With the widespread adoption of new power systems and the increasing demand for automated processing of complex scenarios within these systems

the application of multimodal large models in electric power domain has become critically important. So the application and researches of multimodal data fusion large model technology in the fault detection and diagnosis of power equipment

the smart grid management and the integration of multimodal knowledge graphs with large models are reviewed. Firstly

the current state of AI large model development

the fundamental architecture of mainstream large models and their applications are introduced. Secondly

the types and processing methods of multimodal data are discussed

including audio

text and image data. Furthermore

the application of multimodal large models in fault detection is analyzed and the integration strategy for multimodal data in smart grids is discussed

such as power load forecasting and the use of multimodal knowledge graphs in distribution networks. Finally

the current challenges and future research directions of multimodal large models in electric power domain are summarized

which will provides the theoretical reference for future development.