Sound Crosstalk Cancellation Research On The Transfer Path Analysis

Generally, We have three methods to control noise. They are sound source control, route of transmission control and receiver protection. The most fundamental method is source control. Identifying the main noise source correctly is the most important thing to achieve the goal of source control. So we can take targeted measures to control the sound source.In the actual structure, multiple noise sources spread at the same time, the sound source will produce crosstalk with each other. It will increase the difficulty of getting the dominant noise source.In the field of noise and vibration, transfer path analysis (TPA) has developed rapidly. It is trying to solve the problem of identifying the main excitation source. However, the traditional transfer path analysis method cannot eliminate the crosstalk. Therefore, we need to improve the traditional TPA.In this paper, with the mastering of TPA theory, I first decompose the traditional TPA process, and then add steps to eliminate the sound crosstalk. I combine the incomplete Cholesky decomposition pre-processing and least squares methods to solve pathological equations which can make TPA better to find the noise and vibration excitation source. After that, I start Gui interface programming to make convenient, stable and efficient use in noise and vibration engineering practice.Finding out that the theory very effective, I start a series of experimental verification, the results show that the method of crosstalk cancellation is effective and the optimized TPA is better than the traditional TPA. It can solve sound crosstalk problem and identify the main noise source.