The Fault Diagnosis System On Wind Turbines Based On Data Mining Technology

This paper is based on signal analysis and processing technology, data miningtechnology and database technology,to deeply research the theoretical basis ofcomplex Wind Turbine Generator System(WTGS) fault diagnosis method and itsapplication .This research can ensure WTGS safe and efficient to run, to improvemanufacturing efficiency and management level, to restore fault equipment and saveeconomic loss as a result of fault,to improve economic benefit in society。Large WTGS has complex structure and huge fault type, so acquisition faultsignal and data must be stored in the Database。The more accumulated equipmentmonitoring data, the more important information behind the surge data。These massdata can be used to predict and diagnose to the unit operation state, life cycle, failureof the WTGS. Firstly,this paper introduces the status quo of Wind Power, and thenecessity of WTGS fault diagnosis, introduces the composition, working principle ofthe WTGS, and the various typical faults in the parts of the WTGS; Then focuses onthe Rotor system for fault mechanism analysis, and simulates subsystem of the faultof the signal processing, and extracts of each type of fault signal characteristic;According to these fault data characteristic, design a fault Database including faultknowledge base, fault example base and fault rule base. Putted forward, the fault ruleis built on the basis of fault knowledge base and example base, that using the decisiontree algorithm of data mining, and is preserved into the fault Database. Through theextraction of rules, using the test set, to get the validation of the rules, and ensure theaccuracy of data mining technology in the application of WTGS.This system is in the flat of the Windows Server 2003, I apply DBMS of theSQL Server 2005, the decision tree based on MATLAB toolbox, to constructs rulesare extracted from a decision tree. For WTG several typical fault, I can make muchmore accurate judgement, has some use value. So as to reduce production and testingcosts, increase production stability which has played a positive role.