A multi-step knowledge reasoning model based on information enhanced graph representation and deep learning model

In order to solve the problems of low efficiency, high pressure and error prone caused by manual preparation of new equipment startup task, a multi-step knowledge reasoning model based on information enhanced graph representation and deep learning model is presented. The proposed methodology firstly generates the historical behavior sequences from the historical cases. Then, a scheduling behavior structure diagram is built based on joint behavior statistics data, and a large number of walking behavior sequences are generated using random walk strategy. Finally, these walking behavior sequences are encoded and used for training the knowledge reasoning model based on deep learning. By simulating the intelligent generation of new equipment startup steps in a provincial power grid through case reasoning, the effectiveness of model is verified, which provides a technical foundation for intelligent electric system.