基于天气因素和XGBoost算法的配电线路故障停电预测

恶劣天气是引起电力系统配电设备故障的重要因素,导致配电线路故障停电事件增多。以国网某地市供电公司10 kV配电线路故障停电数据为样本,采用XGBoost算法建立了考虑天气因素的区域线路停电数量预测模型。结果表明:模型虽然对实际停电线路较多的样本预测结果偏差较高,但能够较准确地反映实际停电变化趋势;在超高温、暴雨等恶劣天气条件下,配电线路停电概率较高。这个研究成果对恶劣天气下配网运维检修工作具有较高的实用价值。

Failure Prediction of Distribution Line Based on Weather Factors and XGBoost Algorithm

Severe weather is an important factor causing the failure of power distribution equipment in power system.In the case study of the power failure data of 10 kV distribution line in a municipal power supply company of State Grid,the prediction model of the number of regional line power failures considering weather factors is established by adopting XGBoost algorithm.The results show that the model can accurately reflect the variation trend of actual power failure although it has a relatively high deviation from the predicted results of samples with more power failure lines.Under severe weather conditions such as ultra-high temperature and rainstorm,the probability of power failure of distribution lines is relatively high.This research result is of high practical value to the operation and maintenance of distribution network under severe weather.