Fault Tolerant Control for Nonlinear Networked Stochastic Distribution Systems With Quantized Signals and Packet Dropouts

Abstract

The fault-tolerant tracking control problem is investigated for nonlinear networked stochastic distribution systems (SDSs) under signal quantization, packet loss and system noise. Firstly, a novel logarithmic uniform quantizer is used to quantize the control signal and feedback signal and a data packet loss model is established through Bernoulli random process, which establishes the theoretical foundation for subsequent fault diagnosis (FD). On this basis, a FD observer is designed, which can accurately estimate the fault information under noise and provide reliable fault estimation information for subsequent fault-tolerant control (FTC). Subsequently, to maintain satisfactory tracking performance after fault occurrence, a state feedback fault-tolerant tracking control strategy is proposed based on the parallel distributed compensation technique. Finally, the superiority of the presented FTC method is proven through a numerical example and a practical example. Note to Practitioners—The motivation is to improve the security of SDSs in network environments without changing the system structure or increasing costs. When the fault occurs, quantization and packet loss can result in missing information containing fault characteristics, thereby affecting the real-time and accuracy performance of FD, and also increasing the difficulty of designing fault-tolerant controllers. Regarding to the above problems, a novel FD and FTC strategy is proposed, which can effectively compensate the negative influences of packet loss, actuator fault and quantization error on the system performance.