300M超高强度钢铣削力模型构建及MOPSO优化  

Milling force model construction of 300M ultra-high strength steel and MOPSO optimization

作　　者：张昱 张昌明[1,2] 王运 蒋红元 张爽爽 ZHANG Yu;ZHANG Changming;WANG Yun;JIANG Hongyuan;ZHANG Shuangshuang(School of Mechanical Engineering,Shaanxi University of Technology,Hanzhong 723001,China;Shaanxi Key Laboratory of Industrial Automation,Hanzhong 723001,China)

机构地区：[1]陕西理工大学机械工程学院,陕西汉中723001 [2]陕西省工业自动化重点试验室,陕西汉中723001

出　　处：《兵器材料科学与工程》2021年第1期27-35,共9页Ordnance Material Science and Engineering

基　　金：国家自然科学基金(51505268);陕西省重点研发项目(2020GY-121);研究生创新基金项目(SLGYCX1924)。

摘　　要：用正交试验法对300M超高强度钢进行铣削加工,用直观分析和方差分析探究铣削力随铣削用量的变化规律,建立铣削力的经验指数模型与GA-BP神经网络预测模型,用多目标粒子群优化算法基于铣削力和材料去除率优化铣削参数。结果表明:300M超高强度钢铣削力随铣削速度增大和每齿进给量降低得到有效改善;经优化后的BP神经网络模型预测误差显著降低,两种预测模型对铣削力均有较高预测精度,但后者误差相对较低;使用经优化后的参数,铣削力有效改善。The orthogonal experiment method was used to study the milling process of 300 M ultra-high strength steel.The change rule of milling force with milling amount was explored by using the visual analysis and variance analysis.The empirical index model and GA-BP neural network prediction model of milling force were established.The milling parameters were optimized by using multi-objective particle swarm optimization algorithm based on milling force and material removal rate.The experimental results show that the milling force of 300 M ultra-high strength steel is effectively improved with the increase of milling speed and the decrease of feed per tooth.The prediction error of the optimized BP neural network model significantly decreases.The two prediction models have high prediction accuracy for milling force while the latter error is relatively low.The milling force of the optimized milling parameters is effectively improved.

关 键 词：300M超高强度钢 铣削加工 铣削力 预测模型 MOPSO算法 

分 类 号：TG548[金属学及工艺—金属切削加工及机床]