Research On Fault Diagnosis And Fault Tolerant Control For Stochastic Distribution Systems With Communication Constraints

The stochastic distribution system(SDS)is a kind of systems in which the output probability density function(PDF)is taken as the research objective.It widely exists in chemical industry,papermaking,combustion and other industrial processes closely related to people’s livelihood.In the actual industrial process,due to the harsh industrial environment and the long time operation of the equipment,the system is prone to actuator or sensor fault,which not only affects the reliability of the system,but also may cause huge property loss and serious casualties.Therefore,it is of great theoretical and practical significance to research effective methods for fault diagnosis(FD)and fault-tolerant control(FTC)to ensure the safe and reliable operation of SDSs.With the development of technology and the upgrading of industries,the resource sharing,remote operation and control of SDSs are realized through networked control.However,due to the limitation of network bandwidth,a series of communication constraints such as delay and packet loss will occur during data transmission.Packet loss and delay lead to the loss or delayed delivery of the information containing fault characteristics,which brings new challenges to the research of FD and FTC for stochastic distribution systems.Therefore,the FD and FTC problems for stochastic distribution systems with communication constraints are studied in this thesis.The main contents can be summarized as follows:Firstly,the problem of FD and FTC for the nonlinear SDS with packet loss in feedback channel is studied.The random variable obeying Bernoulli distribution is used to describe the packet loss phenomenon in the feedback channel,and the data retention strategy is used to compensate the packet loss.On this basis,the adaptive observer is designed to estimate the unknown fault.By constructing the second-order sliding mode function,a second-order sliding mode fault-tolerant controller composed of the equivalent control term and the discontinuous term is designed to ensure that the system still has good tracking performance after the fault occurs.Thus,the problem of FD and FTC for a class of Lipschitz nonlinear SDSs with packet loss and the actuator fault is solved.Next,for the nonlinear uncertain SDS with packet loss in two-channel,the problem of FD and FTC when the actuator fault occurs in the system is studied under the condition of disturbance.The existence of packet loss in both the feedback channel and the forward channel makes the design of the FTC algorithm more complicated.Firstly,two independent Bernoulli random processes are introduced to describe the packet loss process in two-channel,and the latest data stored in the buffer is used to compensate the data loss.The augmented adaptive observer is designed to realize the simultaneous estimation of fault and disturbance.On this basis,a model predictive FTC algorithm is proposed to deal with the adverse effects of fault and disturbance on the tracking performance of the system.By comparing the root mean square error(RMSE)of fault estimation with and without packet loss compensation,the influence of packet loss on fault estimation is explored.Then,the problem of FD and FTC of the fuzzy SDS with two-channel constraints(packet loss and time delay)is studied,two cases of additive fault and multiplicative fault are considered respectively.Firstly,the static model of linear fuzzy logic system approximating the output PDF and the dynamic model with the additive fault based on T-S fuzzy model are established.The random packet loss and time-delay caused by the network are described in a unified framework,in which the lost data is compensated with the latest data received by the buffer.The sliding mode predictive fault-tolerant controller is designed to compensate the time delay in the forward channel through the preset control signal.The problem of FD and FTC of the system with the multiplicative fault is further studied.The dynamic model with the multiplicative fault based on T-S fuzzy model is established,and the above FD and FTC strategy is extended to the fuzzy system with the multiplicative fault.The research in this part provides a theoretical basis for the research on FD and FTC of nonlinear SDSs under two-channel constraints.Finally,a state feedback fault-tolerant tracking control strategy based on FD observer is proposed for the fuzzy SDS with packet loss,signal quantization and system noise.The signals from the sensor to the controller and from the controller to the actuator of the system are quantized by a new logarithmic-uniform quantizer,and packet losses are modeled by Bernoulli random processes.An adaptive sliding mode observer is designed to realize fault estimation in the presence of the noise disturbance.In order to ensure that the system still has good tracking performance after the fault occurs,a state feedback fault-tolerant tracking control scheme based on the parallel distributed compensation technology is proposed.The sufficient conditions for the existence of the controller gain matrix are given in the form of linear matrix inequality.