基于MPSO-MLE的变电站设备异常声源定位方法  

Method of Locating Abnormal Acoustic Source of Substation Equipment Based on MPSO-MLE

作　　者：张瑶 罗林根[1] 王辉[1] 盛戈皞[1] 江秀臣[1] ZHANG Yao;LUO Lingen;WANG Hui;SHENG Gehao;JIANG Xiuchen(Department of Electrical Engineering,Shanghai Jiao Tong University,Shanghai 200240,China)

机构地区：[1]上海交通大学电气工程系,上海200240

出　　处：《高电压技术》2020年第9期3145-3153,共9页High Voltage Engineering

摘　　要：变电站电力设备的可听声信号中包含着丰富的振动、放电等信息,对电力设备声音信号的定位及分析是实现设备运行状态评估及故障诊断的有效手段之一。针对变电站全站设备,实现对设备异常声音信号的快速巡检及定位可有效地提升设备故障检测效率。为此基于非接触式声音传感器阵列,从统计分析的角度出发提出了利用极大似然估计法对全站电力设备声源进行定位的方法,并利用变异的粒子群优化算法降低运算量,实现了低信噪比环境下的设备故障声源高效精确定位。仿真分析及实验室测试表明,所提出的设备异常声源定位方法及系统在低信噪比环境中其定位误差比MUSIC方法降低约30%,可快速、有效地定位设备异常声源,为后续设备故障精确定位及诊断提供参考。The audible sound signal of power equipment in substation contains abundant information such as vibration and discharge.The localization and analysis of sound signal of power equipment is one of the effective means to achieve the operation state evaluation and fault diagnosis of the equipment.For the whole station equipment of the substation,how to realize fast inspection and localization of abnormal sound signal of equipment can effectively improve the efficiency of equipment fault detection.Based on the non-contact sound sensor array,this paper proposes a method of locating the acoustic source of the electrical equipment of the whole station by using the maximum likelihood estimation method from the perspective of statistical analysis.Moreover,a mutant particle swarm optimization is adopted to reduce the amount of computation,thus the accurate location of the equipment fault acoustic source can be realized in the low Signal-Noise Ratio(SNR)environment.Simulation analysis and laboratory experiments show that the locating accuracy of the proposed method and system is greatly improved compared with the traditional method.The positioning error is reduced by about 30%compared with the MUSIC method under low SNR.The method can be adopted to quickly and effectively locate the abnormal acoustic source of the equipment and provide references for the accurate positioning and diagnosis of subsequent equipment faults.

关 键 词：电力设备 可听声 最大似然估计 变异粒子群算法 定位 

分 类 号：TM63[电气工程—电力系统及自动化]