集合种群生物地理学优化算法  

Metapopulation Biogeography-Inspired Optimization

作　　者：黄光球[1] 刘权宸 陆秋琴[1] 

机构地区：[1]西安建筑科技大学管理学院,西安710055

出　　处：《系统仿真学报》2014年第6期1217-1224,共8页Journal of System Simulation

基　　金：陕西省重点学科建设专项资金资助项目(E08001);陕西省科学技术研究发展计划项目(2013K11-17);陕西省教育厅科技计划项目(12JK0789)

摘　　要：为了快速求解大规模优化问题,基于集合种群理论构造出了可全局收敛的生物地理学优化算法。在该算法中,每个斑块对应着优化问题的一个试探解;采用正交拉丁方原理构造出了斑块的适宜度特征变量初始化算法,实现了对搜索空间的均衡分散性和整齐可比性覆盖;将局域种群的跳转、混融和静止行为以及斑块的突变和选择现象用于构造斑块的适宜度特征向量演变策略,以便使得斑块的适宜度指数要么保持原状不变,要么向好的方向转移,从而确保了整个算法的全局收敛性;在斑块演变过程中,斑块从一种状态转移到另一种状态实现了对优化问题最优解的搜索。应用可归约随机矩阵的稳定性条件证明了本算法具有全局收敛性。测试结果表明本算法是高效的。To solve large-scale optimization problems（OP）, a metapopulation biogeography-inspired optimization algorithm with global convergence was constructed based on metapopulation theory. In the algorithm, each patch is just an alternative solution of OP; the principle of orthogonal Latin squares was used to construct an initializing algorithm of suitability index vectors on each patch so as to cover search space with balance dispersion and neat comparability; the jump, amalgamation, stillness behaviour of local populations and mutation and selectivity phenomena of patches were used to construct evolution policies of patches so as to ensure patch suitability index（PSI） of each patch to keep either to stay unchanged or to transfer toward better states, therefore the global convergence was ensured; during evolution process of local populations, each patch transferring from one state to another realized the search for the optimum solution. The stability condition of a reducible stochastic matrix was applied to prove the global convergence of the algorithm. The case study shows that the algorithm is efficient.

关 键 词：函数优化 进化计算 生物地理学优化算法 集合种群 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]