膜计算框架下的多目标烟花爆炸算法  

Multi-Object Fireworks Explosion Algorithm Under the Framework of Membrane Computing

作　　者：毛宗杨 陈韬伟[1] 余益民[1] 赵昆[1] MAO Zongyang;CHEN Taowei;YU Yimin;ZHAO Kun(Yunnan University of Finance and Economics Institute of Information, Kunming 650221, China)

机构地区：[1]云南财经大学信息学院,昆明650221

出　　处：《重庆理工大学学报（自然科学）》2018年第11期147-155,共9页Journal of Chongqing University of Technology：Natural Science

基　　金：国家自然科学基金资助项目(61461051;71462036);云南省教育厅一般项目(2015Y278)

摘　　要：膜系统是由多个膜组成的计算单元,具有并行计算和分布式的特性。受膜系统功能和处理化合物方式的启发,将膜系统与烟花爆炸算法结合,提出了膜框架下的烟花爆炸算法。该算法在表层膜中引入拥挤距离和非支配排序来提高算法的局部搜索效率,在外部档案中利用精英反向学习的方法增强算法的全局搜索能力。因此,将膜系统与烟花爆炸算法结合将更好地平衡算法的局部开采和全局勘探能力。仿真实验使用ZDT和DTLZ系列函数对膜计算框架下的烟花爆炸算法进行测试,结果表明:该算法所得的非支配解集更接近真实Pareto前沿,在多样性、收敛性、准确性等方面优于或部分优于其他算法。The membrane system is a computational unit composed of multiple membrane layers,with parallel computing and distributed characteristics.Inspired by the function of the membrane system and the way to process compounds,the membrane system was combined with the fireworks algorithm,and a fireworks algorithm under the membrane framework was proposed.The algorithm introduces crowded distance and non-dominated sorting in the surface film to improve the algorithm’s local search efficiency,and uses the elite reverse learning method in external files to enhance the algorithm’s global search capability.Therefore,combining the film system with the fireworks algorithm,a well-balanced algorithm for local mining and global exploration capabilities is built.The simulation experiment uses the ZDT and DTLZ series functions to test the fireworks algorithm under the membrane computing framework.The results show that the non-dominated solution set obtained by the algorithm is closer to the real Pareto front and is superior in diversity,convergence and accuracy which is better than the other algorithms.

关 键 词：膜计算 烟花爆炸 多目标优化 非支配排序 非支配解集 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]