基于混合细菌觅食算法的多目标优化方法  

Multi-objective Optimization Method Based on Hybrid Swarm Intelligence Algorithm

作　　者：杨尚君[1] 王社伟[1] 陶军[1] 刘学[1] 

机构地区：[1]空军航空大学,吉林长春130022

出　　处：《计算机仿真》2012年第6期218-222,共5页Computer Simulation

摘　　要：针对多目标优化求解过程中多个目标相互制约难以求解的特点,为了提高多目标优化问题的求解速度和精度,并保持最优解的多样性,提出了一种用粒子群改进的混合细菌觅食多目标优化算法。将粒子群算法的寻优更新机制作为细菌觅食算法中趋向性操作的更新机制,将所求得非劣解的拥挤度作为寻优迭代过程中最优值的选取条件。与细菌觅食算法和NS-GA-Ⅱ算法的仿真结果表明,在对多目标测试函数ZDT1～ZDT4和ZDT6的求解过程中,该算法不仅能提高精度和快速地得到Pareto解集,并能有效地保持所求最优解的多样性。To improve solving speed and accuracy, and maintain the diversity of the optimal solution of the multi - objective optimization problem whis is difficult to solve because of targets＇ mutual restraint, we proposed a hybrid swarm bacteria foraging optimization algorithm. In this algorithm, the bacterial foraging algorithm＇s chemotaxis process was replaced by the particle swarm optimization＇s searching optimizing process, the optimal value strategy was determined according to the congestion degree. The simulation was compared with BFO and NSGA - I1 and the re- suits show that, in the process of solving multi - objective test functions ZDT1 - ZDT4 and ZDT6, the algorithm can quickly and effectively obtained the Pareto solution set, and effectively maintain the diversity of the obtained optimal solution set.

关 键 词：多目标优化问题 粒子群优化算法 细菌觅食算法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]