摘要
以产品类型、固溶温度、固溶时间、时效温度、时效时间作为输入层函数,以拉伸性能和耐磨损性能作为输出层函数,采用5×30×6×2的四层拓扑结构构建了优化体育器材热处理工艺的神经网络模型,并对此模型进行了训练、预测、验证和生产线应用。结果表明,该神经网络优化模型预测精度高,预测误差在2.3%~4.2%;用此神经网络模型优化的热处理工艺参数比生产线传统用工艺的试样抗拉强度增大了5.8%,磨损体积减小了54%,拉伸性能和耐磨损性能均得到了明显提高。
Taking product type, solid solution temperature, solid solution time, aging temperature and aging time as input parameters, and taking tensile property and wear resistance as output parameters, the heat treatment process optimization model for sports equipment based on neural network was built by using four layers topology of 5×30×6×2. And the model was trained, predicted, validated and applied to the production line. The results show that the neural network optimization model has high prediction accuracy, and the prediction error is 2.3%-4.2%. Compared with the traditional technology of production line, the tensile strength of the sports equipment treated by the optimization process parameters based on neural network increases by 5.8%, the wear volume decreases by 54%, and the tensile property and wear resistance improve obviously.
作者
蔺建
张强
LIN Jian;ZHANG Qiang(Physical Education Center,Xijing University,Xi'an 710123,China;School of Materials Science and Engineering,Lanzhou University of Technology,Lanzhou 730050,China)
出处
《热加工工艺》
CSCD
北大核心
2018年第16期248-250,253,共4页
Hot Working Technology
关键词
体育器材
神经网络
热处理工艺
拉伸性能
耐磨损性能
sports equipment
neural network
heat treatment process
tensile property
wear resistance