Aiming at the instability of vibration signal of scroll compressors and the difficulty in obtaining a large number of fault samples， a fault diagnosis method based on information entropy fusion and particle swarm optimization-support vector classification （PSO-SVC） for scroll compressors is proposed to determine its fault types. Firstly， the singular spectrum entropy of time-domain， the power spectrum entropy of frequency-domain and the variational mode decomposition （VMD） energy entropy of time-frequency domain are extracted to calculate the information entropy of vibration signals. Then the single evaluation index eigenvector is obtained by fusing the above three types of entropy using factor analysis method. Lastly， the scroll compressor PSO-SVC fault classification model is established through signal data training after the evaluation index is input into the PSO-SVC model. The experimental results show that this method can effectively classify four typical fault types of scroll compressors with 94.5% accuracy in the case of small sample size.