AUTOMOBILE ABNORMAL NOISE DETECTION BASED ON AUDITORY PERCEPTION ANALYSIS

As a key factor affecting ride comfort and the perceived quality of a vehicle, efficient detection of abnormal sounds has become an important research topic in the field of automotive acoustics. To address the low efficiency and poor consistency of traditional subjective detection methods, this study innovatively introduces psychoacoustic auditory models into the field of abnormal sound detection. First, the abnormal sound signals are transformed into perceptual feature representations through biomimetic auditory processing. Then, a dual-threshold detection algorithm based on a differential offset criterion is proposed. By integrating an absolute threshold and a minimum loudness constraint, the method enables precise identification of the abnormal sound boundaries. Experimental results show that the proposed method maintains low computational complexity while accurately detecting the onset and offset of abnormal sounds. Moreover, it demonstrates excellent robustness under complex noise conditions, providing an effective technical solution for active acoustic quality control in intelligent vehicle cabins.