Risk Management of Weather-related Failures in Distribution Systems Based on Interpretable Extra-trees

• DOI: 10.35833/mpce.2022.000430
• Publication Date: 2023-11-23T01:48:12Z
• Authors: Ying Du; Yadong Liu; Yingjie Yan; Jian Fang; Xiuchen Jiang
• Keywords: Interpretability; Extreme Weather

Weather-related failures significantly challenge the reliability of distribution systems. To enhance risk management weather-related failures, an interpretable extra-trees based prediction model is proposed in this study. In model, interpretability successfully introduced to by analyzing and processing paths decision trees extra-trees. As a result, reflected following three respects: it can output importance, contribution, threshold weather variables at high risk. The importance help developing long-term prevention plan. contribution provides targeted operation maintenance advice for next period. critical further preventing risks. Compared with black-box machine learning models, over-comes application limitations. addition generating predicted levels, also provide more guidance information failures.