Granary risk event intelligent early warning and processing method based on ensemble learning and time sequence diagram neural network

The invention relates to a granary risk event intelligent early warning and processing method based on ensemble learning and a time sequence diagram neural network, and the method comprises the steps: S1, collecting granary risk events and processing scheme related data, and constructing a granary safety domain knowledge graph; s2, predicting granary sensor data in a certain period of time in the future based on a multi-model fusion Stacking ensemble learning method; s3, predicting the probability of occurrence of potential risk events of the granary in a certain period of time in the future based on a sensor perception prediction framework of sparse attention of the time sequence dynamic network; and S4, providing a granary risk event intelligent early warning and processing method in combination with the occurrence probability of the potential risk event and the knowledge graph. According to the method provided by the invention, the safety state of the granary can be comprehensively and accurately predicted