Asian Journal of Control最新文献

In this paper, output-feedback containment control problem of unmanned aerial vehicle (UAV) swarm with input saturation is studied under distributed denial of service (DoS) attacks. Any edge in the communication network may be attacked independently. The considered distributed DoS attack may occur at any time, and there is no limit to the number of DoS attacks, which is more in line with the actual attack scenario. The leader UAVs are noncooperative, which are embedded with unknown input. Firstly, to estimate the state signals, distributed observers are designed. Then, a general form of distributed DoS attack model is established. To deal with distributed DoS attacks, compensators are constructed for each follower UAV to estimate its neighbor signals when DoS attacks occur. The compensator calibration mechanism is introduced to calibrate the estimated values once the communication is restored. In this way, the flexible switching between the estimated values and the real values is realized to improve the ability to respond to DoS attacks. Then, an auxiliary system is designed to cope with the possible input saturation phenomenon. Finally, a novel variable-gain distributed containment control law is proposed, based on a quadratic term of the adaptive parameter, which shortens the convergence time and reduces the steady-state error. It is proved that the proposed scheme ensures that all the state errors, including containment error, observation error, adaptive parameter, and auxiliary variable, are uniformly ultimately bounded, and the effectiveness and superiority are verified by simulation experiments and comparative analyses.

This paper presents an Adaptive MIMO Sliding Mode Control (AMSMC) strategy for coordinating Active Front Steering (AFS) and Direct Yaw Control (DYC) systems to enhance vehicle stability and handling under uncertain conditions. Traditional Single Input Single Output (SISO) models fail to capture the complex interactions and nonlinearities inherent in vehicle dynamics, leading to suboptimal performance. The proposed method addresses these limitations by utilizing a Multiple Input Multiple Output (MIMO) framework, which accurately models the nonlinear interactions between AFS and DYC systems. Additionally, the method introduces a dynamic coefficient in the sliding mode control, enabling real-time adaptation to unknown uncertainties and enhancing robustness. The slip angle is estimated by an observer and a first-order delay is introduced into the lateral forces modeling to improve the accuracy of vehicle dynamics representation. The stability of the closed-loop system is further validated using the Lyapunov method. The performance of the proposed controller is evaluated considering the coefficient of road tire friction and parametric uncertainties in vehicle parameters, such as total mass, moment of inertia, and tire stiffness. The efficacy of this method has been rigorously confirmed through MATLAB simulations using a nonlinear 8-DOF vehicle model, which includes parameter uncertainties to ensure the control strategy's resilience to varying road conditions. The simulation results demonstrate significant improvements in the vehicle's stability and handling performance across a variety of driving maneuvers.

This paper focuses on addressing the challenges of finite-time output tracking control for unmanned aerial vehicles (UAVs) with a particular emphasis on security concerns. The potential impact of denial-of-service (DoS) attacks on the integrity of measurement-to-control communication channels is incorporated into our analysis. Such attacks may disrupt the central control unit's capability to access real-time measurement data, which encompasses the system states and the reference signal information. Under this consideration, we have formulated sufficient conditions based on the Lyapunov theory, aimed at ensuring that the system's state trajectory remains confined within a pre-established boundary over a designated time frame. However, note that there are some coupling terms of unknown matrix variables, and the developed conditions are nonlinear and difficult to be solved by some commonly used linear matrix inequality (LMI) techniques. In order to handle the nonlinear and non-convex constraints in the determination of the controller gains and the finite-time $��{�H�}_{�\infty �}�$ performance, we have proposed a novel iterative algorithm. The novel presented algorithm mainly combines genetic algorithm with LMI techniques. To validate the effectiveness of our approach, a case study of a UAV networked output tracking control system is conducted and detailed in the final section of this paper.

This work focuses on developing a mode-dependent estimator for neural networks that have semi-Markov jump parameters and sensor nonlinearities under the imperfect network environment. Semi-Markov jump process with its unique superiority is employed to depict mode switching within neural networks. The data sent over a communication network are considered to be at risk of cyber-attacks, where adversaries could introduce false data in a probabilistic manner. Due to complex practical situations, the sensor may exhibit nonlinear characteristics. A state estimator is utilized to produce estimates for the underlying neural networks, incorporating the sensor's nonlinear characteristics. Furthermore, an applicable estimator is devised by addressing the convex optimization problem. Lastly, the efficacy of the formulated estimator is demonstrated through a numerical illustration.

This paper aims for precise tracking with disturbance rejection and low energy consumption. An event-triggered controller based on a sampled-data enhanced extended state observer (SD-EESO) is designed for a networked tracking system. In this method, the measured output and the given signal are sampled and transmitted to an observer, which is designed to periodically estimate the negative disturbance and tracking errors. Furthermore, a dynamic event-triggered state feedback–feedforward controller is developed. It is shown that as long as the sampling period satisfies the given condition, the estimation error and the tracking error are both globally bounded stable. The numerical simulation and the direct current (DC) brush motor position control example show the superiority of the designed control scheme.

The observer-based event-triggered $��{�H�}_{�\infty �}�$ control for switched systems is studied. The system researched is affected by external disturbance, nonlinear perturbation and time delay, which helps expand the application range of switched system theory. First, to alleviate the high cost or difficulty of obtaining the system state, we construct an improved state observer, which echoes the structure of the considered system. Then, based on the observed state, we adopt a general event-triggered strategy (ETS) to save resources and give a proof of excluding the Zeno phenomenon. Next, using the mode-dependent average dwell time (MADT) and the Lyapunov–Krasovskii functional, the sufficient criteria for the system with $��{�H�}_{�\infty �}�$ performance are obtained. With the singular value decomposition (SVD) of column-full rank matrix, the observer gains, the controller gains, and the event-triggered parameter matrices are co-designed. Finally, a simulation example is implemented to demonstrate the value of this paper.

This paper investigates the problem of observer-based sliding mode control (SMC) of stochastic uncertain Markov jump systems using dual-driven event-triggered mechanism (ETM). The ETM is implemented not only in the system output to the observer channel but also in the feedback channel, forming a dual ETM, which can effectively reduce communication load. By designing a state observer, the error dynamics is derived, and subsequently, sliding mode dynamics (SMD) are formulated by proposing a specific type of integral sliding surface. Moreover, a SMC law is put forward to guarantee the attainment of the sliding surface by the system state within a finite time and to maintain the state within the domain of the sliding surface. Furthermore, sufficient conditions for the closed-loop system are established to ensure both the stochastic stability and a prescribed $��{�H�}_{�\infty �}�$ performance through linear matrix inequalities. Finally, two examples are provided to demonstrate the effectiveness of the proposed results.

The main subject of this paper is to analyze the $��{�H�}_{�\infty �}�$ performance of observer-based hybrid-triggered mechanism network control systems that are affected by dual cyber attacks. Firstly, facing this problem of unobservable system state, an observer model is introduced, which also establishes a solid theoretical foundation for subsequent controller design. Secondly, during data transmission, the systems are vulnerable to dual cyber attacks: The controller may face the threat of false data injection when it receives information from the observer, while the controller faces the threat of denial-of-service attacks when sending commands to the actuator. Thirdly, Lyapunov functions and linear matrix inequalities are employed to ensure that the resulting closed-loop system is asymptotically mean-square stabilized under the prescribed $��{�H�}_{�\infty �}�$ performance. Finally, a simulation example is given to demonstrate the effectiveness of the method in this paper.

An efficient identification method is proposed to estimate the time-varying parameters of a nonlinear dual-control aircraft. Based on the Extended Kalman Filter (EKF), the long-short-term memory (LSTM) neural network is introduced to adjust the filter gain to reduce the filter error caused by model uncertainty and linearization. To solve the heavy training burden of the stochastic gradient descent (SGD), RMSprop, and Adam training algorithm, an efficient training algorithm is presented, which uses Independent Extended Kalman Filter (IEKF) to transform the network weights training problem into an optimal weights identification problem to achieve the convergence speed of second-order training. To further reduce the training burden of the network, the network weights Matrix Factorization (MF) is introduced to reduce the number of learnable weights. In addition, the training performance of the LSTM neural network using MF structure and IEKF training algorithm is analyzed by the Lyapunov method, and the selection method of the artificial process noise matrices is given. Time-dependence data fitting experiments show that the efficient LSTM neural network proposed herein can greatly reduce the training burden of the network, and simulation results show that the Learnable Extended Kalman Filter (L-EKF) achieves better performance than some existing filtering methods in the time-varying parameter identification of the dual-control aircraft.