Fault Diagnosis And Fault Tolerant Control Of Stochastic Networked Control Systems

Due to the complexity of networked control systems,random delay and packet loss will inevitably occur during the operation of the control system.Once the data packet is lost,it is likely to cause incomplete information collection,and even lead to the collapse of the whole system.The problems of packet loss,delay and signal quantization is not considered in traditional control methods,so it is difficult for traditional control methods to be to networked control systems.Based on the above analysis,it is the focus of this thesis to design a kind of control compensator to make the system run normally after the occurrence of time delay or packet loss.Based on the information transmission of networked control systems,the actuator fault in the case of random delay and packet loss is fully considered in this thesis.For the post fault system,different types of observers were designed to obtain the online estimation information of the fault and state of networked control system.Then,based on the online estimation information,appropriate controllers were designed to compensate for the delay packet loss and the impact of the actuator fault.The main contents of this thesis are as follows:1.A class of nonlinear networked control systems with time delay and actuator fault is studied.Firstly,the random time delay block in the networked control systems is processed.The networked control systems with random time delay block is transformed equivalently by Laplace transform and Laplace inverse transform,and the equivalent control system without time delay is obtained.A type of learning observer is designed to estimate the system fault and the state information online.Based on the obtained on-line fault estimation information and the state estimation information,a feedback based fault-tolerant controller is designed to compensate the fault effect of the system,so as to ensure the stability of the system after the fault occurs.Finally,in order to prove the effectiveness of the algorithm,a singlelink robot is taken as an example for the numerical simulation.The theoretical and example simulation results both prove the effectiveness of the algorithm.2.On the basis of the first part,the parameter uncertainties and external disturbances of the model in the networked control systems are considered.At the same time,the phenomenon of random time delays between the controller to the actuator and the sensor to the controller is also considered.For the nonlinear term of the system,a more complex and conservative nonlinear constraint condition is considered,and the one sided Lipschitz condition is used to deal with the nonlinear term.On this basis,an adaptive fault diagnosis observer is designed to estimate the fault information and state information of the nonlinear networked control system online.Then,based on the online estimation information obtained by the adaptive observer,a fault-tolerant control algorithm based on the PI control strategy is designed to compensate the fault effect of the networked control systems.Finally,a switched reluctance motor is taken as an example to verify the effectiveness of the algorithm.3.Further the case of actuator fault under the packet loss occurs is mainly studied.It is assumed that the packet loss process obeys Bernoulli distribution,and the Bernoulli sequence is used to describe the packet loss process.The mathematical expectation and variance of packet loss is the corresponding constant.A kind of discrete proportional integral fault diagnosis observer is designed to estimate the fault information and state information of the nonlinear network control systems on line.Based on online fault diagnosis and state estimation information,a faulttolerant controller based on model reference control strategy is designed to compensate the effect of the fault.Finally,the effectiveness of the algorithm is verified by the numerical simulation of a DC motor.