The invention discloses a method and a system for identification of signal type of cable partial discharge signals. The method includes the steps that noise removal processing of obtained cable partial discharge signals is conducted, the cable partial discharge signals which has gone through the noise removal processing are obtained; the cable partial discharge signals which has gone through the noise removal processing is conducted to optimization processing to obtain a sample data set; a training data set is sent to a support vector machine (SVM) learning model to train and learn, parametersof the SVM model are optimized to obtain optimum parameters by using particle swarm optimization (PSO), a SVM signal type identification model corresponding to the optimum parameters is determined and signal type of the cable partial discharge signals to be checked is identified by using the SVM signal type identification model corresponding to the optimum parameters, identification results are determined. According to the method and system for identification of signal type of the cable partial discharge signals, cable partial discharge signals can be recognized in batches compared to artificial recognition, working efficiency is improved greatly, meanwhile recognition accuracy is improved greatly compared to the artificial recognition.