改进细菌觅食算法在高维优化问题中的应用

发布时间：2019-06-15 03:27

【摘要】：针对以往细菌觅食优化算法自适应步长公式经验性参数过多、无法真正实现自适应的缺点,提出了改进的步长公式,使步长仅与细菌个体当前的进化代数和所求解问题的寻优范围有关,真正实现步长的自适应;其次,将混沌思想和差分进化思想与细菌觅食算法结合,对算法初始化过程和寻优过程进行改进,增加群体多样性,避免算法因为早熟而陷入局部最优值;在高维问题的优化过程中,采用逐维更新细菌位置的方法,将整体问题分维处理,极大地提高了算法效率和精度。通过对多个标准测试函数在多维空间进行测试,表明改进算法在高维空间中寻优时速度快、精度高、求解过程简单可行,在寻得最优解的精度上比其他改进方案有显著提高。
[Abstract]:In view of the fact that there are too many empirical parameters in the adaptive step size formula of the previous bacterial foraging optimization algorithm, an improved step size formula is proposed, which makes the step size only related to the current evolutionary algebra of the bacterial individual and the optimization range of the problem solved, and truly realizes the step size adaptation. Secondly, chaos and differential evolution are combined with bacterial foraging algorithm to improve the initialization process and optimization process of the algorithm to increase population diversity and avoid the algorithm falling into local optimal value because of precocious. In the optimization process of high-dimensional problem, the fractal dimension of the whole problem is greatly improved by updating the bacterial position one by one. By testing several standard test functions in multidimensional space, it is shown that the improved algorithm has the advantages of fast speed, high precision, simple and feasible solving process, and the accuracy of finding the optimal solution is significantly higher than that of other improved schemes.
【作者单位】：兰州交通大学电子与信息工程学院;
【基金】：甘肃省科技计划项目:细菌觅食优化算法在多目标优化中的应用研究(1506RJZA084) 甘肃省教育厅科研项目:菌群优化算法的融合、改进及应用(1204-13) 甘肃省教育科学‘十二五’规划课题:细菌觅食优化算法在高维优化问题中的应用(GS[2015]GHB0907) 兰州市科技计划项目:细菌觅食优化算法在组合优化中的应用研究(2015-2-74)资助
【分类号】：TP18

【相似文献】