一种基于改进多任务联合稀疏表示的道岔故障检测算法
A Turnout Fault Detection Algorithm Based on Improved Multi-task Joint Sparse Representation
王培东1
作者信息
- 1. 广东机电职业技术学院,广东 广州 510550
- 折叠
摘要
为提高道岔维护人员人工检查的工作效率,研究准确率更高的智能道岔故障检测算法,提出一种基于改进多任务联合稀疏表示分类的道岔故障检测算法.该算法以道岔转辙机设备的功率曲线为基础数据,应用小波包分解算法提取电流曲线的频域特征向量,并融合曲线时域特征向量,选用多任务联合稀疏表示分类算法,辅以新型动态时间规整算法衍生核函数进行故障建模.经仿真实验,对收集的实际数据进行检测,模型正确率可达100%,验证了该算法的有效性,实现了智能道岔故障检测.
Abstract
To improve the efficiency of manual inspection by turnout maintenance personnel,a more accurate intelli-gent turnout fault detection algorithm is studied,and a turnout fault detection algorithm based on improved multi task joint sparse representation classification is proposed.This algorithm is based on the power curve of the turnout machine equipment,applies the wavelet packet decomposition algorithm to extract the fre-quency domain feature vector of the current curve,and fuses the time domain feature vector of the curve.A multi task joint sparse representation classification algorithm is selected for fault modeling,supplemented by a new kernel function derived from dynamic time warping algorithm.Through simulation experiments,the collected actual data was tested,and the model accuracy reached 100%,verifying the effectiveness of the algorithm and achieving intelligent switch fault detection.
关键词
故障检测/道岔/多任务联合稀疏表示/核函数/小波包Key words
fault detection/turnout/multi task joint sparse representation/kernel function/wavelet packet引用本文复制引用
出版年
2024
NETL
NSTL
万方数据