Establishment And Realization Of Prediction Model Of Modern Agricultural Equipment In Xinjiang Corps

With the development of 3S technology, automatic control technology, electronic information technology and other new technology, agricultural equipment has gradually developed into a modern and intelligent direction. Agricultural equipment as an important indicator of the modernization of agriculture plays an important role in the development of modern agriculture in Xinjiang Corps. Currently, the Corps has not been built an appropriate indicator and predictive model for the rapid development of agricultural equipment, and cannot assess and predict the impact of the trends of modern agricultural equipment.This paper based on the statistical data of Xinjiang production and Construction Corps in 2001-2014 years of modern agricultural equipment, built Xinjiang modern agricultural equipment level index system including six indicators. It contains sowing equipment level, harvesting equipment level, fertilization equipment level, film equipment level, irrigation equipment level and straw returning equipment level. Using modern agricultural equipment 2001-2012 years statistical data, we established exponential smoothing, BP neural network and GA-BP neural network prediction model, the percentage error of the 2013 and 2014 to verify the model’s prediction accuracy.The prediction results show exponential smoothing model is simple and convenient. When trends in the data is relatively stable or less volatile situations, such as the level of harvesting equipment, model predictive accuracy is higher; when the large fluctuations in the trend of data, such as the level of straw returning equipment, results exist great bias. BP neural network learning expectation output samples and network actual output correlation coefficient is very high, above 0.9. Prediction results are having high accuracy, shows that the model is reasonable and feasible. But the initial weights and thresholds are randomly generated, cannot be obtained accurately. Using genetic algorithm to optimize BP neural network, and genetic algorithm evolution in less than 40 generations can find the optimal weights and thresholds, the result of GA-BP neural network prediction model shows high prediction accuracy.Considering the prediction accuracy and the stability of the model, the GA-BP neural network is used to predict the development of Xinjiang Corps modern agricultural equipment level in the next five years. By 2015-2019 years forecast trend can be seen, the application of new technology of agricultural equipment greatly promoted the development of agricultural equipment level, indicators all have upward trend, modern equipment with higher levels of indicators such as sowing equipment development tends to be smooth, modern equipment with lower levels of the index such as film equipment level will remain volatile state of adjustment, but the adjustment will decrease gradually.