摘要
利用蚁群算法的群体搜索策略,研究了基于蚁群算法的MAS多目标协调优化机制.对每个Agent的目标函数分配一群蚂蚁,使之在问题空间寻优,并对所有的优化解采用谈判机制进行协调,以产生多目标优化问题的Pareto折衷解.采用"误差率"和"空间矩阵"方法对算法的性能指标进行度量.用该方法求解两个典型的多目标优化测试函数,仿真结果表明所提出的方法可成功地解决MAS的多个目标函数的优化问题,收敛速度较快.
A mechanism of multiple objective coordinated optimization based on ant system for MAS is proposed by using colony searching strategy. A family of ants is assigned for each objective of each agent, by which the optimal solution is searched in solution space. Negotiation mechanism is applied to coordinate all the solutions. Performance is measured by usihg "error ratio" and "spacing" metrics. Multiple objective coordinated optimization based on ant system is applied to two typical multiple objective test functions. Simulation results show that this algorithm is able to solve the multiple objective optimization problems successfully and has fast convergence speed.
出处
《控制与决策》
EI
CSCD
北大核心
2007年第8期946-950,共5页
Control and Decision
基金
国家自然科学基金重点项目(60534040)
关键词
蚁群算法
多目标协调优化
谈判机制
Ant system
Multiple-objective coordinated optimization
Negotiation mechanism