机构地区: 河北科技大学材料科学与工程学院
出 处: 《锻压装备与制造技术》 2003年第5期56-59,共4页
摘 要: 采用实验研究与神经网络相结合的方法进行研究,通过大量实验,获取大量数据,在此基础上,建立BP网络模型。通过对开式冷挤压极限变形程度神经网络计算结果与实验结果的比较,其精度较高,证明用神经网络方法既可以实现开式冷挤压工艺的参数预测,又能给出系统完整的可供指导实际生产的工艺参数数据,对于开式冷挤压的实际生产具有指导意义,是一种可行的分析手段。 Based on the method of experimental study combined with ANN and a large amount of experimental data,three BP models have been built with higher precision of prediction results of ANN,compared with the experi-mental results about open-die extrusion maximum reduction.The ANN has been proved not only to realize the pre-diction of the technology parameters of open-die cold extrusion,but also give the systematic complete parameters da-ta,directing practical production,being significance to practice production of open-die cold extrusion hence a feasi-ble analysis means.