Similar-Density-Array-Based Equipment Outage Prediction Method for Distribution Network Considering Weather Factors

Predicting the failure risk of the equipment in distribution network helps to guide operation and maintenance decision-making, which is very useful to boost the resilience against unplanned power failure risks and improve the reliability of distribution network. In this paper, a Similar-Density-Array-Based equipment outage prediction method is proposed to solve the outage prediction problem. In this method, Similar-Density-Array-Based clustering algorithm is used to classify the predictable data and unpredictable data according to the load data and recorded outage data in the power network. Similar-Density-Array-Based equipment outage prediction method is proposed using the Logistic fast minimum error entropy algorithm to predict the failure risk of the equipment considering weather factors. The proposed method is validated in the real distribution dataset. The results proved that the proposed clustering algorithm is effective and the outage prediction method completes the failure risk prediction problem, obtaining good prediction result.