基于信号稀疏分解的局放超声阵列定位法研究

Study on partial discharge ultrasonic array sensor location method based on the sparse signal decomposition

局部放电定位检测是电气设备维修的重要基础。传统超声阵列定位方法中,存在对局放源的位置测向误差较大以及双阵列测向交叉定位原理会出现"异面不相交"的情况,这些都会降低定位的精度。据此,研究了基于圆形超声阵列传感器与信号稀疏分解理论的局放源定位方法。首先研制了相同阵元数下,声学性能优于线形和方形的圆环形阵列;然后将信号稀疏分解理论应用到对局放源的波达方向(DOA)估计当中,并且采用多阵列测向交叉定位及全局搜索的方法实现对局放源定位;最后对该方法进行仿真和实验研究,结果验证了该定位方法的有效性。

Detection of partial discharge（ PD） location is the important basis for the maintenance of electrical equipment. For traditional ultrasonic array location methods,the direction finding error is large,and＂double arrays crossing location principle＂will appear in the situation that two different surface straight lines do not intersect,which will reduce the accuracy of positioning. Accordingly,the research on PD source localization method is carried out based on circular ultrasonic array sensor and sparse signal decomposition theory. Firstly,the circular ultrasonic array sensor is developed,whose acoustic performance is better than the linear array and square array in the same number of array elements. Then sparse signal decomposition theory is applied to the estimation of direction of arrival（ DOA）,and a multi-array direction finding crossing location method with global search is used to achieve the PD source localization. Finally,simulation and experimental research is carried out,and results prove the validity of the method.