Nonconvex Minimization Based Compressive Sensing Approach for Duct Azimuthal Mode Detection

Abstract

Compressive sensing (CS) based fan noise mode detection methods have been developed recently by taking advantage of the fact that the fan noise sound field is usually dominated by a limited set of modes. It has been shown that the CS based methods require remarkably few measurements and can improve mode detection capability. However, the mode sparsity would be easily destroyed either due to strong noise interference or attenuation of spinning modes. In this Paper, a nonconvex CS method based on minimization algorithm is proposed for fan noise azimuthal mode detection. A series of numerical simulations shows that the new based CS method is more robust to measurement noise and can enhance the mode sparsity under background noise interference. In the case in which the background noise level is unknown, the new based CS method with level correction can successfully reconstruct the mode spectrum by using an underestimated noise level. Compared to the previous convex based CS method, the new based nonconvex CS method reduces the number of sensors and requires weaker constraints for sensor array, which is beneficial for optimal array design in practical tests. In addition, the new method can decrease the reconstruction error and improve the dynamic range. An experimental system is implemented to validate the new method. It is demonstrated that the new method can enhance the mode sparsity and improve the dynamic range by about 10 dB.