Forecasting model based on an improved Elman neural network and its application in the agricultural production

On the base of analyzing the dynamic characteristics of Elman neural network, this paper proposes to use an improved Elman neural network to forecast in the agricultural production areas against to the BP neural network´s static defects. We uses the data of rice pest-Chilo to simulate. The experiment shows that the improved Elman neural network has better predictability and stability than Elman neural network and BP neural network.