基于小波变换和随机森林算法的高压断路器机构机械特性监测技术研究

Research on Monitoring Technology of Mechanical Characteristics of Operating Mechanism for High-voltage Circuit Breaker Based on Wavelet Transform and Random Forest Algorithm

针对高压断路器弹簧机构机械特性监测数据类型单一、机械特性状态识别准确率低等问题,以平高集团有限公司CT601-2型弹簧操动机构为基础,提出一种小波变换与随机森林算法结合的高压断路器弹簧机构机械特性监测与状态识别方法。文中首先详细分析了机构机械故障类型,其次通过小波变换对电流、行程信号进行预处理,最后采用随机森林算法实现断路器机械状态识别。通过大量的仿真与试验表明,该高压断路器弹簧机构机械特性状态识别方法识别准确率达到94.67%,具备准确率高、实用性强等优点,可为高压断路器机构机械特性方面的研究提供参考,研究意义重大。

In view of such issues as single monitoring data and low condition identification accuracy rate of mechanical characteristic of spring operating mechanism for the high voltage circuit breaker,type CT601-2 spring operating mechanism from Pinggao Group Co.,Ltd. is taken as the base,the monitoring and condition identification method of spring operating mechanism for high voltage circuit breaker,a combination of wavelet transform and random forest algorithm,is proposed. First,the type of mechanical fault of the operating mechanism is analyzed in detail. Then,the current and stroke signals are preprocessed by wavelet transform. Finally,the random forest algorithm is used to achieve the mechanical condition identification of the circuit breaker. It is shown by a large number of simulations and tests that the identification accuracy rate of the condition identification method of the mechanical characteristics of the spring mechanism for high-voltage circuit breaker is up to 94.67%,which has such advantages as high accuracy and strong practicability,and can provide a reference for the research on the mechanical characteristics of operating mechanism for high-voltage circuit breaker and the research is of great significance.