The present work proposes an integral sliding mode multiobserver (ISMMO) for nonlinear systems in the presence of sensor faults. By augmenting the sliding manifold with an integral term, the sliding mode multiobserver exhibits low-pass filtering dynamics that attenuate high-frequency chattering while preserving the stability of the estimation error dynamics. For the simultaneous estimation of the states and the sensor faults, these faults are considered as auxiliary state variables, thus an augmented multiobserver structure based on a discrete-time decoupled state multimodel is constructed. The stability conditions are presented in terms of Linear Matrix Inequalities (LMIs) using a common Lyapunov function. The superiority and effectiveness of the proposed ISMMO are illustrated through an experimental validation on a chemical reactor and a comparative study.
In this paper, a new discrete-time event-triggered mechanism (DTETM) is proposed, which is used in the framework of designing a discrete-time event-triggered observer (DTETO) for Takagi–Sugeno (T-S) fuzzy models with external disturbances. Existence conditions for the proposed DTETO are derived and formulated in terms of a convex optimization problem, which gives unknown observer matrices and minimized attenuation levels. Finally, an effective algorithm is obtained to estimate the state vector of the uncertain wind turbine system with a permanent magnet synchronous generator, the Chua’s circuits, and the Rossler chaotic system. Compared with the existing T-S fuzzy observers operated based on time-triggered mechanisms or periodic memory sampled-data mechanisms, which may lead to wasteful communication resources, the proposed approach in this paper uses a novel Zeno-free DTETM, which effectively reduces communication and estimation resource usage while ensuring the desired estimation performance. Moreover, unlike the existing discrete-time event-triggered T-S fuzzy observers and fuzzy filters, which are effective in estimating the state vectors of discrete-time fuzzy dynamic models, the one in this paper may be useful in estimating the state vectors of continuous-time fuzzy dynamic models.
The H ∞ control technique aims to attenuate the effect of external disturbances on dynamical systems. This control technique is very important for active suspension systems since these systems are often affected by external disturbances, such as adverse road profiles. So far, the existing control design methods of active suspension systems have not been developed for the design of controllers using an event-triggered mechanism for uncertain active suspension systems subject to external disturbances. To fill this gap, in this paper, we consider the design of a controller using a discrete-time event-triggered mechanism for a half-vehicle uncertain active suspension system with external disturbances. For this, we first propose a discrete-time event-triggered mechanism, and then design an event-triggered controller for the half-vehicle uncertain active suspension system. Finally, we demonstrate the effectiveness of the obtained results by numerical results and simulations. Unlike the existing controllers using periodic sampling mechanisms or static continuous-time event-triggered mechanisms, the proposed controller in this paper uses an extra dynamic variable and depends on discrete supervision, which may effectively save communication resources.
Unmanned surface vehicle (USV) formation control in marine environments is challenged by the underactuated nature of USV dynamics, two-dimensional operating constraints, and unpredictable environmental disturbances. Traditional fault-tolerant control (FTC) methods for unmanned systems are difficult to directly implement in USV formation systems, especially when subject to communication link failures. This paper introduces an innovative distributed FTC framework tailored for the coordination of USV formations, especially in challenging environments. Firstly, to overcome the limitation of insufficient control inputs due to underactuation, head nodes are constructed within the formation, enabling effective command dissemination and simplifying the influence of external disturbances such as friction. Secondly, a virtual leader is introduced to generate an ideal reference trajectory for the formation, which enhances overall robustness. Building upon this, a distributed FTC strategy is developed based on an adaptive state observer and the virtual leader-following control method. The observer adaptively estimates the system states and fault information in real time, while a backstepping-based formation controller compensates for both actuator and communication link faults in a distributed manner. Extensive simulation studies have been conducted to evaluate the fault tolerance mechanisms in distributed systems, which are crucial for maintaining system stability and data integrity in the face of potential node or network failures. Studies demonstrate significant improvements over existing methods in terms of fault resilience, trajectory accuracy, and formation robustness. It can be observed that the FTC approach effectively suppresses steady-state oscillations and maintains errors at the order of 10 −4 m, whereas the non-FTC case exhibits significantly larger oscillations during the same time interval and even shows a tendency toward divergence at certain moments.
This paper investigates the nonlinear dynamic behavior of a three-dimensional energy supply–demand system with government regulation and time delay. A mathematical model is first established. By employing the τ -decomposition strategy, the exact stability boundary associated with the timing of government regulation is determined, and analytical expressions in terms of the system parameters are derived. The conditions for the existence of bifurcation and its direction are then identified. Unlike traditional analytical methods, the proposed approach avoids computing the derivative of the real part of the eigenvalues with respect to the time delay when determining the bifurcation direction. The center manifold theorem and normal form theory are subsequently used to analyze the Hopf bifurcation. In contrast to conventional approaches that require complicated integral calculations, a novel bilinear form is constructed to reduce inner-product operations in infinite-dimensional space to pure matrix algebra. Numerical simulations verify the theoretical predictions, and symbolic computation makes the dependence of key quantities on system parameters explicit. These results reveal the stability boundary of government regulation and the transition from steady states to periodic oscillations, while also providing a more efficient theoretical framework for analyzing other three-dimensional complex systems.
The high-frequency response pilot-operated electro-hydraulic proportional directional valve (HFRPPV) serves as a critical component in the hydraulic systems of high-end power equipment, and its control precision and response speed significantly impact the performance of the hydraulic system. In traditional fuzzy proportional-integral-derivative (PID) control, the thresholds of the fuzzy controller are fixed and cannot automatically adjust their range based on the system error and its change rate. To enhance the control performance of fuzzy PID systems, an adaptive fuzzy PID control with full variable universe and anti-windup (AFPID_FVU_AW) is proposed. The structure and operating principle are introduced, and its mathematical model is constructed. Building upon traditional fuzzy PID control, a scaling factor function was incorporated into the input and output of the fuzzy control algorithm, and an anti-windup module was added to the output of the PID controller. In the simulation model of the HFRPPV control system, the performance of fuzzy PID and AFPID_FVU_AW control was compared and analyzed across four aspects: pulse width modulation (PWM) duty cycle, pressure in the chambers at both ends of the main spool, displacement of the main spool, and output flow of the main valve. An experimental testing platform was established to validate the effectiveness and correctness of the simulation model. Experimental results indicate that during continuous step variations, both control strategies exhibit comparable control accuracy. However, AFPID_FVU_AW demonstrates significant superiority over fuzzy PID in terms of overshoot, response time, and stability, proving that the proposed control strategy can effectively enhance control performance. The AFPID_FVU_AW control strategy addresses the shortcomings of conventional fuzzy PID control, namely its poor adaptability, significant overshoot, and sluggish response. It represents a substantial theoretical innovation with considerable practical application value.
The rapid advancement of autonomous vehicle technology has raised stringent requirements on the accuracy and robustness of vehicle path tracking control systems. However, the nonlinearities of vehicle dynamics, system uncertainties, and bandwidth limitations of in-vehicle networks pose significant challenges to the design of high-performance controllers. Existing hierarchical control methods often treat path tracking calculation and steering actuator execution as independent modules, typically assuming that the lower layer can perfectly execute the steering commands from the upper layer. Moreover, the growing number of onboard sensors challenges the data transmission capacity of the in-vehicle communication network. Conventional adaptive event-triggered (AET) schemes suffer from monotonically evolving thresholds, which eventually causes them to degrade into static event-triggered (SET) strategies and lose adaptive capacity. To address these issues, this paper proposes an integrated path tracking control strategy for autonomous vehicles based on a steer-by-wire (SbW) system, incorporating a nonmonotonic adaptive event-triggered (NAET) mechanism. The proposed strategy unifies path tracking control and steering actuator dynamics into a constrained optimization problem, and the NAET mechanism maintains adaptive threshold adjustment without degrading into an SET strategy. A robust fuzzy dynamic output-feedback (DOF) controller is designed to improve tracking accuracy, reduce reliance on sensors, and conserve in-vehicle network resources. Hardware-in-the-loop (HIL) experiments validate the effectiveness of the proposed method. The results demonstrate that the integrated framework achieves superior path tracking performance, while the NAET scheme significantly reduces redundant data transmissions over the in-vehicle network.
This paper develops a disturbance observer-based adaptive sliding mode control strategy for a class of nonlinear time-delay systems subject to exogenous disturbances and delayed nonlinear uncertainties. A fixed-time nonlinear disturbance observer is designed using dual-power feedback terms to achieve precise estimation of external disturbances. Barrier functions are employed to design an adaptive controller that dynamically adjusts gains for time-delay uncertainty compensation, removing the initial condition dependence of common approaches. The Lyapunov-Razumikhin method is employed to handle large time-varying delays, and the fixed-time convergence of the sliding surface is established through Lyapunov stability theory. The effectiveness of the proposed method is validated through its application to a continuous stirred-tank reactor system and a numerical simulation.
Deep learning approaches have been the subject of considerable interest and challenge in various fields. In this work, we propose a deep learning methodology for automatic modeling and control applications based on an Internal Model Control (IMC) structure. Our study focuses on improving the modeling and control of a transesterification reactor using deep neural network approaches. To emulate the direct dynamics of the system, a feedforward deep multilayer perceptron (MLP) neural network was trained and then used as a neural controller within the IMC framework. Experimental results demonstrate that the proposed modeling approach can accurately capture the reactors nonlinear behavior, while the deep learning-based controller ensures robust and efficient tracking of desired reference trajectories under various operating conditions. Comparative analysis with conventional control techniques highlights the superiority of the proposed method and shows that deep neural networks, when integrated with an IMC strategy, offer a promising framework for the advanced control of complex chemical processes.
The paper presents a fuzzy control strategy for nonlinear direct current (DC) microgrids, taking into consideration intermittent denial-of-service (DoS) jamming attacks, random false data injection (FDI) deception attacks, event-triggered communication mechanism (ETM), quantiser, noise, and mismatching premises. Firstly, using characteristics of attacks and the microgrid's fuzzy model, a resilient ETM is introduced to reduce the data transmission rate, which also excludes the dropout phenomenon induced by attacks. Second, considering the impacts of hybrid attacks, ETM, noise, quantiser, and mismatching premises simultaneously, a fuzzy switched system model is derived. Third, by applying the piecewise Lyapunov theory, mean-square exponentially stable criteria that satisfy performance are presented, which establish a quantitative relationship between the influencing factors and the stability of the system. In addition, sufficient conditions are presented for designing the switched fuzzy controller with mismatching premises. Finally, the effectiveness of the proposed methods is confirmed.
To address the positioning and antisway control problem of bridge crane operations under disturbances, this paper proposes a coupled nonsingular fast terminal sliding mode adaptive tracking antisway control method. This approach enables the trolley to reach the target position accurately and rapidly within a finite time while effectively suppressing the load swing angle. Specifically, a generalized error signal comprising the trolley displacement and the load swing angle is first constructed. Combined with system parameters, an S-curve is employed for trajectory tracking to ensure coordinated control of positioning and antisway. Subsequently, a coupled sliding surface is designed based on the nonsingular fast terminal sliding mode control theory. Meanwhile, an adaptive law is introduced to estimate the upper bound of disturbances online, which is then incorporated into the control law to reduce system chattering. Finally, the Lyapunov method and LaSalle’s invariance principle are utilized to prove the finite-time asymptotic stability of the closed-loop system at the equilibrium point, and the effectiveness of the proposed controller is validated through extensive simulations and experiments. The results demonstrate that the proposed method exhibits excellent control performance, achieving remarkable positioning and antisway effects.
The centerline of a bobsleigh track defines its geometry and is fundamental for simulation modeling. However, publicly available centerline data are limited, and constructing training systems solely from two-dimensional (2D) centerlines is often imprecise. To address this issue, this study establishes a constrained optimization framework for three-dimensional (3D) bobsleigh track generation that explicitly incorporates geometric design constraints and Olympic safety rules, introduces a segment-wise height difference parameterization, and develops a track gradient based projected gradient descent solver validated on real tracks. Within the selected trajectories and parameter ranges considered in this study, the proposed algorithm generates 3D centerlines whose characteristic parameter trends closely follow those of the actual tracks, whether using real or scaled 2D data. Compared with the actual tracks, the maximum deviations in total length, height difference, and average gradient are 0.1%-1.1%, 1.0%-16.4%, and 0.8%-13.5%, respectively. Track sensitivity to segmentation and height difference weighting varies with the data type, and appropriate choices of these parameters effectively reduce deviations and enhance the geometric accuracy of the reconstructed centerlines. Overall, the proposed framework offers a flexible and efficient tool for supporting 3D bobsleigh track centerline generation.
The Multi-Pressure Rail (MPR) system represents an efficient, pressure-controlled alternative to the often-inefficient flow controlled hydraulic control systems present in many agricultural vehicles. This work aims to expand on prior conceptual and simulation work by the authors on such a system by improving the generality of the system's control scheme and further demonstrating savings potential through experimental validation. This is achieved through a novel form of the system's core pressure level and mode selection logic that aims to reduce computational cost and improve scalability to implements with larger numbers of actuators. Two forms of this revised control logic are proposed, one prioritizing efficiency and the other prioritizing stability. The modifications necessary to convert the reference system into a full-scale prototype machine are then proposed, and this prototype is tested in an array of realistic operating conditions. From the tests performed on both the baseline system and the MPR solution it is found that the simplest controller strategy (Fixed HP MPR system) matches the dynamic performance of the baseline solution, while the more aggressive controller (Var HP MPR) does experience a slight reduction in dynamic performance. Both MPR systems achieve close to 50% power reduction and over 90% efficiency gain for the most common operating conditions of the reference vehicles, while around 40% power reduction and 65%-78% efficiency gain are observed at less common high-speed operating conditions.
This article examines the bipartite consensus problem in multi-agent systems affected by actuator saturation and external disturbances. Based on the enlargement of the contractively invariant set, sufficient conditions are derived to ensure bipartite consensus and to evaluate the domain of consensus attraction under input saturation and magnitude-bounded disturbances. For systems subject to energy-bounded disturbances, a criterion is further developed to ensure that the system trajectories starting from an inner ellipsoid remain within an outer ellipsoid while achieving bipartite consensus. Moreover, the analytical problems are reformulated as convex optimization problems constrained by linear matrix inequalities. This formulation improves both the feasibility and computational efficiency of the proposed control method. Finally, numerical simulations are conducted to validate the effectiveness and superiority of the developed bipartite consensus control strategy.
This paper uses a fixed-time (FxT) dynamic event-triggered control(DETC) framework to study distributed multi leader-follower multi-group consensus for multi-agent systems(MASs) under disturbances. A normalized state-dependent variable-exponent (SDVE) design, which avoids the conservative tuning associated with constant-exponent techniques and modifies the convergence rate online in accordance with the system evolution, enforces FxT convergence. To improve resilience against matching uncertainties and external perturbations, a variable-exponent super-twisting algorithm (VSTA) is devised on nonlinear sliding manifolds. In order to facilitate simultaneous intra-group agreement and inter-group coordination while tracking different leader trajectories, agents are divided into several groups and communicate over a complex-valued communication graph. The FxT control architecture incorporates a DETC mechanism that ensures consensus without constant information transmission, hence reducing communication overhead. By establishing a positive infimum of the inter-event intervals and guaranteeing global FxT group consensus using just local information, the suggested technique eliminates Zeno behavior. Comparative numerical studies with conventional sliding mode and event-triggered approaches demonstrate significant communication savings, enhanced resilience, faster convergence, and a substantial reduction in chattering.
Cervical spondylosis is exerting an increasingly severe impact on people's daily lives. Owing to the proven effectiveness of mechanical traction therapy in cervical rehabilitation, considerable research efforts have been devoted to the development of cervical rehabilitation robots (CRR). This paper presents the design and implementation of a CRR driven by pneumatic artificial muscle (PAM). In contrast to conventional motor-driven mechanisms, PAM provides superior compliance and continuous deformation capability, thereby offering cushioning protection for the cervical spine during traction. To guarantee both stability and safety during rehabilitation, an active disturbance rejection sliding mode-admittance control (ADRSM-AC) strategy is proposed for achieving force-displacement coupled compliant control of the PAM. Specifically, an inner-loop active disturbance rejection sliding mode control (ADRSMC) is designed to suppress the inherent nonlinearities of PAM and external disturbances, enabling high-precision displacement regulation. On this basis, an outer-loop admittance control scheme is implemented to evaluate deviations between actual and desired traction forces. By inferring the patient's cervical muscle strength and movement intention, the system adaptively adjusts the target trajectory, thereby ensuring compliant interaction. Comprehensive experimental validation on a robotic prototype confirms the effectiveness of the proposed framework. The results demonstrate that the control scheme achieves accurate position tracking while adaptively regulating the trajectory in accordance with patient intention, ultimately safeguarding cervical spine safety throughout the rehabilitation process.
This paper investigates the secure control problem of networked control systems (NCSs) subject to partially unknown dynamics and network-injected attacks. To address these issues, a novel model-free online learning controller is proposed, which integrates adaptive dynamic programming (ADP) and sliding mode control (SMC) to eliminate reliance on accurate system models. The model-free SMC scheme is designed using only local state information and known bounds of disturbances/attacks, while ADP is employed to approximate optimal control policies. To further alleviate the communication burden of traditional SMC in networked environments, an event-triggered mechanism (ETM) is integrated into the control framework, yielding an ADP-based event-triggered sliding mode control (ETSMC) method, which combines ADP’s adaptive optimization and ETM’s communication efficiency. The proposed ADP-based ETSMC strategy guarantees the reachability of sliding surface and the asymptotic stability of sliding dynamics concurrently, avoiding the need for system identifiers or command generator models. Simulation results demonstrate that the proposed ETSMC ensures control performance and robustness while reducing communication burden, thereby verifying its feasibility and practicality.
Fault diagnosis is critical for ensuring the reliability and safety of industrial systems. Recently, graph convolutional networks (GCN) have gained significant attention for fault classification due to their ability to model complex dependencies in sensor signals. However, existing GCN-based approaches face two major limitations: (1) their adjacency matrices are typically predefined and fixed, failing to adaptively capture the varying importance of motor and gearbox nodes under different fault conditions, and (2) they lack comprehensive multi-scale sampling strategies, limiting their ability to construct informative time-series representations. To address these issues, we propose a multi-scale adaptive graph convolutional network (MSAGCN) for rapid fault classification. MSAGCN introduces an adaptive graph weighting mechanism, enabling the model to dynamically adjust node importance based on fault characteristics. Additionally, a multi-scale sampling aggregation strategy is incorporated to extract rich temporal features, enhancing the model’s discriminative power. Experimental results demonstrate that MSAGCN outperforms conventional GCN-based methods, achieving improved classification accuracy across various fault scenarios. The proposed framework establishes a robust baseline for fast and accurate fault diagnosis, highlighting the importance of adaptive graph structures and multi-scale feature extraction in industrial monitoring applications.
This paper studies the adaptive stabilization problem of a class of uncertain beam PDE-ODE (partial differential equation - ordinary differential equation) cascade system under time-varying distributed and boundary disturbances. To eliminate the influence of distributed and boundary disturbances on the system and achieve stabilization, an adaptive fuzzy boundary control scheme based on the inversion method is proposed. This scheme integrates the uncertainty of system parameters and environmental interference, and dynamically approximates the uncertain nonlinear terms by constructing an adaptive fuzzy logic system, thereby improving the adaptability and robustness of the system. On this basis, a barrier Lyapunov function is constructed to constrain the output state of the system and a virtual control law is designed to ensure that the system state does not exceed the limited range. Finally, without simplifying or discretizing the infinite-dimensional dynamics, the stability of the closed-loop system is analyzed and proved. Numerical examples demonstrate the effectiveness of the proposed control.
This paper investigates the problem of adaptive event-triggered sliding mode control (AET-SMC) for uncertain switched systems constrained by persistent dwell time (PDT) switching rules, in which the C/A channel may be subject to random denial-of-service (DoS) attacks. A Markov chain-based model is employed to accurately characterize the time-dependent nature of DoS attack behavior. An adaptive event-triggering mechanism is designed to dynamically adjust the triggering threshold according to the system state variations, thereby effectively reducing the communication burden while preserving system stability. A sliding mode controller is formulated by incorporating an adaptive event-triggered mechanism and a Markov-based DoS attack model, which ensures robustness against system uncertainties and external disturbances. The sliding surface is designed as a function of the system state and the controller gain, enabling effective tracking and disturbance rejection even under random network attacks. Sufficient conditions are derived to guarantee the mean-square exponential stability of the closed-loop switched system and the reachability of the specified sliding surface by constructing appropriate Lyapunov functions and employing Linear Matrix Inequality (LMI) techniques. Numerical simulations and practical DC motor control case studies demonstrate that the proposed method effectively safeguards robust stability and maintains superior dynamic performance under random network attacks and switching disturbances.
