Abstract: The automotive engine is one of the core components of a vehicle, and traditional engine fault diagnosis methods usually rely on sensor data and physical models, but the installation, maintenance and calibration of sensors, and development and maintenance of physical models require considerable cost investment, and data fusion and complex mathematical models are difficult to adapt to different vehicle models, driving conditions, and environmental changes, which may lead to inaccurate diagnosis in some cases.In this paper, the technology and related methods of automobile engine condition recognition based on audio features are systematically discussed, and the theoretical research of automobile engine condition recognition technology based on audio features is improved. The research results will provide theoretical guidance for practitioners and researchers in the automotive field, and promote the development and innovation in the field of automotive engine fault diagnosis.