基于FUDAN-CCDC模型对新冠肺炎的建模和确诊人数的预测

科学地预测疫情发展趋势对疫情防控至关重要.在新时滞动力学模型(TDD-NCP)的基础上,提出基于随机动力学的时滞卷积模型和离散卷积模型,并基于中国疾病预防控制中心的相关研究结果及公开数据以及Wallinga和Lipsitch的工作,反演出COVID-19的重要参数,拟合了武汉及上海市疫情发展趋势.

Modelling for COVID-19 and the Prediction of the Number of the Infected Based on FUDAN-CCDC

Cheng Jin's group(Fudan University,Shanghai,China)has developed some models for the growth rate of COVID-19(TDD-NCP models and a series TDD-NCP-CCDC models or FUDAN-CCDC models),the main idea is to use the time delay model to fit the real data.Based on these previous work,considering the outstanding agreement with data from Chinese Center for Disease Control and Prevention(CCDC)and the estimations on the growth rate are very stable,we use the observed data reported in CCDC to estimate distribution of the generation interval of the infection and apply the simulation results from the time delay dynamic system as well as released data from CCDC to fit the growth rate.In addition,based on Wallinga and Lipsitch framework,the reproductive number R0 of COVID-19 was estimated.In the last part of the paper,we present some figures of the evolution and predictions of the event computed from the model.The data used in the simulation is ONLY the cumulative infections of different region from CCDC until Feb.16.The results show that the novel dynamic system can well predict the outbreak trend so far.