Abstract: The booming development of the new energy commercial vehicle market has put forward new requirements for traditional fault diagnosis and maintenance. This study describes the characteristics of new energy commercial vehicles from the perspective of fault characteristics and maintenance diagnosis needs. Then, starting from the perspective of multimodal data fusion, a knowledge generation framework based on AIGC is established, and corresponding technical paths are provided for each main link, namely knowledge acquisition, knowledge modeling, knowledge generation, and knowledge verification. Research has shown that applying AIGC technology to the field of maintenance can help to deeply mine structured diagnostic data and unstructured maintenance experience in vehicle systems, thereby achieving an automatically generated maintenance knowledge base based on AIGC, further improving the speed of vehicle fault diagnosis and the quality of maintenance.