基于振电结合的高压断路器特征提取及分类方法研究

Study on the Signal Feature Extraction and Classification of High Voltage Circuit Breaker Based on Vibration Signal and Current Signal

为了对高压断路器操作机构的工作状态进行较准确的评估,提高高压断路器机构的运行稳定性,提出了一种基于振动信号与电流信号结合的高压断路器信号特征提取和分类方法。首先通过对高压断路器分合闸线圈电流信号和振动信号的机理分析,提出利用时间节点参数作为特征向量,然后采用曲线斜率方法提取电流信号时间参数,利用基于短时能量的双门限法提取振动事件的时间参数,将两者的参数作为模式识别的特征向量。最后通过支持向量机(Support Vector Machine, SVM)分类结果表明:线圈电流曲线与振动信号相结合能够准确而全面地反映操作机构的运行状况,利用SVM可以快速准确的判断操作机构的故障类型,对于断路器的故障诊断和检修维护具有重要的意义。

Based on vibration signal and current signal, this paper proposed a method of signal feature extraction and classification of high-voltage circuit breaker. The proposed method can accurately evaluate breaker’s operating conditions and improve its operation stability. Firstly, this paper analyzed the vibration signal and current signal of tripping and closing coils of CBs, advocating to take time-node parameter as characteristic vector. Then, this paper extracted the time parameter of current signal by curve slope and that of vibration signal by double-threshold method based on short-term energy. The two parameters were added as feature vectors for pattern recognition. The classification results of SVM indicated that the integration of vibration signal and current signal of tripping and closing coils precisely reflects the operation state of high voltage circuit breaker. In addition, SVM quickly tells the fault type of operating mechanism, contributing to fault diagnosis and maintenance of CBs.