Observer-based Fault Diagnosis And Fault-tolerant Control Of Stochastic Distributed Systems

With the continuous improvement of the level of industrial automation,the complexity of modern control systems increase rapidly,which also lead to the increase of the possibility of systems failure.Once the fault occurs,it will reduce the performance of the systems,lead to the instability of the systems,and even cause serious property losses and casualties.Therefore,the implementation of fault diagnosis and fault tolerant control is of great significance to improve the reliability and safety of the systems.Meanwhile,due to the influence of systems nonlinearity and random disturbances,most of the variables in the stochastic systems are non-Gaussian variables.The traditional control methods with expectation and variance as optimization indexes are no longer suitable for non-Gaussian stochastic systems.Therefore,the stochastic distribution systems with output probability density function(PDF)as control objective are proposed.Therefore,this paper studies the problem of fault diagnosis and fault tolerant control for stochastic distribution systems based on observer,the main contributions of this paper are summarized as follows:(1)For the linear stochastic distribution systems where the linear B-spline is used to approximate the output PDF,the unknown disturbance is taken into consideration,the iterative learning observer is designed to estimate the fault,using the estimated value of the fault,a fault tolerant controller based on PI tracking is designed to realize the fault tolerant control.Meanwhile,H_?performance index is introduced to ensure the robust performance of the systems,the gain matrices of the observer and the controller can be calculated by solving the corresponding linear matrix inequality(LMI).Finally,the effectiveness of the design algorithm is verified by numerical simulation.(2)For the nonlinear stochastic distribution systems where the square root B-spline is used to approximate the output PDF,a fault diagnosis and fault tolerant control method based on intermediate observer is proposed.Firstly,an intermediate variable is introduced,and then an intermediate observer is designed to estimate the systems state and actuator fault simultaneously.Using B-spline model,the PDF shape control problem is transformed into a dynamic weight vector control problem,and a fault tolerant controller with PID structure is further designed,so that the PDF can still track to the desired PDF after the fault occurs.Finally,the simulation results verify the effectiveness of the designed observer and controller.(3)For the nonlinear stochastic distribution systems where the rational square root B-spline is used to approximate the output PDF,assuming that the nonlinear dynamic of the systems is unknown,the neural network functions are used to approximate the nonlinear dynamic,and an adaptive observer is designed to estimate the state and actuator fault of the systems.Furthermore,the robust PI fault tolerant controller is designed to ensure the robustness and tracking performance of the systems.Finally,the designed observer and controller are verified by simulation experiments.