An approach for Condition Based Maintenance strategy optimization oriented to multi-source data
Youyuan Wang,Hang Liu,Jiangang Bi,Feng Wang,Chunyu Yan,Taiyun Zhu
DOI: https://doi.org/10.1007/s10586-016-0626-1
2016-01-01
Cluster Computing
Abstract:Condition Based Maintenance (CBM) scheduling for power equipment is based on some types of data in power system, such as the health condition of components, the load level of the substation and Life Cycle Cost (LCC), etc. Due to the lack of necessary data, traditional CBM strategy optimization models are usually established without enough data, and the optimization results are usually difficult to obtain the enough accuracy. With the development of Chinese power industry, the data volume in power system is increasing dramatically in recent years, and the power industry is undoubtedly entering the era of Big Data. To handle and store these increasing data, more and more data management systems have been developed for different application. For example, Energy Management System, Production Management System and Management Information System have been widely applied. The multi-source data in these systems gives an opportunity to improve the accuracy and rationality for CBM optimization results. Thus, it is of great significance to make full use of these data in maintenance strategy optimization process. This paper presents a new approach to optimize the CBM strategy for components in the substation. After analyzing the interconnection relations of different types of components, the substation is divided into different maintenance units, the components in the same unit can be maintained together. To quantitatively evaluate the reliability of the components before and after repaired, two failure rate calculation models based on Health Index (HI) and age reduction factor are established respectively, and all the alternative maintenance strategies for the abnormal components are proposed based on the location and severity of the faults or defects of the components. According to the theory of LCC, the CBM optimization model is established when determines the minimum total cost as the optimization goal during the maintenance period. The total cost consists of the reparation cost, the interruption cost and the maintenance cost. Finally, an application example in a 220 kV substation is proposed, and these multi-source data are fully applied in the optimization process. The calculation results indicate that these types of data have benefit to improve the accuracy and rationality of the optimization results, and the strategy after optimized can obviously improve both the power supply reliability and the economy of the substation.