Research on Technical Route of AI Large Models for Electric Power Knowledge Services

Artificial intelligence (AI) large language models have shown great potential and value in various vertical fields. In response to the demand for the intelligent development of new power systems, this paper summarizes the difficulties faced in the application of AI large language models in the field of power such as large language model illusions. To address these difficulties, starting from the two dimensions of large language model response enhancement and knowledge enhancement, combined with technologies such as fine-tuning, prompt engineering, and retrieval-enhanced generation, this paper proposes an engineering framework for the implementation of AI large models for electric power knowledge services, providing support for empowering the construction of new power systems in China with AI large models. Finally, this paper gives the corresponding solutions to the challenges faced in building large language models for power systems, including difficulties in data acquisition, data security, low interpretability, and high research and development cost.