摘要
提出一种改进的多目标微粒群优化算法来求解人力资源分配问题。通过对种群进行正交初始化,保证了个体在整个可行解空间上的均匀分散,使得算法能够在整个可行解空间上进行均匀搜索;通过基于网格技术的外部存档非劣解删选策略,有效地保留了逼近Pareto前沿的非劣解;引入一种广义的学习策略来提升粒子向Pareto前沿收敛的概率。实验结果表明,提出的多目标微粒群算法能有效地解决多目标人力资源分配问题,具有较好的应用价值。
This paper proposed a novel multi-objective particle swarm optimization algorithm for solving the human resource allocation problem.It ensured that the individual equably dispersing in the feasible solution space by using the population quadrature initialization,effectively retained approximation Pareto off-the-press non-inferior solution by using external archive non-inferior solution delete selected strategy based on the grid technology,and promoted the probability of particles convergence to the Pareto frontier by introducting a generalized learning strategies.Results of the numerical experiment show that the proposed algorithm is effective and useful in solving the human resource allocation problem,and has good application value.
出处
《计算机应用研究》
CSCD
北大核心
2011年第9期3338-3340,共3页
Application Research of Computers
基金
鲁东大学基金资助项目(032812)
关键词
微粒群算法
多目标优化
人力资源分配问题
particle swarm optimization algorithm
multi-objective optimization
human resource allocation problem
