy based on switching among backup communication topologies is developed. With this strategy, even if some agents are temporarily disconnected from the leader during an attack, MASs can restore communication through topology switching and maintain consensus without relying on the commonly used a priori constraints on attack frequency and duration. Finally, simulation studies on a formation of five aircraft demonstrate that the proposed approach can achieve stable and reliable leader–follower consensus under multiple practical adverse factors, thereby verifying its feasibility and practical value.

ABSTRACT For stochastic nonlinear high‐order systems with unmodeled dynamics together with time‐varying both state and input delays, this paper investigates the fixed‐time tracking control problem for the first time. First, an adaptive neural network‐based controller is designed by integrating the neural network method with adaptive backstepping to solve algebraic loop issues under non‐strict feedback. An improved Lyapunov–Krasovskii function is constructed to compensate for time‐varying state delays, while dynamic and compensation signals are used to handle unmodeled dynamics and input delays, respectively. Then, based on the semi‐global practical fixed‐time stability theory, the boundedness of all signals in the closed‐loop system is proven, demonstrating that the convergence time is independent on system initial values. Finally, simulation examples are given and demonstrate the effectiveness of the devised control scheme.

Emotional modeling allows machines to become more aligned with human perceptual patterns. In light of this, a neural conduction mechanism–based computation model of emotional cortex, referred to as the Amygdala-Based Bionic Model (ABBM), is proposed to achieve a mathematical representation of emotional cognition. First, the neural dynamics neurodynamics modeling method of the K-model is adopted for the construction of ABBM. Second, the dynamical characteristics of emotional information transmission are characterized through ABBM, thereby establishing a mathematical expression for emotional cognition. Finally, when subjected to external stimuli, ABBM reproduces significant fast-response and 1/f characteristics, exhibiting bionic performance, which validates the effectiveness of the ABBM modeling. Cognition of emotion system is achieved using the ABBM. The significance of ABBM modeling lies in its provision of a mathematical tool for analyzing the mechanisms of the human emotional system, offering a theoretical foundation and technical direction for the development of highly bionic machine perception systems.

This paper presents an observer design method based on Fourier decomposition and rotational exosystem reconstruction to address periodic intermittent faults that are challenging to handle in existing studies. The research focuses on the fault-tolerant control (FTC) of multi-agent systems (MAS) subject to physical-layer periodic intermittent faults and network-layer Denial-of-Service (DoS) attacks. First, considering the prevalence of periodic faults and the non-differentiability of certain fault signals at switching instants due to instantaneous level transitions, a Fourier observer based on a rotational exosystem model is designed to estimate actuator faults and system states. Second, a FTC protocol integrating an event-triggered mechanism and an open-loop estimator is developed to mitigate the impact of DoS attacks, with further analysis conducted on the attack frequency and duration. Finally, the feasibility of the proposed observer and FTC scheme is validated through a simulation case study.

In this article, a fixed-time prescribed performance control (PPC) scheme is proposed for vehicular platoons with uncertain dynamics and actuator faults, dead zones, and saturation. A novel sigmoid function is employed to smoothly approximate asymmetric actuator saturation and dead zones, and a unified framework is constructed that can simultaneously account for actuator faults. To handle model uncertainties and external disturbances, an observer is designed to estimate disturbances within a prescribed time. The proposed scheme uses adaptive sliding-mode control to ensure that tracking errors converge to prescribed steady-state regions within a fixed time to achieve the fixed-time individual vehicle stability and string stability without considering initial errors between vehicles. Simulation experiments are performed to verify the effectiveness of the proposed scheme.

Whereas model-free, data-driven inference of chaos in dissipative systems has been extensively studied, little attention has been paid to the inference of chaos in conservative systems. Here, leveraging the technique of parameter-aware reservoir computing (RC) in machine learning, we present a model-free approach to the inference of measure synchronization (MS) in coupled bosonic Josephson junctions. We demonstrate that, using time series of several system states generated at different coupling strengths, a machine can be trained which, guided by the coupling coefficient, is able to infer not only the critical couplings at which MS emerges, but also the variation of the system order parameters during the transition from desynchronization to synchronization. The results highlight the potential of the parameter-aware RC technique in inferring the dynamics of high-dimensional Hamiltonian systems, and lay the groundwork for model-free analysis of the collective dynamics of quantum many-body systems.

This paper addresses fault-tolerant control and disturbance rejection for linear systems with actuator saturation and communication constraints. A method based on the Fourier series is proposed to handle the non-differentiability of periodic intermittent faults at jump points. A window function is used to reduce the Gibbs phenomenon and a criterion for selecting the expansion order is provided. By designing a dynamic event-triggered composite observer, system states, faults, and disturbances are estimated simultaneously while reducing the communication burden. The effects of actuator saturation are handled using a dynamic anti-saturation compensator combined with L2 gain, and a fault-tolerant controller is developed based on the estimated information. Finally, the effectiveness of the proposed scheme was validated through a numerical simulation and a practical example.