In order to detect the problems of food safety in time and reduce the incidence, a food safety supervision of real-time analysis and prediction model based on the storm cloud platform and the Elman neural network optimizing by the bat algorithm was constructed from the perspective of food supply chain. Through analyzing the factors affecting food safety, the model was used to analysis the food safety data in real-time. By the analysis of the model, an effective food safety supervision mechanism in real time was formed to further reduce the food safety hazards.