改进的多目标快速群搜索算法及其在桁架优化中的应用

发布时间：2018-04-29 22:20

本文选题：改进的多目标快速群搜索算法 + 桁架结构优化　；参考：《河北工程大学》2017年硕士论文

【摘要】：近年来,我国进行了大规模的基础设施建设,建筑行业的发展得到了强有力的推动。经济全球化的深入发展,给建筑行业的发展带来了众多机遇,同时也面临了巨大的挑战。人们对建筑结构的使用要求也更加严格。桁架结构具有施工简单、拆装方便、运输便利等优点,且刚度和整体稳定性好、抗震能力强,因此被广泛应用于各种领域。研究桁架结构的优化设计具有很好的现实意义。传统的结构优化设计模式为先根据经验判断,然后进行重新假定,这种传统方法效率低且需消耗大量的人力和时间,已不能满足复杂结构的优化需求。人们在研究和模拟生物觅食行为的基础上,提出了一种基于计算机编程软件的群体智能优化算法。该类算法由于其计算简单、参数设置少等优越性,已被广泛应用于机械设计、建筑结构设计、电力系统优化等领域。本文针对多目标快速群搜索优化算法(MQGSO)的不足之处进行了改进,提出了改进的多目标快速群搜索优化算法(IMQGSO),并尝试将该算法应用于多个桁架结构的优化设计中。本文对MQGSO算法改进的几个方面主要有:一,种群初始化时引入了混沌的思想,降低了算法初始化的随机性,提高了算法的收敛速度;二,约束处理方面引入半可行域的概念,充分利用最优解附近的不可行解的有价值信息,并保证算法搜索方向的正确性;三,引入比例阀值R,严格控制有利个体的比例,保证算法最终收敛于可行域内。本文首次将改进后的算法应用于桁架结构优化设计中,分别对3个桁架结构进行了多目标连续变量优化设计、2个桁架结构进行了多目标离散变量优化设计,并将优化结果与MQGSO算法及其他算法对比。结果表明,改进的快速群搜索优化算法收敛速度快、收敛精度高、Pareto Front分布均匀广泛,对桁架结构优化效果显著,能广泛应用于桁架结构多目标优化设计中。
[Abstract]:In recent years, our country has carried on the large-scale infrastructure construction, the construction industry development has obtained the powerful impetus. With the development of economic globalization, it brings many opportunities and challenges to the development of construction industry. The use of building structures is also more stringent. Truss structure has the advantages of simple construction, convenient disassembly and assembly, convenient transportation, good stiffness and overall stability, strong seismic capacity, so it is widely used in various fields. It is of great practical significance to study the optimal design of truss structures. The traditional structural optimization design pattern is judged first by experience and then re-assumed. This traditional method is inefficient and requires a lot of manpower and time, so it can not meet the optimization requirements of complex structures. Based on the research and simulation of foraging behavior, a swarm intelligence optimization algorithm based on computer programming software is proposed. This kind of algorithm has been widely used in mechanical design, building structure design, power system optimization and other fields because of its advantages such as simple calculation, less parameter setting and so on. In this paper, the shortcomings of multi-objective fast group search algorithm (MQGSO) are improved, and an improved multi-objective fast group search algorithm (IMQGSOO) is proposed, and the algorithm is applied to the optimization design of multi-truss structures. In this paper, we improve the MQGSO algorithm in several aspects: first, we introduce chaos in population initialization, reduce the randomness of initialization, and improve the convergence speed of the algorithm; second, we introduce the concept of semi-feasible domain in constraint processing. It makes full use of the valuable information of the infeasible solution near the optimal solution and ensures the correctness of the search direction of the algorithm. Thirdly, the proportional threshold R is introduced to strictly control the proportion of the beneficial individuals, so as to ensure that the algorithm will converge in the feasible region. In this paper, the improved algorithm is applied to the optimization design of truss structures for the first time. The multi-objective continuous variable optimization design for three truss structures and the multi-objective discrete variable optimization design for two truss structures are carried out respectively. The optimization results are compared with MQGSO algorithm and other algorithms. The results show that the improved fast group search algorithm has the advantages of fast convergence speed, high convergence precision and uniform and wide Front distribution. It can be widely used in the multi-objective optimization design of truss structures because of its remarkable effect on the optimization of truss structures.
【学位授予单位】：河北工程大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TU323.4

【参考文献】