Aiming at the current problem of difficult to choice the classifier parameter of SVM for the power quality disturbance identification and classification, put forward a new kind of power quality disturbance identification method based on the particle swarm optimization (PSO) SVM. Using MATLAB software to model five kinds of common power quality disturbance signal, the detected voltage signals were used as the input samples of the PSO-SVM training and testing after complex wavelet transform. The simulation results show that the method can identify and classify the power quality disturbance quickly and reliably, it has a high application value for power quality monitoring of the power grid.