Propagation Prediction of COVID-19 Based on Immune Differential Evolution Algorithm and Dynamic DSEIR Model

Li, Z.; Feng, F.

In this paper, we developed a dynamic DSEIR infectious disease dynamics model and used an Immune Differential Evolution Algorithm to analyze the effect of specific measures on the spread of the epidemic. For the prediction of the number of confirmed cases and deaths, a dynamic time-lagged kinetic model (DSEIR) is developed and an Immune Differential Evolution Algorithm is used to estimate the unstatistical parameters of the model, and finally the model parameters are adjusted to analyze the effects of health measures. We first improve the traditional SEIR model to find the predicted number of cases based on the time-lag characteristics of the epidemic, and estimate the uncounted model parameters by fitting the predicted results to the actual statistics using an Immune Differential Evolution Algorithm. The model parameters were adjusted according to the behavior of implementing the home order in the United States and the strict quarantine measures adopted in Hubei Province, China in advance and delay to analyze the change pattern of the prediction results. It can be concluded that: early adoption of strict quarantine will significantly reduce the epidemic losses in Hubei Province, and vice versa will significantly increase the losses;if the home order is implemented immediately will significantly reduce the epidemic losses in the United States. The advantages of this paper are: the traditional SEIR model is improved by combining the actual characteristics of COVID-19, and the parameters are estimated using the Immune Differential Evolution Algorithm, which makes the prediction model closer to the actual situation of this epidemic. © 2021 IEEE.