【摘要】 对神经网络ＢＰ算法的一些改进作了大量的数值试验研究，并组合了部分改进方法用于动力配煤问题中状态预测过程的建模．最后把神经网络方法纳入到非线性混合离散变量优化设计方法（ＭＤＯＤ法）中，对杭州配煤场的动力配煤过程作了应用介绍．更多还原

【Abstract】 A large quantity of numerical experiments were conducted for some improvement of BP algorithm of neural networks,some improved methods were combined for setting up model of state estimation process for steam coal blending,at last,the neural networks method is brought into non linear mixed discrete variables optimum design method and its application for steam coal blending process in Hangzhou blending yard is described.更多还原