摘要
传统多目标优化问题通常是以加权或约束方式将其转换为单一目标,未能反映多目标间的复杂关系,不利于随时根据需求作出有效的决策。为了更合理地确定切削用量,采用多目标粒子群算法首先求得问题的pareto最优前沿,经过后期多准则决策得到满足不同要求下的最优方案。采用这种先寻优后决策的方法,能有效弱化先验知识不足的影响,较传统多目标优化方法更为实用有效。并经与多目标遗传算法比较,多目标粒子群算法具有更优良的性能。
Traditional multi-objective optimization,through simply convening to a single goal, often fails to reflect the more complex relationship between goals,and decision-make effectively at any time on demand. For selecting optimum cutting parameters,MOPSO（Multi-Objective Particle Swarm）is applied to obtain the pareto optimum front,and then the best answer is got according to multi-criteria decision. The method, making decision after searching optimum solutions, is more applicable and effective and can weak designer＇ s transcendental information deficiency problem. Compared with M OGA （Multi-Objective Genetic Algorithm）,MOPSO showed better performance.
出处
《组合机床与自动化加工技术》
北大核心
2010年第3期27-29,33,共4页
Modular Machine Tool & Automatic Manufacturing Technique
关键词
切削用量
多目标粒子群算法
多准则决策
优化
cutting parameters
Multi-Objective Particle Swarm
multi- criteria decision
optimization
