基于布谷鸟搜索优化特征选择的轴承故障诊断

扫码查看

原文链接

NETL
NSTL
万方数据

中文摘要：针对高维特征导致的滚动轴承故障诊断精度低问题，提出一种基于改进二进制布谷鸟搜索的特征选择方法。首先，采用Hilbert-Huang变换技术提取电机信号的时频域信息，建立高维特征集合。其次，基于布谷鸟搜索算法提出一种基于互信息的特征加权初始化方法，通过计算特征互信息对原始特征集合进行相关性评估，达到快速去除无关特征的目的。同时，引入局部螺旋开发策略，并采用动态切换概率算子实现全局探索和局部开发的平衡，加快算法收敛。最后，采用KNN分类器对轴承状态进行诊断，实验研究表明，上述方法能有效提取更具价值的特征信息，识别精度高达99。05%，相比于其它同类方法诊断准确率高，性能更稳定。

外文标题：Bearing Fault Diagnosis Based on Cuckoo Search Optimization Feature Selection

外文摘要：Aiming at the problem of low accuracy of rolling bearing fault diagnosis caused by high dimensional features,a feature selection method based on improved binary cuckoo search is proposed.Firstly,the Hilbert-Huang transform technique was used to extract the time-frequency domain information of the motor signal and establish a high-dimensional feature set.Secondly,based on the cuckoo search algorithm,a feature weighting initialization method based on mutual information was proposed.By calculating the mutual information of features,the correlation e-valuation of the original feature set was carried out to achieve the purpose of quickly removing irrelevant features.At the same time,the local spiral development strategy was introduced,and the dynamic switching probability operator was used to achieve the balance between global exploration and local development to accelerate the convergence of the algorithm.Finally,the KNN classifier was used to diagnose the bearing state.The experimental results show that the method can effectively extract more valuable feature information,and the recognition accuracy is as high as 99.05%.Compared with other similar methods,the diagnosis accuracy is high and the performance is more stable.

外文关键词：

Feature selectionBinary cuckoo search algorithmFault diagnosisMutual information

作者：

辛艳、孙丽君、陈天飞、冯斌斌

展开 >

作者单位：

河南工业大学粮食信息处理与控制教育部重点实验室,河南郑州 450001

河南工业大学郑州市机器感知与智能系统重点实验室,河南郑州 450001

河南工业大学电气工程学院,河南郑州 450001

关键词：

特征选择二进制布谷鸟搜索算法故障诊断互信息

基金：

国家自然科学基金国家自然科学基金中原科技创新领军人才资助项目中原英才计划中原青年拔尖资助项目河南省优秀青年科学基金河南省高校科技创新人才支持计划河南省科技攻关项目河南省科技攻关项目河南省科技攻关项目郑州市协同创新专项河南工业大学自科创新基金支持计划

项目编号：

6217312761973104224200510008[2023]1121230041003621HASTIT02921210221016921210221008622210232020921ZZXTCX062020ZKCJ06

出版年：

2024

计算机仿真

中国航天科工集团公司第十七研究所

计算机仿真

CSTPCD

影响因子：0.518

ISSN：1006-9348

年,卷(期)：2024.41(7)

参考文献量5