Research On Fault Diagnosis And Fault Tolerance Control Methods Of Control System Based On Model

With the rapid development of science and technology,modern industrial equipment has developed toward large-scale,intelligent,integrated,complicated and automated.Thus,higher requirements have been put forward for the equipment's safety and stability.If one part of the system in fault,it will cause the whole system to run in error,or even collapse and then result in serious economic losses or significant casualties.In order to prevent the fault influence and ensure safe and reliable operation of the system,these faults must be detected and isolated timely and effectively.From the perspective of energy flow,this paper discusses model-based fault detection,location,and fault-tolerant control methods based on bond graph theory.Firstly,the controllability and observability analysis method of hybrid system based on bond graph theory is described.This method uses the basic principle of bond graph theory to analyze the controllability and observability of the system structure.A lot of examples are used to verify the validity and accuracy of the method.Secondly,considering the complexity of the control system,a signed directed graph model is proposed based on the causal relationship between the key elements in the bond graph.The qualitative redundancy relationship is generated according to the qualitative state among the model variables.Then the qualitative fault features of the system are extracted,and the source of fault is located by analyzing the status of the related components and the observed abnormal behavior.Compared with the analytic redundancy relationship fault diagnosis algorithm based on bond graph,this method reduces the complexity of the system model,the number of analytic redundancy relationship and the amount of calculation.Finally,in order to improve the robustness of the control system,a robust diagnostic observer is designed,which combines linear fractional bond graphs and proportional integral observer to achieve robust fault diagnosis and fault estimation,and then adaptive threshold is used to evaluate residual.The observer can track the dynamic behavior of system variables in real time,adjust and update the thresholdeffectively,reduce the false alarm rate and false negative rate,and improve the accuracy of fault diagnosis.Then,an active fault-tolerant control algorithm based on state estimation and residual feedback is proposed to ensure the performance index of the system works within or below the scope when the system component in fault.