Sound pressure level real-time monitoring of outdoor directional noise

Abstract

A single microphone can’t select an efficient target signal from a mixed sound source when measuring the sound pressure level (SPL) of a target noise source in an outdoor environment. Therefore, a microphone array is used to enhance the target signal through beamforming to restrain the interference of non-target signals effectively. Owing to their advantages of small size and low price, micro-electro-mechanical system (MEMS) microphone arrays are widely used. However, MEMS microphone mismatch commonly occurs due to production tolerances as well as aging effects. To calibrate the SPL measurement difference caused by the whole mismatch microphone array when outdoor noise monitoring is performed, a real-time calibration method was introduced in this study by compensating for the SPL difference between the reference microphone and MEMS microphone array. The circular harmonics domain-multiple signal classification sound source localization algorithm was utilized to select the effective target signal component in the time–frequency domain, and the standard transfer functions of the reference microphone were utilized to correct the SPL measurement bias of the array. To verify the feasibility of the method, the consistency of the compensating SPL difference between multi-noise and single-target noise environments was experimentally investigated. The calibration method exhibited good accuracy and robustness based on the results and can also be applied to spherical microphone array noise monitoring device.