IET Control Theory and Applications最新文献

This paper presents decentralized fuzzy filter design techniques for discrete-time nonlinear large-scale systems with uncertain interconnections. The subsystems of the nonlinear large-scale system and the decentralized filters are represented using Takagi–Sugeno (T–S) fuzzy model. Based on the closed-loop system with estimation error, the decentralized fuzzy filter design problem is addressed to guarantee the $�H_{\infty }$ filtering performance. A sufficient condition is derived to satisfy both the stability condition and the $�H_{\infty }$ filtering performance, and its constructive condition is reformulated into various linear matrix inequality (LMI) formats. Finally, the effectiveness of the proposed methods and procedures is demonstrated through a simulation example. The main novelties of this paper can be broadly summarized into two points: First, the decentralized fuzzy filter design techniques are effectively proposed for uncertain interconnection by using the maximum interconnection bound. Second, the decentralized fuzzy filters are designed when the premise variable is not only measurable but also non-measurable.

This study addresses the adaptive dynamic quantization control problem for Markov jump systems (MJSs) subject to multimode injection attack (IA). For the controlled system, the Markov chain and multimode IA, modelled as a categorical-distributed stochastic process, are jointly characterized by a hidden Markov model. Firstly, a sliding mode controller without quantization signals is designed and a mode-dependent Lyapunov function is constructed to establish the exponential ultimate boundedness condition of the MJSs. And then, under the frame that the state signal is quantized before transmitted to controller, an adaptive dynamic quantization strategy is further proposed such that the above controller can achieve the same control performance as the non-quantized case, eliminating the need for re-designing new controller. Finally, simulation results validate the effectiveness and robustness of the proposed quantized control strategy under combined dynamic quantization and IA scenarios.

Control barrier functions (CBFs) offer provable safety guarantees for dynamical systems. However, the existing control syntheses heavily depend on safety priors, which limits their applicability to data-driven scenarios. To deal with this issue, we propose an optimization-based approach to obtain data-driven high-order CBFs (HOCBFs) from the offline dataset. We first construct a group of safety-guaranteed sets from safe trajectories and provide verification conditions for valid HOCBF candidates via Lipschitz properties. Next, we propose a novel optimization framework to obtain a data-driven HOCBF with enhanced feasibility. To solve the proposed optimization problem, we provide convex and non-convex optimization approaches with different neural network implementations. Finally, the proposed approaches are illustrated via two numerical examples from the safe exploration of an unmanned aerial vehicle in static but unknown environments.

This study introduces a two-stage estimation framework designed to mitigate the online computational burden associated with traditional state estimation algorithms, such as the Kalman filter and extended Kalman filter, which require continuous computation of time-varying gains. The proposed framework integrates a fixed-gain Luenberger observer for base estimation with a lightweight neural network-based compensation module, which is trained offline and deployed online to enhance accuracy by correcting residual errors caused by fixed-gain limitations. To handle both linear and non-linear systems effectively, distinct loss functions are constructed by integrating prior knowledge of the system model, enabling the network to produce physically meaningful and interpretable outputs. Numerical simulations validate the effectiveness of the proposed method, demonstrating improved estimation accuracy and reduced computational cost compared to conventional estimators.

This study presents a robust state estimator for linear impulsive switched systems subject to unknown input signals and measurement noise. Such systems exhibit highly complex behaviour due to sudden dynamic changes and external disturbances, which significantly complicates the state estimation process. The presence of unknown inputs and measurement noise may not only degrade estimation accuracy but also challenge the stability of conventional observers. To overcome these difficulties, a novel full-order observer design is proposed that does not require output differentiation, thereby eliminating impulsive effects typically caused by differentiating noisy outputs. Moreover, the proposed framework explicitly accounts for measurement noise, an aspect that has rarely been addressed in the context of impulsive or switched systems. The proposed estimator ensures exponential convergence of estimation errors while simultaneously enhancing accuracy and robustness under impulsive switching conditions. The effectiveness of the method is demonstrated through simulations on an electrical circuit with variable loads. The results confirm that accurate and stable state estimation is achieved even under fast switching and in the presence of disturbances, thereby highlighting the practical applicability of the proposed approach.

The challenges of communication interruptions have caused unmanned vehicles to lose the real-time transmission of essential data. The traditional control architecture, designed for continuous measurement, is inadequate to address the constraints posed by the fragility of facilities in harsh maritime conditions. Therefore, it is necessary to enhance the transient stability performance in non-ideal conditions and environments. This work proposes a novel intermittent predefined-time control algorithm. This approach eliminates the need for frequent measurement and recalculation of control laws at every moment, and extends the usage scenarios of intermittent control from first-order systems to second-order systems. During the holding interval, the control input remains as the last control signal in the driving interval, provided that the validated stability conditions are satisfied. The designed controller also ensures stability with zero control inputs during the holding interval. Furthermore, the proposed method aims to optimise energy consumption in control and measurement. Using the Lyapunov theorem and mathematical induction, sufficient conditions for achieving predefined-time convergence in intermittent control are derived and verified with unmanned vehicles. The proposed method demonstrates a 5.4% reduction in tracking error compared to the fixed-time control method, and a 7.9% reduction in tracking error compared to the event-triggered predefined-time control method. This approach reduces the frequency of control law recalculations, measurements and energy consumption.

This paper proposes a novel fixed-time active fault-tolerant control method for second-order affine nonlinear systems. The key innovation lies in the integration of a super-twisting barrier function based adaptive sliding mode control (ST-BFASMC) with a radial basis function neural network (RBFNN) observer, achieving simultaneous improvements in convergence speed and fault tolerance. Firstly, a RBFNN observer with weight and centre value update strategy is designed. Secondly, a non-singular fast terminal sliding surface and control law with a super-twisting term are constructed. Furthermore, the fixed-time convergence properties of both the controller and observer are rigorously proven using Lyapunov stability theory. Experimental studies on quadrotor UAV attitude control demonstrated that the adopted RBFNN observer achieved over 60% improvement across all performance metrics compared to baseline methods, while the control algorithm exhibited more than 30% enhancement in multiple indicators relative to conventional ASMC and barrier function based ASMC (BFASMC) approaches. These results validate the algorithm's strong robustness and fault-tolerant capability in the presence of actuator failures.

This paper presents the first study on the discrimination problem of symmetric games with respect to a permutation group. Based on the algebraic representation of games, the necessary and sufficient conditions for symmetric games with respect to a permutation group are first derived. Subsequently, by the properties of the semi-tensor product of matrices and swap matrices, a minimal discriminant equation system comprising the fewest equations for verifying symmetric games with respect to a permutation group is constructed. Finally, three examples are provided to illustrate the main theoretical results presented of the paper.

This paper studies the graphical game for multi-agent systems with Lipschitz nonlinear dynamics over directed graphs. First, a distributed optimal control policy is presented to ensure the leader-following consensus. Then, a modified cost function in the framework of graphical games is designed such that the weighting matrix is time-varying and relies on the Lipschitz nonlinearity, leading to the global Nash equilibrium. A simulation study is finally provided.

In this paper, we present two adaptive control frameworks for physical human–robot interaction without direct force measurements. One enabling the robot to guide human motion using tracking control, and the other allowing it to follow human motion via admittance control. Both frameworks incorporate adaptive algorithms to estimate interaction forces, which are then integrated into controller designs to enhance performance and adaptability. In the first framework, the estimated forces are used to compensate for human-applied forces, enabling the robot to accurately follow a predefined trajectory. In contrast, the second control framework uses the estimated forces to adjust the robot's reference velocity, aligning its motion to human intentions. We employ Lyapunov's technique to demonstrate the stability and uniform ultimate bounds of the responses. Simulation and experimental results are presented to validate the proposed control algorithms. These results indicate that the approaches provide promising solutions for human–robot interaction with reduced cost and complexity.