Research Of Grain Production Prediction Based On AIGA-BP Neural Network

China is a large agricultural country, the grain problem is the most important in the agricultural production. The fluctuation in Grain production will inevitably cause the unstable national economy Grain production system is a very complicated system, involving many factor, The purpose of research is to adopt a more advanced theory to improve the present food production prediction method, and creat a new prediction model.Firstly, this thesis reviewed the basic condition of grain production in Henan province, and then introduced the theory of principal component analysis and steps to carry out it. SPSS17.0 software was used to analyze grain production influencing factors between 1978 and 2009 in detail at three stages.9 and 12 grain production factors were selected to do the principal component analysis in this paper, and nine factors were found not good enough to reflect grain production, therefore,12 factors were selected as the grain prediction input value of Henan province.Secondly, described in detail the theory of BP neural network, explained the advantages of using the BP neural network to predict grain production. The results of using BP neural network to predict grain productionwere that the seven years of average absolute error is 267.14 ten thousand tons, and specific error about each year is more larger. Detailed analysis of the reason that the BP network is vulnerable to fall into local minimum was made, then the new method was adopted to solve the problem of BP network. Finally, this paper studies the adaptive immune genetic algorithm (AIGA), and then use the AIGA to optimize the BP neural network weights and thresholds values, used the AIGA global search method to solve the local minimum values of BP network, and meanwhile established the AIGA-BP network prediction model about Henna province's grain production. The simulation experiment results were that the average absolute error of grain production predicted by the new model is 127.02ten thousand tons, the result shows that the AIGA-BP neural network model has a higher prediction accuracy than the BP network model.This paper analyzed the rules of influencing grain production prediction factors, and set the grain influence factors data between 2010 and 2012 in accordance with rules. Disaster factors of three different situations were considered, and then the data of the three different situations were set accordingly. Finally, the AIGA-BP model was used to predict the grain production of Henan province between 2010 and 2012.