| Artificial Intelligence is a dream when human being develops into computer age. It provides the probability for automated learning and acquiring of knowledge,common adapting of knowledge expressing way, high efficiency and entirely of searching solution and intelligence body activation into environment. To meet the need of model established for complexity system,this paper has applied artificial neural network technology in research of crop response to water-sal(tCRW) & soil moisture forecasting under mulched drip irrigation. The main conclusions are as follows:(1) A model of crop response to water based on BP neural network for cotton has developed by using under mulched drip irrigation experiment data. The simulation results of the model shows that it is able to express correctly the relationship between the yield and water use of cotton and the BP neural network is a new method suitable to simulate the crop response to water under mulched drip irrigation.(2) Two types of model with three layers structure based on BP artificial neural network to forecast soil moisture have been set up. The model BP(6,8,1) for forecasting soil moisture in 0-60cm depth has a maximal relative error 10%, and average relative error 2.0%, respectively 5.0% is the maximal absolute error and 1.0% is the average absolute error respectively. The model BP(5,8,1) for forecasting soil moisture in 20-60cm depth has the maximal relative error, average relative error, maximal absolute error, average absolute error 15%,1.0%,11%, 1.0%,respectively. The two models can be used in practice.
|
PDF Full Text Request