摘要
在复杂的通信电源电路中,由于信号的非线性和时变性,传统的短路识别方法往往难以达到理想的诊断效果。针对上述问题,研究基于谐波小波包分解的通信电源电路短路识别方法。通过谐波注入法采集谐波响应信号,通过小波包分解算法将谐波响应信号分解为若干个小波包系数,基于小波包系数计算能量特征值,根据能量特征值,利K-means聚类方法进行通信电源电路短路识别。结果表明,针对3种不同的电源短路故障类型,所研究方法的轮廓系数一直保持在0.9以上,说明该方法一直保持相对较高的准确性和稳定性。
In the complex communication power supply circuit,due to the nonlinear and temporal variation of the signal,the traditional short-circuit identification method is often difficult to achieve the ideal diagnostic effect.Therefore,the short-circuit identification method based on harmonic wavelet packet decomposition is studied.The harmonic response signals were collected by the harmonic injection method,the harmonic response signal is decomposed into several wavelet packet coefficients by the wavelet packet decomposition algorithm,the energy eigenvalues were calculated based on the wavelet packet coefficient.According to the energy characteristic value,the K-means clustering method identifies the short circuit of the communication power supply circuit.The results show that the contour coefficient of the studied method is kept above 0.9 for three different power short circuit fault types,indicating that the method has maintained relatively high accuracy and stability.
作者
王硕
WANG Shuo(Zhengzhou Shengda University,zhengzhou 451191,China)
出处
《传感器世界》
2025年第8期37-42,共6页
Sensor World
关键词
谐波信号
小波包分解
通信电源
电路短路
K-MEANS聚类
harmonic signal
wavelet packet decomposition
communication power supply
short circuit
K-means clustering
