Evaluation of Optimization Method for Inspection Scheduling of Power Distribution Facilities Using Maintenance Data Accumulated by Power Utility

Power distribution companies must establish an optimal maintenance schedule that balances the reliability and efficiency of a number of widely scattered facilities. Accordingly, we have developed an optimization method for inspection scheduling by which optimal inspection schedules are planned using 1) a method to predict the inspection priority for each facility and 2) a method to establish an inspection schedule that minimizes the inspection cost. In this study, for step 1, a practical method using both inspection history and facility attributes is proposed. The method consists of two approaches: a relation analysis and a defect prediction. In the relation analysis, relations between facility defect and other maintenance data are analyzed using a multiple correlation analysis method on the basis of chi-square testing. And then, facility defect is predicted using the data items that correlated with the facility defect obtained from the relation analysis. A modified quantification method type II introducing biased posterior probability is proposed for the prediction. Furthermore, proposed method is demonstrated by utilizing maintenance data accumulated by a certain Japanese power utility company. Evaluation results show that inspection cost can be reduced by 36-58% compared with the conventional twice-every-two-years inspection while retaining reliability.