Study On The Maintenance Strategy Of Wind Turbine Drive Train

Wind turbine drive train consist of gearbox, main shaft, rolling bearing andother rotating parts, which is an important part of the energy transfer between windturbines and generators. Due to the complexity of its load, running in harshenvironment, coupled with inadequate design and other reasons, the failure rate ofequipment is high. If there is problem, the maintenance is complex, and the cost ofmaintenance is high. Therefore, it is important that providing for the conditionmonitoring and making the optimal maintenance strategy to ensure reliableoperation and lower maintenance cost. The key of the condition-based maintenanceis to establish a more precise correspondence between the online monitoring dataand the degree of equipment. In this paper, the proportional intensity ofmaintenance model for equipment running status is build, the equipment failure ratefunction is determined, and then the optimal maintenance strategy is proposed.Finally, through the simulation example using the actual failure data on theMATLAB simulation data platform, the correctness and rationality of the model areproved. The key of condition-based maintenance strategy study include thepretreatment of online monitoring data pretreatment, modeling, the estimation ofmodel parameter, the determination of maintenance threshold, the formulation ofmaintenance strategy and so on.(1) Since the structure of wind turbine drive train is complex and there aremany points, the obtained online monitoring data is numerous. There is strongcorrelation between them, while the proportional intensity of maintenance modelrequires independent covariates reflect the effects of the historical maintenance dataand online monitoring data to equipment health status. In this paper, themathematical statistics software SPSS1.7is carried on analysis the correlation, andprincipal component analysis is used to extract the independent covariate as thebasis of proportional intensity model.(2) Currently, most of the condition based maintenance model is focus onperfect maintenance, but the imperfect maintenance is used in practice since thecost of perfect maintenance is expensive, which make the system status betweeninstalling new equipment and the status before maintenance. Considering the affection of the maintenance effect and frequency to the equipment status, the paperintroduces several maintenance types, such as perfect maintenance, minimummaintenance and imperfect maintenance, and their influence to the equipment status,and puts an emphasis on the imperfect maintenance model. The accuracy ofparameter estimation is a key factor affecting the rationality of the model, this paperapplied the mathematical statistical software SAS to study the robustness of themodel in terms of the sample size censoring rate, and proposed the scope of themodel.(3) The key of the optimize maintenance strategy is the acquisition ofmaintenance threshold. In traditional condition based maintenance, the monitorvalue is compared with the set value, and the maintenance is made when the valueis beyond the range, not considering the model of reliability and maintenance cost.In this paper, taking the equipment reliability and maintenance cost as theoptimization goal, then it propose an approach based on physical programming todeal with the tradeoff between the minimizing maintenance costs and maximizingreliability to determine the optimal threshold value of maintenance. he maintenancecost of the life cycle is reduce on the premise of guarantee the reliability. Finally,the wind turbine drive train component is analysis in form of maintenance decisiondiagram, and the correctness and rationality of the model are proved.