基于改进BP神经网络对江苏省粮食产量的仿真预测

Application of Improved BP Neural Network in Forecasting Grain Yield

该文分析了传统BP神经网络收敛速度慢的原因,提出了一种BP神经网络的改进算法,在此基础上建立了时间序列对象的粮食产量预测模型,并运用Matlab对江苏省粮食产量进行了预测,仿真测试结果证明了该方法可行,具有实用性。

Cause for slow convergence about the traditional BP neural network are analyzed,and an improved BP algorithm is proposed in this paper. Moreover,we develop a forecast model of time sequence object on this basis,and then this model is studied on the Grain Yield in Jiangsu Province with Matlab.The results from the simulation test proved that the algorithm is feasible and practicable.