Font Size: a A A

Research Of Wind Turbine Fault Features And Based On Alarm Fault Diagnosis

Posted on:2018-12-10Degree:MasterType:Thesis
Country:ChinaCandidate:D LiFull Text:PDF
GTID:2322330512978408Subject:Power engineering
Abstract/Summary:PDF Full Text Request
With the development of wind turbine(WT)of scale and technology,improving the reliability of WT will become the inevitable requirements and development trend of wind-powe field in China.In order to improvie the reliability of WT,this paper is mainly about fault features extraction in different control strategies and fault diagnosis based on SCADA alarm of WT.In order to compare the WT fault features in different control strategies,This paper uses MATLAB Simulink to establish WT simulations in optimum tip-speed ratio control and optimum torque control.The transmission efficiency reduction caused by the gearbox fault is the failure mode.Respectively extract power,rotor torque and generator torque,which is under health status and under gearbox transmission efficiency 90%,80%.Next,compare with the fault features and analysis results,to prepare for the research about alarm fault diagnosis.As a part of SCADA data,alarm is often not effectively used.This paper compares the traditional method of high frequency signals fault diagnosis,and proposes a fault diagnosis method whitch is based on low frequency signal SCADA alarm.Alarm is not equal to fault,due to WT is a complex system and the sub components of the system is in close contact,.when a fault occurs,the mutual influence is between the each alarm,and it will lead to the occurrence of mass alarm.Therefore,this paper through the fuzzy decision-making method to research the root alram,and you can find out the root alarm in a large number of alarm.However,fault has no contact with alarm,as a result it have no way to make maintenance workers in advance formulate a maintenance scheme in advance.Therefore,this paper proposes a fault diagnosis method of DS evidence theory.At first,we consider the maintenance of fault records as the identification framework.and use the maintenance and alarm through the Bias method to extract basic probability assignment(BPA),Then we use the improved DS evidence theory to diagnose fault.Alarm is contacted with fault at this time.The maintenance workers can diagnose fault with alarm,so then it can guide the maintenance workers i to make the plan in time,and achieve saving time and cost.Finally,this paper introduce GUI MATLAB programming interface with DS evidence theory fault diagnosis method,and have a horizontal verification to this method.It prove this method has feasibility,validity and accuracy.Through the research of this paper,it povide adequate theoretical basis for the the reliability improvement of WT in the future.
Keywords/Search Tags:Wind turbine, different control strategies, fault features, SCADA alarm, DS evidence theory, fault diagnosis, reliability
PDF Full Text Request
Related items