摘要
将BP神经网络的理论和算法应用于轧钢力学性能的预测研究,采用实际的轧钢现场生产数据,建立工艺参数、化学成分与力学性能的映射模型。经过测试和评估,该BP网络能较好地预测轧钢产品的力学性能。
Based on the theory of artificial neural network, BP algorithm is used for the training of networks, and the relationship between the controlling parameters in hot rolling (temperature, chemical composition,strain quantity, ere) and the parameters of mechanical properties is established. The calculation results are in good agreement with the experimental results.
出处
《荆门职业技术学院学报》
2008年第3期34-37,共4页
Journal of Jingmen Technical College
基金
湖北省自然科学基金项目(项目编号:2000J016)
关键词
BP神经网络
热轧钢材
性能预测
BP artificial neural network
hot rolled plate
mechanical propertiesp rediction