摘要
根据输电线断股引起的辐射场的变化和巡检机器人运行环境状况,选用热释电远红外传感器对输电线断股信号进行检测。在数据处理时,针对断股信号的非平稳性质和小波基的时频特征,选用db4小波基对断股信号进行了6层小波分解。为解决单纯采用小波变换带来的诸如分解信息量大,故障特征值数目多、提取需要人工干预,难以实现在线检测与诊断的问题,本文引入了小波能量熵的概念,从而使描述输电线断股故障的特征值数目大大减少,并且易于计算机自动提取,利用较小规模的BP网络就实现了巡检机器人对铝绞导线断股损伤的在线精确诊断。在线断股检测实验证明了该方案的正确性。
Based on the radiation field changes caused by broken strand signal of transmission line and the operating situation of inspection robot, the pyroelectric far-infrared sensor that inspects broken strand signal of transmission line is selected. During data processing, broken signal is decomposed by six level wavelet using db4 wavelet base, aiming at the non-stationary nature of strand broken signal and the time-frequency character of wavelet bases. In order to solve problems caused by simply adopting wavelet transform, such as more decomposition information, too many eigenvalues and hard to realizing online inspection which needs manual intervention, the wavelet energy entropy is introduced in the paper. So that the eigenvalues of broken strand fault of transmission line are greatly reduced and easy to automatically extract by computer. When smaller BP network is adopted, the online precise diagnosis to the broken strand of aluminum conductor by inspection robot is realized. The experiment of online detection of broken strand shows the validity of the new method.
出处
《电工技术学报》
EI
CSCD
北大核心
2010年第6期185-191,共7页
Transactions of China Electrotechnical Society
基金
国家"863"计划资助项目(2002AA420110-4)
关键词
输电线
巡检机器人
小波分析
熵
断股检测
Power transmission line
inspection robot
wavelet analyse
entropy
detection of broken strand