Application of an autoregressive model in cucumber downy mildew disease forecasting
HUA Lai-qing 1 , XIONG Lin-ping 1* ,SHEN Guang-rong 2 ,MENG Hong 1 ,HU Ya-ping 3 ,ZHAO Sheng-rong 4 (1. Department of Health Statistics, Facuty of Health Service,Second Military Medical University, Shanghai 200433, China; 2. School of Agriculture and Biology, Shanghai Jiaotong University,Shanghai 201101; 3. Agro-technical Extension of Pudong New District of Shanghai,Shanghai 201201; 4. Agro-technical Extension of Songjiang District of Shanghai, Shanghai 201613)
Objective:To explore an autoregressive model of forecasting the cucumber downy mildew disease morbidity(CDMDM). Methods: Based on the theory of infective ordinary differential equation (SIR model), we constructed an autoregressive curvilinear model of CDMDM containing a square term. The parameters of the model were simulated by a time series of cucumber downy mildew disease, and another time series of cucumber downy mildew disease was used to validate the model. Results: The first-order autoregressive model of CDMD was obtained as I (t+1)= 0.006 1 I (t)(100-I (t))+ 0.996 5I (t)+ 0.908 4. Fitting goodness test showed that sum of squared errors was 133.168 7, the coefficient of determination was 0.989 7, and the root mean squared error was 4.080 0. The first-order autoregressive model was better than the second-order model. Conclusion: The model has satisfactory fitting goodness and can be employed to forecast the trend of CDMDM.