基于小生境的菌群算法的改进

Advanced Bacterial Foraging Optimization Based on Niche Technology

将进化论中的小生境技术与菌群优化算法结合,提出了小生境菌群算法,通过细菌之间的距离找到具有相似距离的例子个体组成小生境种群,利用菌群优化算法在该种群里模拟细菌的行为,所有细菌经过小生境群体的进化后,找到最优的个体存入到下一代菌群中,直到找到满意的适应值。该算法克服了菌群算法收敛效率低,容易收敛到其他极值点的问题。用Shubert函数验证了该算法的性能,并与原算法进行了比较,通过更多的优化函数说明了该组合算法可以获得较好的解,极大地提高了收敛成功率,且代价较小。

The niche technology in Evolution Theory was proposed to combine with the bacterial foraging optimization （BFO） , and niched bacterial foraging optimization （NBFO） was formed. The NBF cluster formed from sample individuals whose distance is similar to that among bacteria; then the original BFO was use to emulate the actions of bacterial. Bacterial with the best fitness in the niche was collected as the new generation of bacterial and was then used in the iteration next time. The algo- rithm overcomes the problem of slow convergence and the problem that BFO will find other climax point as the minimal points. Shubert function was used to verify the properties of the NBFO and the cooperation between NBFO and BFO in other optimization functions shows that the combined algorithm can achieve better results with higher convergence rate and lower time orice.