The tractor engine is a key component to ensure the normal operation of the tractor.At present,vibration sig-nals are mainly used for engine fault prediction and condition identification.This study proposes a GRU-based recurrent neural network model to build a tractor engine condition recognition system based on audio features by analysing the audio signals generated by the tractor engine under different operating conditions and extracting Mel as the main feature.The prediction results show that the system can accurately identify the normal operation status of the engine and different types of fault conditions,and the recognition rate of abnormal tractor engine can reach 97.15％.The results can be used to help improve the safety and reliability of tractor operation,reduce downtime and improve agricultural production efficiency.
维普
万方数据