The Methods Study Of Tobacco Yield Prediction And Early Warning Of Diseases In Heilongjiang Province

Tobacco is an important economic crop, its taxes is one of the important source of national and local government revenue, ranking first in the industry. Heilongjiang province is one of the main producing areas of quality fill tobacco, favored by many domestic and foreign cigarette factory. The stability of the tobacco production in Heilongjiang province is particularly important, we depth study which factors have a significant impact on tobacco production and How to prevent the occurrence diseases, and on this basis, establishing tobacco production forecast model and the early warning model of major pests and diseases is of great significance.These research results, practicing proper guidance and management to provide a basis for the implementation of the tobacco and technical personnel, at the same time, there is a useful exploration for tobacco production forecast and early warning of tobacco diseases. It can provide the corresponding reference for researchers in other related fields; it also can be applied to other areas of the forecast and Early Warning. The main research content is as follows:1、Analysis of literature data, according to the characteristics of climate regions of Heilongjiang region,after stepwise filtration, select the transplanting time xl,pure nitrogen application x2,pure phosphorus application x3,pure cilium application x4,The cake fertilizer application x5,and May-August average temperature x6six fundamentals of regression model, in addition to these six basic factors. The dependent variable leaf production y and independent variables xl-x6data are derived from the Harbin tobacco companies in2009the branch of statistics, the theory of stepwise regression, make Mathematical Statistics Toolbox of MATLAB mathematical software, establish quadratic function stepwise regression model between Heilongjiang Province tobacco production and these factors, then apply the stepwise regression model predicted the2010tobacco production.2、Further analysis of each single factor on the yield of each single factor marginal yield effect, the last in-depth analysis of the optimal cultivation practices, and get the maximum value of tobacco production in the optimal cultivation conditions.3、According to the warning theory as well as early warning indicators of system optimization principle, establishment of early warning indicators of tobacco wildfire disease in Heilongjiang province, to the extent of police intelligence, police sources and warning signs indicators system, and use the time difference correlation analysis to determine the index system leading, coincident, lagging properties.4、On the basis of complete the selection and classification, first determine the police sentiment indicators warning limit and police, the user feedback to determine the range of warning signs warning limit indicators and warning degree. The application identified by police intelligence indicators and warning signs indicators warning limit and warning degree build a decision tree model of Heilongjiang Province, tobacco wildfire disease warning, reasonable a decision tree, appropriate analysis and early warning results, further warning of the decision tree model,5、The validity of discriminate, through the establishment of the time series model to predict the meteorological data of2012in Heilongjiang province, forecast meteorological data combined with decision tree model Heilongjiang Province, in2012the incidence of wildfire disease early warning.6、The results of forecasting and early warning, there is corresponding discussion finally in the paper, providing appropriate advice and recommendations for tobacco production in Heilongjiang province.