Analysis and prediction: A research on the operation and maintenance cost of power distribution network

Abstract

Operation and maintenance of electric distribution network is a critical component of the power grid enterprise investment. To optimize the investment strategy, it is necessary to estimate the exact cost of network operation and maintenance. However, the estimation involves various potential impact factors, categories of which including social, economic, policy, resources, etc. Furthermore, the mechanism among those factors is so complicated that can be hardly described using linear models. According to that, in this study we present an accurate cost predictive model with the combination of Grey Relational Analysis (GRA) and Artificial Neural Network. The main factors affecting the operation and maintenance cost of distribution network are extracted by the grey relational analysis and are collected as the input variables of artificial neural network model. The empirical test shows that the grey relational analysis can efficiently reduce the number of input variables of artificial neural network prediction model, which subsequently improve the predicting accuracy of artificial neural network model.