利用B P神经网络对作物生育动态进行模拟。在对B P神经网络的结构和训练算法加以选择后 ,网络模型具有拟合精度高、收敛速度快和泛化能力强等特点。B P网络可用于作物生育进程中单输入单输出、单输入多输出、多输入单输出和多输入多输出动态关系的模拟和作物生育动态的预测 Using Bp neural network to simulate crop growth dynamics. After choosing the structure and training algorithm of BP neural network, the network model has the characteristics of high fitting precision, fast convergence rate and generalization ability. B P network can be used to simulate the dynamics of crop monocropping, single-input multiple-output, multiple-input-single-output and multiple-input-multiple-output