基于声音识别的电网变电站设备故障检测方法
Fault Detection Method for Power Grid Substation Equipment Based on Voice Recognition
翁杰 1陈周 1陈胜泉1
作者信息
- 1. 国网湖北省电力有限公司超高压公司,湖北 武汉 430000
- 折叠
摘要
针对电网变电站设备故障检测中传统方法的局限性,提出一种基于声音识别的设备故障检测方法.该方法通过高精度麦克风阵列采集变电站设备的运行声音,并对音频信号进行预处理、特征提取与机器学习模型训练,最终实现故障的准确识别.实验证明,该方法在多种故障类型的检测中表现出极高的准确率,特别是在变压器过热故障识别中,准确率高达 99.8%.该方法为电网变电站设备的智能故障检测提供了一种有效的解决方案.
Abstract
Aiming at the limitations of traditional methods for equipment fault detection in power grid substations,a method for equipment fault detection based on voice recognition is proposed.In this method,the running sound of substation equipment is collected by high-precision microphone array,and the audio signal is preprocessed,feature extracted and trained by machine learning model,and finally the fault is accurately identified.Experiments show that this method has a very high accuracy in the detection of various fault types,especially in the identification of transformer overheating fault,the accuracy is as high as 99.8%.This method provides an effective solution for intelligent fault detection of power grid substation equipment.
关键词
声音识别/电网变电站设备/故障检测Key words
voice recognition/power grid substation equipment/fault detection引用本文复制引用
出版年
2024
NETL
NSTL
万方数据