异响声纹特征模拟下的220kVGIS异响声学成像识别方法

220kVGIS abnormal sound acoustic imaging recognition method based on abnormal sound pattern feature simulation

为了准确识别220kVGIS异响,为有效诊断220kVGIS故障,为保障电力系统安全稳定运行提供可靠依据与参考,研究异响声纹特征模拟下的220kVGIS异响声学成像识别方法。采用多通道传输的数字麦克风阵列声纹特征模拟信号采集技术,采集220kVGIS声纹特征模拟信号,基于DAMAS反卷积波速形成算法,利用常规波速形成算法输出的有效声模拟信号数据,结合220kVGIS声源分布状况以及数字麦克风阵列声音传播函数间的卷积关系,建立并求解线性方程,求解220kVGIS异响声学成像识别结果。实验结果表明:该方法可实现220kVGIS异响声学成像识别,识别准确性较高,能够获取较为准确的异响位置信息,满足实际工作需要。

In order to accurately identify the abnormal sound of 220kVGIS,provide reliable basis and reference for effectively diagnosing 220kVGIS faults and ensuring the safe and stable operation of power system,the acoustic imaging recognition method of 220kVGIS abnormal sound under the simulation of abnormal sound pattern characteristics is studied.Adopting the multi-channel transmission digital microphone array voiceprint feature analog signal acquisition technology,the 220kVGIS voiceprint feature analog signal is collected.Based on the Damas deconvolution wave velocity formation algorithm,the effective acoustic analog signal data output by the conventional wave velocity formation algorithm is used,combined with the 220kVGIS sound source distribution and the convolution relationship between the sound propagation functions of the digital microphone array,the linear equation is established and solved,Solve the recognition results of 220kVGIS abnormal sound acoustic imaging.The experimental results show that this method can realize 220kVGIS abnormal sound acoustic imaging recognition,with high recognition accuracy,and can obtain more accurate abnormal sound location information,which can meet the needs of practical work.