Detection method of abnormal load state of diesel generator set based on PANet network and IMF

Aiming at the problems of missing smooth data, limited detection range and high false alarm rate in diesel generator set load state anomaly detection, an improved method based on PANet network and IMF is proposed. Through multi-sensor data acquisition and pre-processing, the anomaly detection model is constructed by combining PANet network and IMF to achieve comprehensive coverage and high accuracy of load state detection. Experiments prove that the method has excellent performance and is close to the ideal detection effect.