【摘要】 利用BP神经网络对斜纹夜蛾幼虫的存活过程进行建模分析,给出了BP神经网络建模的Matlab算法程序。结果表明,BP神经网络能充分逼近斜纹夜蛾幼虫的存活过程曲线,使用BP神经网络可对斜纹夜蛾幼虫的存活过程进行有效的建模。利用所建立的模型,得到了关于斜纹夜蛾幼虫存活过程的模拟预测结果。不同温度下斜纹夜蛾幼虫种群的死亡过程有所不同。在20,24,36和40℃温度下,斜纹夜蛾幼虫种群的死亡高峰分别出现在卵孵化后20d左右,5～20d,5～10d及3～6d。最后,对神经网络建模的特点,以及影响BP神经网络建模效果的主要因素进行了分析讨论。更多还原

【Abstract】 BP neural network may approximate non-linear functions with desired accuracy. This paper deals with neural network modeling of larval survivorship of Spodoptera litura F. It was found that BP neural network could be effectively used to model the larval survivorship-time relationship. Some results for dynamic process of larval survivorship were obtained. Matlab codes for BP network model were given. The features of neural network modeling, and the factors influencing the efficiency and goodness of modeling were discussed.更多还原