Research On The Incidence Regularity Of Major Tobacco Diseases And Insect Pests In Hunan And Development Of Forecasting App

Tobacco(Nicotiana tabacum L.)is an important economic crop that is grown on a large scale in many provinces in China,of which Hunan is the third largest tobacco producing area in China.Tobacco pests and diseases are one of the most important factors affecting the yield,quality and economic efficiency of tobacco.At present,the production mainly relies on chemical pesticides to control tobacco diseases and insect pests,but the lack of accurate prediction of tobacco diseases and insect pests leads to the misuse of chemical pesticides,threatening the environment and human health.Therefore,it is urgent to exploit an intelligent and accurate detection and reporting system to guide the prevention and cure of tobacco diseases and insect pests.With the rapid development of modern artificial intelligence and information technology and the continuous progress of measurement and forecasting methods and models,it is possible to build an accurate intelligent measurement and forecasting system.In this study,the main diseases and insect pests of Tobacco in Hunan were taken as the research object,and the occurrence regularity of main diseases and insect pests of tobacco in Hunan was studied based on the investigation data of main diseases and insect pests and meteorological data of 7 stations in Hunan from 2009 to 2015 and 2021.The detection and reporting model of main diseases and insect pests of tobacco in Hunan was built,and the intelligent detection and reporting App of main diseases and insect pests of tobacco in Hunan was developed.Lay a foundation for effective and accurate control of tobacco diseases and insect pests.The specific research results are as follow:Ⅰ.Investigation and analysis of the occurrence regularity and species change of tobacco diseases and insect pests in Hunan province.The historical survey data of tobacco diseases and insect pests in 7 survey sites in Hunan province from 2009 to 2015were collected,and the field incidence of major tobacco diseases and insect pests in Hunan province was investigated from March to August 2021.The results showed that the main diseases and insect pests of tobacco in Hunan from 2009 to 2015 were tobacco mosaic virus,tobacco cucumber mosaic virus,tobacco black shank,tobacco bacterial wilt,tobacco brown spot,Myzus persicae,Agrotis ipsilon,Helicoverpa assulta and Spodoptera litura.Potato virus Y was identified as a major disease in 2021.From the overall change,the year-to-year variation of tobacco mosaic virus,tobacco cucumber mosaic virus,tobacco black shank,Helicoverpa assulta and Agrotis ipsilon were small.The changes of other pests and diseases were generally large.The occurrence time and peak of most diseases and pests in central and southern Hunan were 35 days earlier than that in northwest Hunan.Compared with 2009 to 2015,the incidence of tobacco cucumber mosaic virus,tobacco brown spot,Myzus persicae,Helicoverpa assulta and Spodoptera litura decreased in 2021,and the potato virus Y was aggravated.Ⅱ.The forecasting and reporting model of tobacco diseases in Hunan province was constructed.For the Hunan province,the prediction models of tobacco mosaic virus,tobacco black shank,tobacco bacterial wilt and tobacco brown spot were constructed based on historical survey data.This model is based on fully connected neural network and constructed by stochastic gradient algorithm.The results showed that the four disease detection and reporting models passed the test.The correlation values of the tobacco brown spot detection and reporting model were greater than 0.4 and less than0.7,indicating that the relationship was close;the correlation values of the other three disease detection and reporting models were all greater than 0.7,indicating that the relationship was very close.The regression model fitting degree R~2 value and DW value were calculated,and the DW value was close to 2,indicating that the model fitting effect was good.The results showed that the prediction results of the four diseases were good,and the detection and prediction models performed well,which could be applied in the tobacco disease detection and prediction App.Ⅲ.Preliminary research and development of Hunan tobacco main diseases and insect pests intelligent App.In this study,the design idea and system interface of the App for measuring and reporting tobacco main diseases and insect pests were described.The App mainly included four systems,namely,disease and insect pest identification system,data entry system,measurement and reporting system and disease and insect pest control suggestion system,and completed the function and interface display of the four systems.The App can identify the types of diseases and insect pests and input the data of diseases and insect pests quickly and conveniently.The results of disease and insect pests measurement and prevention methods are displayed in the client for users to view,laying a foundation for subsequent production and application.