差分细菌觅食算法

BACTERIA FORAGING OPTIMISATION ALGORITHM BASED ON DIFFERENTIAL EVOLUTION

为解决细菌觅食优化算法BFO(bacteria foraging optimization)迭代过程中因种群多样性损失较快而易陷入局优的问题,将差分进化思想和BFO结合。首先在BFO的趋向性操作和繁殖操作中使用差分策略更新细菌位置,从而保证群体内的多样性不会随着算法迭代的进行而过快降低。其次是对算法中细菌位置更新方式的改进,通过逐维更新每一个维度信息而非一次更新一个细菌所有维度信息的方式,充分利用每一次位置改变带来的有利信息,从而大幅提高了算法寻找到全局最优值的效率。与其他4个算法对10个标准优化函数的测试结果对比表明,改进后的算法在寻得最优值的精度、效率、稳定性方面表现更好。

Bacterial foraging optimisation（ BFO） algorithm is easy to fall into local optima in iteration process due to rapid loss in population diversity. To solve this problem,we combine the differential evolution idea with BFO. First,in operation of chemotaxis and reproduction of BFO,we employ differential policy to update bacteria locations so as to ensure that the diversity within the population will not degrade too fast along with algorithm iteration process. Secondly,we improve the way of bacteria locations update in the algorithm,by updating the dimensionality information per dimension one by one rather than updating all the dimensionalities information of a bacterium once,we make full use of the favourable information brought by every location change,therefore greatly improve the efficiency of the algorithm in searching global optimum value. It is demonstrated by contrasting the testing results of the proposed algorithm with other four algorithms on ten benchmark optimisation functions that the improved algorithm performs better in terms of accuracy,efficiency and stability of the searched optimum value.