摘要
以钒钛改性镁合金成分、模具预热温度、始锻温度、终锻温度、锻造比为输入层参数,以拉伸性能为输出层参数,采用5×35×7×1四层拓扑结构构建出钒钛改性镁合金汽车轮圈锻造工艺优化的神经网络模型。结果表明,神经网络模型相对预测误差在2.6%~4.2%,平均预测相对误差为3.3%。与生产线传统工艺相比,采用神经网络模型优化工艺成形的汽车轮圈抗拉强度增大了53%。
Taking magnesium alloy composition modified by vanadium-titanium, die preheating temperature, initial forging temperature, final forging temperature and forging ratio as input parameters and taking tensile property as output parameter, the neural network model of forging process optimization of the magnesium alloy modified by vanadium-titanium automobile wheel rim was built by using the 5×35×7×1 four layer topology. The results show that the relative prediction error of the neural network model is between 2.6% and 4.2%, and the average relative prediction error is 3.3%. Compared with the traditional process in the production line, the tensile strength of automobile wheel rim formed by using the neural network model optimum process increases by 53%.
作者
潘玲
李明
PAN Ling1, LI Ming2(1. Chongqing Industry Polytechnic College, Chongqing 401120, China; 2. Department of Materials Science and Engineering, Chongqing University, Chongqing 401331, Chin)
出处
《热加工工艺》
CSCD
北大核心
2018年第3期167-169,173,共4页
Hot Working Technology
基金
重庆市教育委员会科学技术研究项目(KJ122101)
关键词
钒钛改性镁合金
汽车轮圈
锻造工艺
神经网络
拉伸性能
magnesium alloy modified by vanadium-titanium
automobile wheel rim
forging process
neural network
tensile property
