Study On The Prediction Model And Control Index Of Three Tobacco Leaf Diseases In Jilin Province

Tobacco brown spot disease,wildfire disease and angular leaf spot disease occurred in the world,and were three more serious diseases in tobacco growing areas in Jilin Province,China.At present,the control of tobacco diseases mainly relied on the fungicides and bactericides.However,due to the wide variety of pesticides and the influence of climatic factors,tobacco farmers were blinded in spraying pesticides.In order to clarify the regularity of occurrence and the best time to use pesticides,the growth models,the prediction models and the control indixes of of these three tobacco diseases were established in this paper.The results were as follows:Through the investigation of the diseases indexes of tobacco brown spot,wildfire and angular leaf spot,the regularity of the indexes of these three diseases changed with time was simulated with SPSS20.0 software,It is found that the cubic function and logistic function have good simulation effects on tobacco brown spot disease,and the cubic function has good simulation effects on wildfire disease and angular leaf spot.According to the disease growth model,the regularities of this three diseases were deduced.The results showed that the epidemic trends of tobacco brown spot,wildfire and angular leaf spot were roughly same,those were,the diseases spreaded slowly when the disease started,and then entered into the stage of rapid disease development,the disease index increased rapidly.After the first half of August,the disease index increased slowly or no longer.The key predictors were screened by the coincidence method and the correlation coefficient method,and the disease index at mid-June was also as a key predictor,and the SPSS20.0 software was used to analyze the regression relationship between the disease indexes of tobacco brown spot at mid-August and key predictors from 2016 to2020,and 5 tobacco brown spot prediction models were established.The average degrees of fit of model 1 to model 5 were 91.69%,91.32%,96.18%,98.20% and88.02% respectively.Using tobacco brown spot disease index at mid-August and key predictors data in 2016-2019 to establish three tobacco brown spot prediction models as model 6-model 8,the average degrees of fit were 76.90%,77.42%,and 82.07%,respectively,and Model 6 to Model 8 had respectively 73.03%,69.26% and 79.01%accuracy rates for forecasting tobacco brown spot disease index in 2020(mid-August).Using the 2016-2020 tobacco wildfire disease index data(mid-August)and key predictor data to establish a tobacco wildfire disease prediction model,which was model 9,its average degree of fit was 86.17%,using the 2016-2019 tobacco wildfire disease index(mid-August)and key predictor data establish a tobacco wildfire disease prediction model,which was model 10.The average accuracy rate was 76.22% for forecasting the tobacco wildfire disease index in mid-August 2020.SPSS20.0 and Eviews9 software were used to establish an ARIMA(1,2,0)model of tobacco wildfire disease by time series analysis,and its degree of fit was87.57%.The average accuracy rate of time prediction was 88.45%.The prediction model ARIMA(2,2,0)of tobacco angular leaf spot was also established by time series analysis method,and it had a fitting degree of 85.06%,and the average accuracy rate of the five time predictions for the disease index of tobacco angular leaf spot from August 14 to August 26 was 94.15%.By calculating the allowable loss rate,the output value loss rate and the allowable level of economic damage,the SPSS20.0 software was used to perform regression analysis on tobacco brown spot disease,wildfire and angular leaf spot disease index and output value loss rate data,the control indexes of these three diseases were determined.The results showed that the control index of tobacco brown spot disease was between disease index 1.48 and 4.57,it meaned that when the disease index in the tobacco field reached 1.48,the first control was started,and the control index of tobacco wildfire was between disease index 1.43 and 4.38.it meaned that when the disease index in the field reached 1.43,the first control was started.and the control index of tobacco angular spot was between disease index 1.51 and 4.53,it meaned that when the disease index in the field reached 1.51,the first control was started.After the first control,according to the development of the diseases,it needed to determine if a second or third control was necessary.