Aiming at the problem of motor rolling bearing fault diagnosis accuracy, a motor fault diagnosis method based on wavelet packet analysis and artificial fish (AFSA) optimized support vector machine (SVM) was proposed. Wavelet packet multi-resolution analysis method was used to decompose and reconstruct the motor vibration signals, and the time-frequency diagrams of different frequency bands were obtained. Then, AFSA algorithm is used to optimize the parameter penalty parameter(C) and kernel parameter() in the support vector machine. Finally, the fault diagnosis model of AFSA-SVM is established, and the frequency band energy value in the signal time-frequency graph is extracted as the input feature vector of AFSA-SVM for learning and testing. Finally, the simulation experiment verifies that the fault diagnosis accuracy is up to 98.7%, and compared with the particle swarm optimization support vector machine (PSO-SVM) and the unoptimized support vector machine, the results show that the proposed method has a higher fault diagnosis and recognition effect.