具有混合执行器故障的多智能体分布式有限时间自适应协同容错控制

针对多智能体编队系统执行器发生故障时,所引起的参数不确定以及系统瞬态不稳定问题,本文采用径向基函数神经网络(radial basis function neural networks,RBFNNs)对不确定参数(未知函数)进行估计.同时,基于反推技术设计出合理的自适应容错控制器,并通过有限时间理论保证系统实现瞬态稳定....

Distributed finite-time adaptive cooperative fault-tolerant control for multi-agent systems with integrated actuators faults

Aiming at the uncertain parameters and transient instability of the system caused by the actuator fault for the multi-agent formation system, this paper adopts the radial basis function neural networks (RBFNNs) to approximate the uncertain parameters (the unknown nonlinear function). A reasonable adaptive fault-tolerant controller is designed based on the back-stepping technology, and the transient stability of the system is guaranteed by the finite time theory. Firstly, this paper adopts ten agents as the controlled objects, and constructs a nonlinear multi-agent system model based on the theory of directed communication topology. Secondly, based on the approximation characteristics of the RBFNNs, a reasonable fault-tolerant controller is designed by combining the back-stepping technology and the dynamic surface technology to compensate the unknown nonlinear actuator faults in the multi-agent, and the finite-time theory is used to solve the system transient state instability problem. Then, the stability and fast convergence of the controller are analyzed based on the Lyapunov stability theory. Finally, through the comparison of two calculation examples, it is verified that the performance of the designed controller is better than the traditional back-stepping technology, which provides an effective research idea for engineering practice.