摘要
针对极限学习机连接权重和阈值随机选取存在的很大不确定性,提出将麻雀搜索算法与极限学习机结合搭建故障诊断模型(FASSA-ELM)。在原麻雀搜索算法的基础上引入Sine混沌映射优化初始种群,结合萤火虫算法(FA)对麻雀种群的位置以及最优解位置进行扰动更新,将改进后的麻雀搜索算法用于优化极限学习机的权值和阈值。采用集合经验模态方法提取出高压断路器分合闸线圈电流中的故障特征量,对断路器故障特征的仿真分析表明,FASSA-ELM的诊断准确率达到了100%,将训练样本和测试样本互换后该模型诊断准确率为84.5%,与其他三种模型相比,该方法具有更高的准确率和更好的稳定性。
This paper combined an improved sparrow search algorithm with the extreme learning machine to construct a fault diagnosis model (FASSA-ELM) for solving the uncertainty of the weight of connection of the extreme learning machine and the random threshold selection.This study introduced the Sine chaotic map based on the original sparrow search algorithm to optimize the initial population.It is combined with the firefly algorithm (FA) to disturb and update the position of the sparrow population and the optimal solution position.Moreover,it used the improved sparrow search algorithm to optimize the weights and thresholds of the extreme learning machine.This research employed the ensemble empirical mode method to extract the fault feature quantity of the switching coil current of the high voltage circuit breaker and conducted a simulation of the circuit breaker fault characteristic.The analysis results show that the diagnostic accuracy of the FASSA-ELM gets up to 100%.However,if exchanging the training sample for the test sample,the diagnostic accuracy of the model is 84.5%.Compared with the other 3 models,this method has higher accuracy and better stability.
作者
张莲
贾浩
张尚德
赵梦琪
赵娜
黄伟
ZHANG Lian;JIA Hao;ZHANG Shang-de;ZHAO Meng-qi;ZHAO Na;HUANG Wei(Chongqing Energy Internet Engineering Technology Research Center,Chongqing 400054,China;School of Electrical and Electronic Engineering,Chongqing University of Technology,Chongqing 400054,China)
出处
《电工电气》
2022年第10期50-56,共7页
Electrotechnics Electric
关键词
断路器
极限学习机
故障诊断
分合闸线圈电流
circuit breaker
extreme learning machine
fault diagnosis
switching coil current
