摘要
针对数控机床故障诊断精度不高、实时性差等问题,提出一套基于声音增强技术的数控机床故障分析与维修方法。该方法通过多通道声音传感器采集机床运行声音信号,采用深度学习算法进行声音增强与特征提取,利用支持向量机分类器实现故障类型识别与故障严重程度评估。实证研究结果表明,该方法的故障识别准确率达到97.7%,平均诊断时间为18.7 s,维修效率达到89.6%,各项指标均优于行业标准要求,为数控机床智能化维护提供了有效的技术支撑。
Aiming at the problems of low accuracy and poor real-time diagnosis of numerical control machine tools,a set of fault analysis and maintenance methods of numerical control machine tools based on audio enhancement technology is proposed.In this method,multi-channel audio sensors are used to collect the operating audio signals of machine tools,and deep learning algorithm is used for audio enhancement and feature extraction,and support vector machine classifier is used to realize fault type identification and severity evaluation.The empirical research results show that the fault identification accuracy of this method reaches 97.7%,the average diagnosis time is 18.7 s,and the maintenance efficiency reaches 89.6%.All the indexes are better than the requirements of industry standards,which provides effective technical support for the intelligent maintenance of numerical control machine tools.
作者
李开艳
刘发江
张志
李吴荣
LI Kaiyan;LIU Fajiang;ZHANG Zhi;LI Wurong(School of Intelligent Manufacturing,Zhaotong Vocational College,Zhaotong 657000,China;School of Mathematics and Statistics,Zhaotong College,Zhaotong 657000,China)
出处
《电声技术》
2025年第9期198-200,共3页
Audio Engineering
基金
云南省教育厅科学研究基金项目(2024J1990)
2024年昭通职业学院校级课题“基于QBE理念的‘电工基础’课程教学改革探索与实践”。
关键词
声音增强技术
数控机床
故障分析
维修方法
深度学习
audio enhancement technology
numerical control machine tool
fault analysis
maintenance method
deep learning
