多峰函数优化的改进群居蜘蛛优化算法

发布时间：2018-11-24 12:06

【摘要】：针对群居蜘蛛优化(SSO)算法求解复杂多峰函数成功率不高和收敛精度低的问题,提出了一种自适应多种群回溯群居蜘蛛优化(AMBSSO)算法。引入自适应决策半径概念,动态地将蜘蛛种群分成多个种群,种群内适应度不同的个体采取不同的更新方式,提高了种群样本多样性;提出回溯迭代进化策略,在筛选全局极值的基础上,根据进化程度执行回溯迭代更新,保证了算法全局寻优能力。高维多峰函数仿真结果表明,同SSO算法、PSO算法等优化算法相比,AMBSSO算法具有较快的收敛速度和较高的收敛精度,尤其适用复杂高维多峰函数优化问题。
[Abstract]:In order to solve the problems of low success rate and low convergence precision of colony spider optimization (SSO) algorithm, an adaptive multi-colony backtracking (AMBSSO) algorithm is proposed. By introducing the concept of adaptive decision radius, the spider population is dynamically divided into several populations, and the individuals with different fitness within the population adopt different renewal methods to improve the diversity of population samples. A backtracking iterative evolution strategy is proposed. Based on the selection of global extremum, the backtracking iterative update is performed according to the degree of evolution, which ensures the global optimization ability of the algorithm. The simulation results of high dimensional multimodal function show that compared with the optimization algorithms such as SSO algorithm and PSO algorithm, AMBSSO algorithm has faster convergence speed and higher convergence precision, especially suitable for complex high dimensional multimodal function optimization problems.
【作者单位】：晋中学院信息技术与工程学院;山西大学物理电子工程学院;
【基金】：教育部高等学校教学指导委员会项目(No.JZW-14-JW-09) 山西省高等学校教学改革项目(No.J2014108) 山西省科技攻关计划项目(No.20110321025-02) 晋中学院教学改革项目(No.ZL2016jg04)
【分类号】：TP18

【相似文献】