基于超声阵列传感器与遗传MUSIC的局放源波达方向估计

Source DOA Estimation Based on Ultrasonic Array Sensor and Genetic MUSIC

基于超声阵列传感器的局放源波达方向估计是局部放电准确定位的基础。传统MUSIC测向算法的精度和搜索成功率较低,据此提出一种基于超声阵列传感器与遗传MUSIC算法的局放源波达方向估计新方法。遗传MUSIC算法利用空间谱估计公式构造适应度函数,基于遗传迭代搜索的思想对方位角进行谱峰搜索,以获得全局最优解。通过建立信号模型、阵列模型开展了大量仿真实验,并采用寻优精度、搜索成功率和搜索时间等三个指标为依据,对经典MUSIC和遗传MUSIC的测向结果进行对比。结果表明,基于遗传MUSIC算法提高了搜索成功率,绝对精度相较传统MUSIC算法误差减小0.3°,可以获得更高精度的局放源波达方向角。

Partial discharge direction of arrival （ DOA） estimation based on ultrasonic array sensor is the basis for ac-curate positioning of the partial discharge. Considering the problem of low precision and searching success rate of tra-ditional MUSIC direction finding algorithm, the paper presents a new method based on ultrasonic array sensor and ge-netic MUSIC algorithm for partial discharge DOA estimation. The genetic MUSIC algorithm constructs fitness function with spatial spectrum estimation, and searches spectral peak of the azimuth based on the idea of genetic iterative searching for global optimal solution. This paper carried out a large number of simulation experiments through the es-tablishment of signal model and array model, then uses three indicators of searching precision, searching success rate and searching time as the basis to compare the direction finding results of classical MUSIC and genetic MUSIC. Re-sults show that compared with traditional MUSIC algorithm, the genetic MUSIC based algorithm improves the searching success rate and reduces the error by 0. 3＆#176; in absolute precision, therefore obtains higher accuracy of partial discharge DOA angle.