基于稀疏表示和M-ELM的断路器故障诊断模型

发布时间：2018-04-27 14:16

本文选题：断路器 + 目标跟踪　；参考：《系统仿真学报》2017年11期

【摘要】：针对传统断路器机械特性评估方法的不足,构建了基于稀疏表示和M-ELM(Memetic-Extreme Learning Machine)的断路器故障诊断模型,该模型首先利用稀疏表示方法跟踪断路器分(合)闸过程中与动触头同步运动的连杆或主轴上辅助标志物运动的轨迹;然后,根据该轨迹获取断路器操动机构的行程时间曲线并据此计算断路器的分(合)闸速度等各种机械特性参数;最后,将动触头运动机械特性参数作为M-ELM的输入进行断路器的故障诊断。通过对12 k V真空断路器的测试实验证明了该模型诊断断路器状态的有效性和优越性。
[Abstract]:A fault diagnosis model of circuit breakers based on sparse representation and M-ELM(Memetic-Extreme Learning Machine is constructed to overcome the shortcomings of traditional methods for evaluating the mechanical characteristics of circuit breakers. Firstly, the sparse representation method is used to track the track of the auxiliary marker movement on the connecting rod or spindle which moves synchronously with the moving contact during the opening (closing) process of the circuit breaker, and then, According to the track, the travel time curve of the operating mechanism of the circuit breaker is obtained and the mechanical characteristic parameters such as the opening (closing) speed of the circuit breaker are calculated accordingly. Finally, The fault diagnosis of circuit breaker is carried out by using the moving contact mechanical characteristic parameter as the input of M-ELM. The experimental results of 12 kV vacuum circuit breaker show that the model is effective and superior in diagnosing the state of the circuit breaker.
【作者单位】：华北电力大学计算机系;华北电力大学电力工程系;国网河北省电力公司;
【基金】：中央高校基本科研业务费专项资金(2017MS156)
【分类号】：TM561

【相似文献】