期刊文献+

铝合金板带材力学性能极限学习机预测模型

Prediction model of mechanical properties extreme learning machine for aluminum alloy sheet and strip
下载PDF
导出
摘要 为探究极限学习机(ELM)预测模型在铝合金板带材力学性能预测上的可行性,在数据预处理的前提下,首先通过试验确定了极限学习机网络的最优参数,将最优的极限学习机网络应用于铝合金抗拉强度、屈服强度、伸长率的预测中。结果表明,在测试集上,极限学习机模型在抗拉强度、屈服强度、伸长率上的平均绝对百分比误差(M_(APE))分别为7.41%、14.09%、14.99%。极限学习机预测模型在铝合金板带材力学性能预测中具有可行性,且该预测模型针对抗拉强度的预测精度优于对屈服强度和伸长率的预测精度。 The paper is to to explore if it is feasible for the extreme learning machine(ELM)prediction model to predict mechanical properties of aluminum alloy sheet and strip.On the premise of data preprocessing,the optimal parameters of ELM network were determined through experiments,and the optimal ELM network was applied to the prediction of tensile strength,yield strength and elongation of aluminum alloy.The results show that the mean absolute percentage error(M_(Ape))of ELM model in tensile strength,yield strength and elongation is 7.41%,14.09%and 14.99%,respectively.The ELM prediction model is feasible in predicting the mechanical properties of aluminum alloy sheet and strip,and the prediction accuracy of the model for tensile strength is better than that of yield strength and elongation.
作者 熊振强 李家栋 李勇 赵鹏 XIONG Zhen-qiang;LI Jia-dong;LI Yong;ZHAO Peng(State Key Laboratory of Rolling and Automation,Northeastern University,Shenyang 110819,China)
出处 《轻合金加工技术》 CAS 2022年第12期19-24,共6页 Light Alloy Fabrication Technology
基金 国家自然科学基金(51790485) 山东省重点研发计划项目(2019JZZY010401) 南宁市创新创业领军人才“邕江计划”资助项目(2019002) 南宁市科技重大专项项目(20201041)。
关键词 铝合金 性能预测 机器学习 极限学习机 aluminum alloy property prediction machine learning extreme learning machine
  • 引文网络
  • 相关文献

参考文献10

二级参考文献195

共引文献456

;
使用帮助 返回顶部