Sensor Fault Diagnosis And Fault Tolerant Control For Non-Gaussian Stochastic Distribution Systems

Improving the safety and reliability of the system in operation has become an urgent need in the fields of aviation,aerospace,high-speed rail,industrial process,etc.once an accident occurs in such a system,it may cause huge property loss or even serious personal injury.Fault diagnosis and fault-tolerant control technology play an important role in improving the reliability and safety of dynamic system.Therefore,the research on fault diagnosis and fault-tolerant control technology is of great theoretical significance and practical value.Stochastic distribution control systems have wide application background in agriculture,chemical industry,mineral industry and other industries,such as crystallization process,emulsion polymerization,coating,pigment,latex paint manufacturing,grinding process,papermaking process,chemical reaction process,boiler combustion and flame control process.In the above practical industrial systems,there are many production sites accompanied by high temperature,high pressure,toxic and other environments.The research of fault diagnosis and fault-tolerant control algorithm for this kind of system is helpful to improve the safety and reliability of the system,avoid personal injury and property loss.In this thesis,sensor fault diagnosis and fault-tolerant control schemes for stochastic distribution systems are studied in depth,and a series of research results are obtained.The main work and contributions of this paper are shown as follows:(1)The first part mainly focuses on sensor fault diagnosis and fault-tolerant control of stochastic distributed system,and uses learning observer to diagnose sensor fault.Fault estimation information is used to compensate the faults,and sliding mode control algorithm is used to track the expected probability density function(PDF)of the system output.The Lyapunov stability theorem is used to prove the stability of the observation error dynamic system and the whole control process,respectively.Finally,the mathematical model of industrial control is established and numerical simulation is carried out by computer,which further verifies the effectiveness of the algorithm.(2)The second part mainly focuses on the sensor fault diagnosis and fault-tolerant control of stochastic distribution systems with random delays and disturbances.First,the original system is transformed into an equivalent system by using Laplace transform,and the sensor fault is diagnosed by learning observer.The fault estimation information is used to compensate fault,and the PI control algorithm based on equivalent system design makes the PDF of the system output track desired PDF.The Lyapunov stability theorem is used to prove the stability of the observation error dynamic system and the whole control process respectively.Finally,the numerical simulation is carried out with the mathematical model established by the actual industrial control,which further verifies the effectiveness of the algorithm.(3)The third part mainly focuses on the fault diagnosis and fault tolerant control for nonlinear stochastic distribution systems subject to the sensor fault and actuator fault simultaneously.A new fault diagnosis and fault tolerant control algorithm based on fuzzy modeling is proposed for a class of complex stochastic distribution processes.The proposed framework includes linear fuzzy logic system(FLS)?Takagi Sugeno(T-S)fuzzy models?sensor and actuator faults.A fuzzy fault diagnosis observer is designed to estimate the fault information,and a fuzzy controller is designed to meet the control requirements,which including stability and the performance of tracking.Numerical simulation with computer shows the correctness and validity of the theoretical results.