Abstract: By analyzing sound signal,this paper analysis the sound signal of a kind of compact horizontal driver in advanced automobile seats to evaluate the sound quality.The sym6 wavelet was employed to de-noise given sound signal,and a band-pass filter was used to acquire specific frequency band of the de-noised signal.Hilbert transform was then adopted to demodulate filtered signal the main component of the demodulated signal was the characteristic signal containing fault information of the driver;consequently,a low-pass filter was utilized to eliminate high frequency noise and the anticipated signal containing fault information thus was acquired;fourier transform was performed for the low-pass filtered signal to obtain its frequency spectrum.Finally,the amplitude of the signal at characteristic frequency was compared with a preset threshold to determine if the driver was defective or of poor sound quality.This method had high detection accuracy and anti-noise ability over the whole frequency band.