论文部分内容阅读
对鞍钢冷轧厂四机架冷连轧机轧制压力模型进行了认真分析，指出了其存在的缺陷．把遗传算法（GeneticAlgorithms，简称GA）和神经网络有机结合，设计出了具有遗传算法性能参数优选、网络结构参数优选、网络性能参数优选以及GA-BP算法联合进行网络权值修改几种功能的遗传神经网络，建立了基于遗传神经网络的新冲连轧机轧制压力模型．通过原模型计算值、新模型计算值与实测值之间的对比分析可知，遗传神经网络模型计算精度优于传统轧制力模型．
The rolling pressure model of 4-stand tandem cold rolling mill in Anshan Iron and Steel Co., Ltd. was carefully analyzed, and its defects were pointed out. By combining genetic algorithm (GA) with neural network, we designed a genetic algorithm that has some functions of genetic algorithms, such as optimization of performance parameters, optimization of network structure parameters, optimization of network performance parameters and GA-BP algorithm combined with network weight modification Neural network, a rolling pressure model of a new stamping mill based on genetic neural network was established. By comparing the calculated value of the original model and the calculated value of the new model with the measured value, it can be seen that the calculation accuracy of the genetic neural network model is better than the traditional rolling force model.