Research on GRU Based Emergency Material Delivery Prediction Algorithm

In order to solve the problem of the influence of the large-scale flow mode of all kinds of emergency materials on the epidemic situation, this paper first analyzes the changes of the epidemic situation before and after putting in the vegetable package, and uses the time series clustering method to cluster the daily epidemic situation. It is determined that the vegetable package can control the epidemic situation to a certain extent. Then, the GRU prediction algorithm is established, and the epidemic data of unput vegetable bags in other regions at home and abroad are used to train GRU to predict the epidemic situation of unput vegetable bags in Changchun, and compared with the actual data, the accuracy of the algorithm reaches 97.8%. The results showed that the predicted data were significantly higher than the real data, both confirmed cases and asymptomatic infections. on the 32nd-40th day, the predicted confirmed cases were 2.7 times higher than the actual data. On the 23rd-81st day, the predicted asymptomatic data was 1.2 times higher than the actual data, which further verified that the vegetable package could effectively inhibit the epidemic.