Application of signal sparse decomposition in the detection of partial discharge by ultrasonic array method

The detection of partial discharge of power transformer is important for the safe operation of a power system. The direction of arrival is an important part of ultrasonic array detection. Classic DOA estimation algorithms based on subspace analysis do not perform well under the conditions of low SNR and small snapshot numbers. The problem with using an improved algorithm is that it requires a huge amount of calculation. In this paper sparse signal decomposition is applied to spatial spectrum estimation. The authors derived the model of array signal direction finding by the use of matching pursuit algorithm, setting up overcomplete dictionaries for different types of sensor array and simulating different types of sensor array receiving signals. Simulations and experiments show that direction finding based on MP decomposition method performs better than using the MUSIC algorithm under the conditions of low SNR and small snapshot numbers, and that the influence of array types is less.