Wind Power Converter Reliability Assessment Considering Different Temperature Variation Influence Factors

Wind farms have evolved rapidly nowadays, the security and reliability running areessential, as one of the crucial device of wind energy conversion system, theperformance of wind power converter has a significant importance. Since the operationstates varie with the randomly changing wind speed, the failure rate figure is muchhigher than other traditional power converter machineries. The wind power converterused for direct-drive permanent magnet synchronous generator (PMSG) system isstudied, for which the reliability impacts factors, the calculation time scale and thereliability assessment model are further presented. All the achievements are useful forwind power converter accuretly reliability assessment and converter designoptimization.The main contents of this thesis are followings:â‘ Based on the power losses calucalation formulas and the thermal net circuit, thejunction temperature of power modules simulation model is constructed in the platformof MATLAB/Simulink, and the junction temperature varation characteristics have beenanalyised in the scenario of different wind speed. Firstly, the junction temperaturevariation charachteristics in constant wind speed are studied. Secondly, the impact ofwind speed component of ramp, gust and noise are compared respectively, andconsequently the relationship of wind speed accumulation time and junctiontemperature variation is concluded. Finally, junction temperature variations understochastic wind speed are simulated and analyised. All the above studies are dedicatedto the wind power converter reliability assessment model.â‘¡Considering the long time scale junction temperature variation, the turbinemechanical response time is investigated and is proposed to be used as the wind speedsampling time interval. Firstly, the sampling calculated results and the simulation resultsare compared to indicate the importance of wind speed sampling time interval. Secondly,the turbine mechanical response time is calculated under tiny wind speed step operationscenario, and the response time is proposed to be used as the time interval choosedreference. Finally, taken a real wind turbine as example, the relationship between windspeed and wind sampling time interval is calculated.â‘¢Considering different time scales of device junction temperature variations,corresponding failure rate calculation methodologies are presented. Firstly, in respect to long time scale junction temperature variation, based on the multi-state probabilityanalysis method, two dimension allocation factors that corresponding their effects onthe operation states of wind power converter are taken. Combining with the cyclinginformation extraction method of rainflow, the conformance failure rate calculatedmodel is proposed. Secondly, in respect to short time scale junction temperaturevariation, the power cycling and thermal cycling fitted curves are adopted to carried outa linearly cumulation damage law based failure rated calculation model. Finally,combining the above analysis, the wind power converter reliability assessemt model isproposed, based on which the operational parameters as well as heat sink designprarmeters adjustment strategies are preseted in reliability improvement perspectives.The above achievements have provided an assement method for wind powerconverter reliability, and will offer a technical support for conveter reliability designoptimization.