基于改进遗传算法的多模型加工路径规划  

Multi-model Machining Path Planning Based on Improved Genetic Algorithm

作　　者：雷伟军[1] 程筱胜[1] 戴宁[1] 郭保苏[1] 李向佳[1] 

机构地区：[1]南京航空航天大学机电学院,南京210016

出　　处：《机械工程学报》2014年第11期153-161,共9页Journal of Mechanical Engineering

基　　金：国家科技支撑计划(2012BAI07B04);国家自然科学基金(51205192;81271181);国家高技术研究发展计划(863计划;SS2013AA040801-02)资助项目

摘　　要：在数控加工中,为提高整体的加工效率,往往将多个模型在一块坯料中进行排布,然后进行整体加工规划。为解决多模型整体加工中的路径规划问题,提出对模型及其连接点进行多次调整的方法。在将模型位置简化为加工轮廓的几何中心点的基础上,采用遗传算法对加工轮廓的几何中心点进行排序,确定整体的最短加工顺序。然后在保证加工模型间不发生干涉的情况下,通过对模型的姿态和加工起始点的位置进行调整,进一步缩短模型间的空间跳刀长度。为解决遗传算法易于早熟和难以跳出局部最优解等问题,采用父子代参与竞争和自适应遗传算子等方式对遗传算法进行改进。试验结果表明,该方法能有效地缩短多模型加工的整体路长,并且改进的遗传算法具有很好的收敛效果。In order to increase the total machining efficiency in NC machining,multiple models are machined in a piece of blank frequently.To solve the machining path planning of multiple models,a method by adjusting the machining model and its connected points several times is presented.On the basis of simplifying the machining contour to the geometrical center of the machining model,genetic algorithm is adopted to plan the general machining route of the multiple models.On the situation of avoiding the interference between machining models,the posture and the start point of the machining model are transformed to shorten the machining length between adjacent models.To solve the problems of genetic algorithm premature and hard to jump out of local optimal solution,parent and child are set to participate in competition and adaptive genetic operators is adopted to improve the genetic algorithm.The experimental results show that this method can effectively minimize the overall processing length and the improved genetic algorithm has great convergence results.

关 键 词：多模型加工 遗传算法 路径规划 

分 类 号：TH164[机械工程—机械制造及自动化] TP391[自动化与计算机技术—计算机应用技术]