The partial discharge fault identification method driven by multiple types of fault information is of great significance for improving the accuracy and fault tolerance of fault identification. In this paper, typical partial discharge types in switchgear are identified, and four typical partial discharge models (corona discharge, surface discharge, floating discharge, and air gap discharge) are set up. Ultrasonic (Ultra) method, very-ultra-high frequency (V-UHF) method, and pulse current method (PCM) are used to collect partial discharge signals generated by different discharge types. First, the deep convolution neural network (CNN) algorithm is used to train the measurement data of different sensors, and then the Dempster Shafer (D-S) evidence theory is used to fuse the recognition results of multidimensional information sources, and the final decision is made. The results show that compared to the fault identification mode based on a single information source, the fault identification mode based on multidimensional information sources has higher accuracy and can correctly identify the discharge type even when one of the multidimensional information sources makes a false judgment, with better fault tolerance and better recognition effect.