Equipment Fault Acoustic Source Direction of Arrival Estimation with Microphone Arrays Using SRP-PHAT Method

Abstract

The localization and detection of equipment fault in air-insulated substations is important for the electrical equipment statue analysis. The equipment fault detection method based on acoustic signal has been widely studied due to its simplicity and ease of application. In this paper, a novel acoustic equipment fault localization method based on directional of arrival (DOA) estimation is proposed. The main idea of design is to relate the position of the fault source to the established steered response power (SRP) map. Because the closer the position is to the fault source, the higher the SRP amplitude. Thus, by performing peak search of the spectrum is performed to get the directional angle of the fault source acoustic signal. Afterwards, the simulated fault source is moved to another place to repeat the estimation method. The double exponential model for acoustic signal is used for simulation tests and actual estimation system is established for field tests. Results of simulation tests and field tests have shown that the proposed method could detect fault source with high accuracy. The localization error is around 5% of the field of view under actual noise environment, which has verified the feasibility of the proposed method.