有向动态拓扑混合作用力微粒群优化算法及可靠性应用
发布时间：2018-08-12 09:26
【摘要】：针对微粒群优化算法易陷入局部最优、出现早熟等不足,从作用力规则和种群拓扑结构两方面进行研究。提出一种混合作用力微粒群优化(Hybrid force PSO,HFPSO)算法,将算法的搜索过程划分为前期和后期两个阶段,分别构造引斥力规则和双引力规则,使算法搜索前期具有良好种群多样性、搜索后期有较高寻优精度。进一步将生物趋利避害的行为选择机制融入HFPSO算法,提出有向动态拓扑混合作用力微粒群优化算法,赋予微粒主观能动性使其靠近适应值较好微粒、远离适应值较差微粒,提出适应值驱动边变化的有向动态拓扑(Fitness-driven edge-changing unidirectional dynamic topology,FEUDT)结构,并将FEUDT结构与HFPSO算法以结构演化和算法进化同步进行的方式结合,进一步提升算法的优化性能。利用Benchmark函数对所提算法与标准PSO、搜索后期斥力增强型混合引斥力微粒群优化(LRPSO)算法进行性能对比测试,结果表明,所提算法具有较好的寻优能力和较快的收敛速度。通过桥式系统可靠性优化实例和供应商参与的某汽车产品子系统可靠性设计优化实例,验证了所提算法求解实际复杂优化问题的有效性。
[Abstract]:The particle swarm optimization (PSO) algorithm is easy to fall into local optimum and premature, so it is studied from two aspects: force rules and population topology. A hybrid force particle swarm optimization (Hybrid force PSO-HFPSO) algorithm is proposed. The search process of the algorithm is divided into two stages: the early stage and the later stage. The repulsive force rule and the double gravity rule are constructed, respectively, so that the algorithm has good population diversity in the early stage of search. In the later stage of searching, the accuracy of searching is high. Furthermore, the behavior selection mechanism of biological convergence and avoidance is incorporated into the HFPSO algorithm, and a hybrid force particle swarm optimization algorithm with directed dynamic topology is proposed, which gives the particle subjective initiative to make it close to the better adaptive value and away from the poor adaptive particle. A novel adaptive edge driven oriented dynamic topology (Fitness-driven edge-changing unidirectional dynamic topology FEUDT) structure is proposed, and the FEUDT structure is combined with the HFPSO algorithm in the way of structure evolution and algorithm evolution synchronization to further improve the optimization performance of the algorithm. The Benchmark function is used to compare the performance of the proposed algorithm with that of the standard PSOs, and the performance of the (LRPSO) algorithm is compared with that of the (LRPSO) algorithm. The results show that the proposed algorithm has better optimization ability and faster convergence speed. The effectiveness of the proposed algorithm for solving the complex optimization problem is verified by the reliability optimization examples of the bridge system and the reliability design of a vehicle product subsystem with the participation of the supplier.
【作者单位】： 燕山大学河北省工业计算机控制工程重点实验室;燕山大学河北省重型机械流体动力传输与控制重点实验室;先进锻压成型技术与科学教育部重点实验室(燕山大学);
【基金】：国家自然科学基金(51405426,51675460) 河北省自然科学基金(E2016203306)资助项目
【分类号】：TB114.3;TP18
本文编号：2178637
[Abstract]:The particle swarm optimization (PSO) algorithm is easy to fall into local optimum and premature, so it is studied from two aspects: force rules and population topology. A hybrid force particle swarm optimization (Hybrid force PSO-HFPSO) algorithm is proposed. The search process of the algorithm is divided into two stages: the early stage and the later stage. The repulsive force rule and the double gravity rule are constructed, respectively, so that the algorithm has good population diversity in the early stage of search. In the later stage of searching, the accuracy of searching is high. Furthermore, the behavior selection mechanism of biological convergence and avoidance is incorporated into the HFPSO algorithm, and a hybrid force particle swarm optimization algorithm with directed dynamic topology is proposed, which gives the particle subjective initiative to make it close to the better adaptive value and away from the poor adaptive particle. A novel adaptive edge driven oriented dynamic topology (Fitness-driven edge-changing unidirectional dynamic topology FEUDT) structure is proposed, and the FEUDT structure is combined with the HFPSO algorithm in the way of structure evolution and algorithm evolution synchronization to further improve the optimization performance of the algorithm. The Benchmark function is used to compare the performance of the proposed algorithm with that of the standard PSOs, and the performance of the (LRPSO) algorithm is compared with that of the (LRPSO) algorithm. The results show that the proposed algorithm has better optimization ability and faster convergence speed. The effectiveness of the proposed algorithm for solving the complex optimization problem is verified by the reliability optimization examples of the bridge system and the reliability design of a vehicle product subsystem with the participation of the supplier.
【作者单位】： 燕山大学河北省工业计算机控制工程重点实验室;燕山大学河北省重型机械流体动力传输与控制重点实验室;先进锻压成型技术与科学教育部重点实验室(燕山大学);
【基金】：国家自然科学基金(51405426,51675460) 河北省自然科学基金(E2016203306)资助项目
【分类号】：TB114.3;TP18
【相似文献】
相关期刊论文 前2条
1 李军军;王锡淮;黄有方;肖健梅;;基于混合微粒群优化算法的配送中心选址问题求解[J];物流科技;2006年08期
2 ;[J];;年期
相关会议论文 前1条
1 苗广祥;牛玉广;陈向阳;张海萍;;基于微粒群优化算法模型的改进策略的研究[A];第二十六届中国控制会议论文集[C];2007年
相关硕士学位论文 前2条
1 菅立川;辐射反问题的微粒群优化算法[D];哈尔滨工业大学;2007年
2 周喜虎;微粒群优化算法[D];陕西师范大学;2012年
，本文编号：2178637
本文链接：https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2178637.html
