论文部分内容阅读
通过将回归网络结构和移动窗口技术加入反向传播神经网络（BP）算法程序中，对预测橡胶硫化控制时变系统的训练样本进行了分析.最后在实验的基础上提出了选取训练样本的两条准则.“,”This paper analyzes the training samples of a time_varying system used in the forecast of rubber curing control by adding the recurrent network and moving window to the algoitnm of the back_propagating(BP) neural network, two criteria to choose training samples being put forth on the basis of experi_ments.