摘要
针对高速干切滚齿过程中的工艺参数优化决策问题,提出一种基于加工工艺样本预测和多目标遗传优化算法的工艺参数优化决策方法。基于实际加工工艺样本集,以改进的多目标遗传算法(improved NSGA-Ⅱ)为主体模型,以最大刀具寿命、最小加工能耗为优化目标,以加工质量、加工时间为约束条件,利用遗传反向传播算法(GABP)神经网络建立关于加工优化目标的预测模型,将其作为多目标优化模型的适应度函数;通过DBSCAN算法获取待优化滚齿工艺问题的相似样本集,建立多目标优化问题输入区间;构建面向待优化滚齿工艺问题的多目标优化模型,迭代搜索最优工艺参数集。
A method was proposed to optimize and decide parameters in the situation of high speed dry hobbing,supported by example prediction and multi-objective genetic optimization algorithm.Based on actual processing sample sets,the subject model was a variant of NSGA-Ⅱ,the optimization goal was the maximum tool life and the minimum energy consumption.The improved GABP neural network was used to construct a prediction model for the processing optimization fitness function,and a similar instance set of the hobbing problem was obtained through the DBSCAN clustering algorithm,so as to establish multi-objective optimization constraints,and construct the multi-objective optimization model for optimizing and deciding process,which may search the optimal processing parameters iteratively.
作者
刘艺繁
阎春平
倪恒欣
牟云
LIU Yifan;YAN Chunping;NI Hengxin;MOU Yun(State Key Laboratory of Mechanical Transmission,Chongqing University,Chongqing,400030)
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
2021年第9期1043-1050,共8页
China Mechanical Engineering
基金
重庆市技术创新与应用发展专项(cstc2019jscx-mbdxX0041)。
关键词
高速干切
滚齿工艺参数
遗传反向传播算法神经网络
改进的多目标遗传算法
最大刀具寿命
最小加工能耗
high speed dry cutting
hobbing process parameter
genetic algorithm-back propagation(GABP)neural network
improved non-dominated sorting genetic algorithm-Ⅱ(NSGA-Ⅱ)
maximum tool life
minimum machining energy consumption
