基于粒子群算法的加工参数多目标优化技术研究  

Technical Study on Multi-objective Optimization of Processing Parameters Based on Particle Swarm Algorithm

作　　者：赵静[1] 李丽[2,3] 王国勋[1] 

机构地区：[1]沈阳理工大学机械工程学院,沈阳110159 [2]东北大学机械工程与自动化学院,沈阳110819 [3]沈阳城市建设学院交通与机械工程系,沈阳110167

出　　处：《机械制造》2014年第11期42-46,共5页Machinery

基　　金：国家863高技术研究发展计划项目(编号:SS2012AA041303)

摘　　要：在数控加工中,为了尽可能提高生产效率和降低生产成本,采用粒子群优化算法对加工参数进行多目标优化。以切削速度、切削宽度和每齿进给量为决策变量,以加工时间和成本为目标函数,并以机床性能、刀具参数、工件质量等为约束条件,建立优化模型。采用罚函数法对约束条件进行处理,将多目标问题转化为单目标优化问题进行求解。为解决粒子群优化算法优化效果受参数影响较大的问题,提出了参数自适应协同粒子群优化算法(WCVPSO),算法参数按照一定规律变化,提高了优化算法的精度和收敛速度。实际加工试验表明,提出的优化方法提高了加工效率,降低了加工成本。In order to maximize production efficiency and reduce production costs in CNC machining, particle swarm is used for multi-objective optimization of processing parameters. It takes the cutting speed, cutting width and the feed engagement as decision variables, the processing time and costs as the objective function, and the machine tool performance, tool parameters and workpiece quality as constraints to establish the optimization model. It also adopts penalty function method to deal with the constraints, and solve them by translating the multi-objective issue into single objective issue for optimization. Aiming at the issue that the optimization effect of particle swarm optimistic algorithm has suffer from the major impact of the parameters, a parameter adaptive and cooperative particle swarm optimization （WCVPSO） is proposed where the algorithm parameters are changing according to certain regular pattern to improve the accuracy of optimistic algorithm and convergence rate. The actual processing test demonstrates that the proposed optimistic method can improve the working efficiency and reduce the processing costs.

关 键 词：多目标优化 粒子群算法 加工参数优化 参数自适应 

分 类 号：TH161[机械工程—机械制造及自动化] TP273[自动化与计算机技术—检测技术与自动化装置]