International Journal of Adaptive Control and Signal Processing最新文献

This paper focuses on the parameter estimation problem for pseudo-linear systems with autoregressive moving average noise. In order to reduce the computational complexity of the identification algorithms, the original pseudo-linear system is decomposed into three submodels and a three-stage auxiliary model-based recursive generalized extended least squares (3S-AM-RGELS) algorithm is proposed based on the hierarchical identification principle. The convergence analysis is provided to show that the parameter estimation error can converge to zero under the presented 3S-AM-RGELS algorithm. Finally, numerical simulations demonstrate the effectiveness of the proposed algorithms.

Sensor fault detection and isolation in multi-agent systems (MAS) with uncertain dynamics and undirected, connected communication networks is addressed in this article. The proposed approach involves a two-step process: First, fault detection, and then likelihood-based fault isolation. A novel fault reconstruction technique is introduced by tuning the unscented Kalman filter (UKF) noise covariance matrices within the Q-learning framework. This adjustment helps reconstruct the uncertain states of the MAS and train the internal parameters of a neural network using historical measurements. This innovative method is referred to as Enhanced reinforced UKF (ERUKF). To reduce neural network approximation errors, a robust control term utilizing the hyperbolic tangent function is applied. The stability of ERUKF, when combined with the robust control method, is mathematically proven using the Lyapunov theorem. Simulations illustrate that ERUKF exhibits lower estimation errors compared to adaptive UKF, achieving a 96.67% success rate in fault isolation under Monte Carlo (MC) simulations.

For stochastic nonlinear time-delay systems with full-state constraints and input saturation, the existing control methods cannot be directly applied. To solve the control problem, an adaptive tracking control scheme is proposed here. The barrier Lyapunov function (BLF) and Lyapunov-Krasovskii functional (LKF) are combined in a unified framework. The former is to prevent the violation of state constraints and the latter is to address the impact of time delay. A suitable auxiliary system with the same order as the considered system is constructed to compensate for input saturation. Moreover, the multi-dimensional Taylor network (MTN) is introduced to estimate unknown nonlinear functions in the process of controller design. With the help of the Lyapunov stability theorem and the backstepping technique, an adaptive MTN tracking controller is developed to ensure that all the closed-loop signals remain bounded in probability and that the system state constraints are never violated while the output signal tracks the reference one successfully. Simulation examples with three types of reference signals ($��{�y�}_{�d�}�=�0�.�5�\sin �\left(�2�t�\right)�$, $��{�y�}_{�d�}�=�0�.�9�$ and the square reference signal) are provided to verify the validity of the designed control approach.

This article proposes a fixed-time adaptive tracking control strategy for a class of high-order non-strict feedback nonlinear systems with asymmetric output constraints. The key technique of this strategy is to guarantee that the output of the system does not break the pre-set constraints through the barrier Lyapunov function (BLF) while approximating any unknown nonlinear term through the fuzzy logic systems (FLSs). We design a fixed-time fuzzy adaptive controller to ensure that the closed-loop error signal converges to a tunable neighborhood near zero in a fixed-time instant. Compared with the existing results, the fixed-time adaptive control is realized and the semi-global fixed-time stability of the closed-loop system is proved theoretically. Finally, the effectiveness of the proposed scheme is verified by a numerical simulation.

An adaptive preset performance tracking control scheme is proposed for a class of uncertain strict feedback stochastic nonlinear systems with full state constraints and input dead zones. A new coordinate transformation and obstacle Lyapunov function are adopted to solve the problem of full state constraints and preset performance. The nonlinearity of the input dead zone is solved by splitting the dead zone nonlinear term into a linear term and a bounded disturbance term. A fuzzy logic system is used to approximate the unknown function in the system. A command filter is introduced to solve the “differential explosion” problem, and a compensation signal is used to compensate for the filtering error. An adaptive command filter controller is designed using the adaptive backstepping method and stochastic stability theory, which can ensure that all the closed-loop signals are ultimately bounded with semiglobal consistency in terms of the probability of quarter moments of an appropriate compact set. Meanwhile, the tracking error satisfies the predefined performance and converges to a small neighborhood close to zero. Simulation results show the effectiveness of the method.

This paper is concerned with the predefined-time tracking control problem for vehicular platoon systems composed of heterogeneous vehicles with arbitrary initial states under the influence of nonlinear disturbances. To counteract the influence of system uncertainties and disturbances, a practical predefined-time disturbance observer (PTDO) is constructed for lumped disturbance compensation in the closed-loop system. Meanwhile, a predefined-time control scheme is developed incorporating a predefined-time performance function (PTPF) and a composite error transformation mechanism to ensure that the position and velocity errors of the vehicle platoon system are constrained within a predetermined accuracy within a predefined time frame. The proposed approach eliminates reliance on initial conditions inherent to conventional prescribed performance control methods while demonstrating adaptability to diverse application scenarios. The efficacy and practicality of the developed control protocol are substantiated through rigorous theoretical analysis and numerical simulations.

In this study, a novel perspective on an adaptive bipartite consensus tracking scheme for multi-agent systems (MASs) in the face of sensor attacks is presented. The bipartite consensus tracking problem of MASs is transformed into the tracking problem of a single agent based on the shortest path method, and the assumption of structural balance for a digraph is removed. A significant challenge encountered in this article is that the original sensor measurement states cannot be directly and accurately measured and used after being subjected to malicious deception attacks. Therefore, to solve this issue, a novel coordinate transformation method including compromised states is proposed. Furthermore, the Nussbaum function is only introduced in the final step to avoid the effects of unknown attack gains resulting from deception attacks in the system, and a set of tuning functions are designed to resolve the singularity issue encountered during the adaptive iterative process of previous studies. A distributed security control scheme is designed using backstepping and Lyapunov stability theory, ensuring that all closed-loop system signals remain globally and uniformly bounded. Simulation results with a group of single-link robots validate the efficacy of the presented approach.

An adaptive fixed-time extended state observer (AFESO)-based nonsingular fast terminal sliding mode control (NFTSMC) method is proposed for a class of underactuated overhead crane systems with unknown external disturbances. Firstly, an improved fixed-time nonsingular fast terminal sliding mode surface is designed, which not only solves the singularity problems in terminal sliding mode control but also achieves system stabilization within bounded convergence time. Moreover, an adaptive fixed-time extended state observer is presented to estimate both matched and unmatched disturbances; then, the estimated value is used for feedforward compensation to compose the controller. Subsequently, to improve control quality and enhance robustness of the control system, a weakly reinforcement learning (RL) algorithm is used to update parameters of a chattering free reaching law parameter, upon which a novel fixed-time auxiliary error compensation system is constructed to solve the input saturation problem. The fixed-time stability analysis of the closed-loop system is derived by Lyapunov theory. Finally, the rapidity and effectiveness of the developed control method are demonstrated through simulation results.

This study focuses on adaptive fuzzy finite-time control for nonstrict-feedback nonlinear systems afflicted with actuator faults and asymmetric dead-zone, utilizing an event-triggered approach. Fuzzy logic systems are used to approximate the system's unknown terms. A relative threshold event-triggering mechanism is devised to reduce communication requirements. This mechanism ensures that the actuator only receives system input when a significant event occurs, thereby optimizing control efficiency. The proposed strategy integrates event-triggered adaptive fuzzy control with the backstepping method to achieve finite-time stability. This approach guarantees that the tracking error converges to a region near the origin, and ensures semiglobal practical finite-time stability for all signals within the closed-loop system. A numerical example and a real-world example of a pendulum system are used to illustrate the efficacy of the control approach.

This study proposes an adaptive gain scheduled attitude control strategy for the reentry phase of reusable launch vehicles (RLVs), which is model-independent and offers a convenient implementation. It originally integrates an extended state observer (ESO) with a sigmoid estimator to address the wide envelope of velocities and altitudes, along with the strong nonlinearities and aerodynamic uncertainties. In particular, the ESO is used for disturbance estimation, whereas the sigmoid estimator is applied for disturbance classification. The proposed design achieves the adaptive online estimation and compensation of the RLV control gain and reduces the model dependence. In addition, the closed-loop finite-gain stability is investigated based on the Lyapunov theory, and the tracking boundedness can be proved. Finally, the effectiveness and robustness of the proposed design are verified based on numerical simulations and Monte Carlo tests, and the advantages of weak model dependence and strong ability on disturbance rejection are highlighted in the proposed design compared with the PID control and linear active disturbance rejection control.