The Method Study Of Liquid Level System's Sensor Fault Diagnosis And Fault Tolerance Control

With the continuous development of automated production,people's demand for product quality is also getting higher and higher.However,the system will lose effective control due to the failure during operation.Therefore,how to improve the reliability and stability of the control system has become an urgent problem need to study and solve.Considering that the actual industrial system is mostly complex and non-linear,it is difficult to obtain accurate model,and the sensor is prone to failure.This thesis takes the single-tank as the controlled object,based on data mining technology of knowledge discovery,a series of research on the fault diagnosis and fault-tolerant control method about liquid level sensor is carried out.And the main research work is as follows:(1)On the basis of understanding the research background and significance of the project,analyzing the current research status of fault diagnosis and fault-tolerant control technology at home and abroad,we take the sensor of the liquid level control system as the research object,and study the fault diagnosis and fault-tolerant control method.Based on OPC(OLE for Process Control)communication technology,the hardware in the loop experiment platform of PID controlled single liquid level system is built,and then the experiment of the sensor under normal and fault conditions is carried out by means of the experimental platform,the experimental data are collected and used to study the fault diagnosis method.(2)Considering it is difficult to establish precise model of complex nonlinear system,a new method of knowledge discovery based on data mining is proposed to establish mathematical model of liquid level sensor's fault diagnosis.In the first place,the historical output data of each sensor's fault type are analyzed and processed to extract the fault characteristics.Next,the fault detection is carried out by using residual knowledge,after detecting the fault of the sensor,the fault setting model is established by analyzing the relationship between the change rate of the liquid level and the fault value,which is used to determine the value of the fault.Then,the fault location model is established by using the previous fault features to distinguish the sensor fault types.Finally,the experimental results show that the proposed fault diagnosis method is accurate and effective.(3)Aiming at the result of the fault diagnosis,a fault-tolerant control method of series fault compensation model is proposed,Which is to compensate the error of sensor output and offset the fault.And the fault-tolerant control experiment is designed to verify that the liquid level system with fault compensation can run stably regardless of whether the fault occurs or not.(4)Based on the original experimental platform,the S-function is called to embed fault diagnosis and fault-tolerant control algorithm,and establish a liquid level system with real-time fault diagnosis and fault tolerance control.Then,the real-time fault diagnosis and fault-tolerant control simulation experiments are carried out on the normal no fault,additive and multiplicative faults of the sensor.The experimental results show that the method studied in this thesis can be effectively applied in the actual liquid level system,and achieve the purpose of real-time fault diagnosis and fault-tolerant control of the sensor.