摘要
动态鲁棒优化问题广泛存在于各个领域,且难以求解。动态鲁棒粒子群优化(PSO)算法是一种有效的求解方法。但是,现有算法存在全局搜索能力弱和无法对个体进行综合评价的问题。为有效求解动态鲁棒优化问题,在研究的基础上提出一种混合差分进化的动态鲁棒粒子群(DRPSO-DE)算法。该算法不仅使用差分进化(DE)算法的变异策略提升粒子群算法的全局搜索能力,还提出一种综合指标来对种群个体进行评价。此外,为提高动态鲁棒粒子群算法的搜索效率,采用一种基于排序的选择策略挑选最佳个体,并将它们用于指引种群进化。为验证DRPSO-DE的有效性,选取五个动态标准测试函数对其进行测试。从试验结果来看,所提出算法的整体性能要优于原有算法,能够有效求解动态鲁棒优化问题。
Dynamic robust optimization problems are widely found in various fields.They are hard to solve.Dynamic robust particle swarm optimization(PSO)algorithm is an essential methods in the field of dynamic optimization.However,existing algorithms have poor global search capability and cannot effectively evaluate the individual in the population.To effectively solve dynamic robust optimization problems,a dynamic robust particle swarm optimization integrated with differential evolution(DRPSO-DE)alogrithm is proposed in this study.In the DRPSO-DE alogrithm,a mutation strategy of the differential evolution(DE)alogrithm is used to improve the global exploration capability of the PSO,and a comprehensive performance metric is proposed to evaluate individuals.Additionally,a ranking-based selection strategy is employed to select individuals to guide the population evolution based on the proposed performance metric.To demonstrate the effectiveness of the DRPSO-DE,five dynamic robust optimization problems are selected.Experimental results indicate that the overall performance of the proposed algorithm is better than that of the original algorithm,and the DRPSO-DE is an effective method to solve dynamic robust optimization problems.
作者
杨霞
廖青
范勤勤
王维莉
YANG Xia;LIAO Qing;FANQinqin;WANG Weili(School of Logistics Research Center,Shanghai Maritime University,Shanghai 201306,China;School of Electronic Information and Electrical Engineering,Shanghai Jiao Tong University,Shanghai 200240,China)
出处
《自动化仪表》
CAS
2020年第8期30-35,共6页
Process Automation Instrumentation
基金
国家重点研发计划基金资助项目(2016YFC0800200)
国家自然科学基金资助项目(61603244)
中国博士后科学基金资助项目(2018M642017)。
关键词
动态优化
鲁棒优化
评价指标
粒子群优化
差分进化
鲁棒最优解
变异策略
选择策略
Dynamic optimization
Robust optimization
Performance indication
Particle swarm optimization(PSO)
Differential evolution
Robust optimal solution
Mutation strategy
Selection strategy
