Maize response to water with BP neural network method based on limited water

Ecological environment facing the great task, soil and water heavy loss, unbalanced annual rainfall and water supply and demand in semi-arid areas of the Loess plateau, harvested rainwater is very limited in terms of collecting and storing rainwater. Physiological-biochemical characteristic of water effect on crop yield is quite complex, and quantitative calculation is difficult. Applying strict mathernatic at and physical equation is difficult. Professors brings forward many models for crop response to water, but these models have special trait of terrain tract and time domain and can not be used conveniently. There is obviously different model of crop response to water base on limit water which is according to ETmin/ETa defined relative evapotranspiration. Employing relative yield and evapotranspiration can offset effect of apart factor to model, so the paper takes relative yield and evapotranspiration as input and output sample of BP Neural Network, through compared training and analyzing many times, establishing model of maize response to water based on BP Neural Network, and comparing with Jensen model which is used always in China and gained satisfied result.