基于功率最小原理和遗传算法的钢轨万能轧制过程宽展预测  

Spread prediction in rail universal rolling based on the principle of minimum power and the genetic algorithm

作　　者：董永刚[1] 李鸿全[1] 张江涛[1] 

机构地区：[1]燕山大学机械工程学院,秦皇岛066004

出　　处：《塑性工程学报》2014年第6期31-37,50,共8页Journal of Plasticity Engineering

基　　金：教育部博士点基金资助项目(20070216006);国家自然科学基金资助项目(50775196);河北省自然科学基金基地专项资助项目(08B017)

摘　　要：为得到钢轨万能轧制过程轨头和轨底的精确宽展,以轨头和轨底宽展为待定参数,在坯料断面尺寸、设备参数、轧制工艺参数已知的基础上,基于上限原理推导得出包含轨头和轨底宽展系数的钢轨万能轧制过程总功率表达式,并应用MATLAB软件遗传算法工具箱编写M文件对钢轨万能轧制总功率进行单目标优化,得到轧制功率最小时的轨头和轨底宽展系数。为验证理论模型,对18kg/m轻轨进行万能轧制实验,并用DEFORM软件模拟其万能轧制过程,将优化后宽展系数与实验结果进行对比,两者误差小于2%。For obtaining the spread of rail universal rolling exactly ,the spread of rail head and rail basewas set as two undeter‐mined parameters ,the size of beam blank ,the device parameters and the processing parameters were known in advance ,and based on the upper‐bound principle ,the equation of total rolling power was expressed by the spread coefficient of head of rail and base of rail .Moreover ,the M documents based on the MATLAB were programmed and the total rolling power was optimized ,then the corresponding spread coefficients can be obtained when the total rolling power is the minimum value .For verifying this theoretical model ,the experiments of 18kg/m light rail universal rolling were conducted ,and the rolling process was simulated by the rigid‐plastic FE software DEFORM .By comparing the simulation results with the optimized results ,the error between two results is no more than 2% .

关 键 词：最小轧制功率 宽展 钢轨 万能轧制 优化 

分 类 号：TG335[金属学及工艺—金属压力加工]