基于多特征评估与 XGBoost 的高压断路器故障诊断

High-voltage circuit breakers are key equipment in the power system and their reliable operation is an important guarantee for maintaining the security of the power grid. Under the construction of smart grid，the fault diagnosis research of high-voltage circuit breakers has received extensive attention.In this paper，the mechanical vibration signals of high-voltage circuit breakers under different working conditions is researched and a kind of fault diagnosis model based on multi-feature assessment and XGBoost is proposed. The multi-dimensional feature quantity of the vibration signal from the time domain，frequency domain and time-frequency domain is extracted. The method of fuse feature importance is adopted and combined with the XGBoost model to assess and filter the multi-dimensional feature and remove the redundant feature quantity.At the same time，the Bayesian optimization algorithm is adopted to optimize the parameters in the XGBoost model，which improves the accuracy of classification. The results show that the diagnostic method based on multi-feature evaluation and XGBoost has a high accuracy rate and can effectively achieve accurate classification of mechanical vibration signals of high-voltage circuit breakers.