Abstract: Aiming at the problem of abnormal sound and vibration caused by workers' installation errors in refrigerator production, a method of refrigerator sound and vibration fault diagnosis based on improved variational modal decomposition-support vector machine (IVMD-SVM) is proposed to realize the fault self-check for sound and vibration of refrigerators. First, use the collector to obtain the acoustic and vibration signals of the refrigerator under different working conditions, and then perform variational modal decomposition, feature extraction and other preprocessing on the acoustic and vibration signals to obtain the feature data set, which is divided into training set and test set, and then use the training set. Set the training SVM model, and finally apply the trained model to the test set, and output the fault identification result. The results show that the method has high diagnostic accuracy (95.6%, 96.8%), and is a sound-vibration fault diagnosis method for refrigerators suitable for actual working conditions.

Key words: Refrigerator, Fault diagnosis, Variational mode decomposition, Support vector machine, Acoustic and vibration