Research On Fault Diagnosis Of Wind Turbine Based On Adaptive Observer

Wind power is the world's fastest growing renewable energy,and with its development,the maintenance of wind turbines will be particularly important.In recent years,the occurrence of turbine faults is frequent,which may cause certain economic losses,damage or even casualties to the turbines.Therefore,in order to ensure the normal operation of the units,it is of great significance to carry out fault diagnosis research on them.This paper studies from the perspective of fault observer,designs an adaptive observer and applies it to the fault estimation of wind turbines,mainly including the following contents:1.This paper summarizes the development of fault diagnosis and current fault diagnosis methods,and elaborates the principle of fault diagnosis method based on observer.Furthermore,an adaptive state observer design method is proposed for the nonlinear system with unknown parameters in the state equation and dynamic output,which can estimate the system state effectively,so as to establish a theoretical basis for the follow-up study on the fault diagnosis of wind turbines.2.A fault estimation method based on adaptive observer is proposed for the fault sensor of drive system of wind turbine.Firstly,we analyze the model of each part of fan transmission system,and establish a comprehensive state space model of drive system.Then,based on its model we design a fault estimation observer and discuss the system fault estimability,at the same time,we use the residual signal to give a fault estimation algorithm of adaptive rate.When the system fails,the fault vector can be estimated according to the designed fault estimation method.According to the magnitude of the fault amplitude,which part of thesystem fails can be detected,and the form of fault generation can be diagnosed,which is conducive to fault maintenance.Finally,the simulation results show the accuracy and effectiveness of the proposed fault estimation method.3.Taking into account uncertain factors such as noise interference,for variable pitch system with actuator failure and sensor failure for wind turbine,a fault diagnosis method based on adaptive decoupling observer is proposed.Firstly,we establish the variable pitch system model,and the model uncertainty and noise interference are used as the unknown state input of the system.Then we design the adaptive decoupled observer,which through the decoupled processing technology the state equation of the system no longer contains the fault estimation vector,and the solution of the system fault estimation and the system state estimation can be separated.In this way,the residual signal which is not sensitive to disturbance factors and is sensitive to fault quantity is generated,at the same time,the threshold evaluation function of this residual signal is used to establish the fault decision logic,and the system fault detection is conducted under the condition of determining the stability of the system.The simulation results show that the designed decoupling observer can accurately and effectively detect and estimate the faults of the complex variable pitch system with uncertainties such as disturbance.