中国科学院沈阳自动化研究所机构知识库
Advanced  
SIA OpenIR  > 信息服务与智能控制技术研究室  > 学位论文
题名: 人工蜜蜂群优化算法研究及应用
其他题名: Research and Application on Artificial Bee Colony Optimization Algorithm
作者: 邹文平
导师: 朱云龙
分类号: TP301.6
关键词: 人工蜜蜂群优化算法 ; 生物启发式计算 ; 群体智能 ; 粒子群优化
索取号: TP301.6/Z94/2012
学位专业: 机械电子工程
学位类别: 博士
答辩日期: 2012-05-29
授予单位: 中国科学院沈阳自动化研究所
学位授予地点: 中国科学院沈阳自动化研究所
作者部门: 信息服务与智能控制技术研究室
中文摘要: 通过对社会型生物的观察和对其群体行为的研究,学者们已经发明了许多群体智能算法。所谓群体智能就是在群体中每个个体都非常简单,但是这些简单个体组成的一个群体却涌现出了十分复杂的智能行为。这些复杂行为已经被广大的研究人员成功建模,发展出了各种算法并用于求解大量的复杂实际问题。     20世纪90年代研究人员开发出了基于蚂蚁的蚁群优化算法和基于鸟群、鱼群的粒子群算法,并且在二十年间被用于解决各种领域内的优化问题。然而在最近的10年里,蜜蜂表现出的智能行为引起了学者们极大的兴趣,进行了广泛深入的研究,并促使研究者发展出了新的算法。     2005年Karaboga基于蜜蜂的觅食行为提出了一种人工蜂群(Artificial Bee Colony, ABC)的优化算法。该算法自提出以来由于其概念简单、易于实现并且控制参数少。因此引起了国内外相关领域众多学者的关注和研究并在许多领域都得到了应用。但ABC算法也有许多不足。其在高维问题、离散问题与多目标问题等方面的开展还不够理想。在种群个体之间的结构上也存在可研究空间。 因此本论文的研究目的一方面针对原始人工蜜蜂群优化算法中存在的缺点和不足,给出其改进方法或提出新的优化模型,使之更为有效可靠;另一方面将提出的新模型新算法应用于实际工程领域,拓展人工蜜蜂群算法的应用领域。     研究的内容包括:离散二进制人工蜜蜂群优化算法、基于外部档案的多目标人工蜜蜂群优化算法、多群体协同进化人工蜜蜂群优化算法、基于不同拓扑结构的人工蜜蜂群优化算法以及各个算法在实际问题中的应用。具体的研究内容和创新性成果概括如下:     (1) 离散二进制人工蜜蜂群优化算法     原始人工蜜蜂群优化算法在连续搜索空间有良好的优化效果,但对于离散的搜索空间则不能直接应用,必须对标准ABC算法加以改进。因此提出了用于解决离散二进制问题的人工蜜蜂群优化算法(Binary Artificial Bee Colony, BABC)。BABC算法修改了原始ABC算法产生新解的方式,使其在多维度上同时进行更新。赋值新解与原始解之间维度距离为概率意义,表示个体飞行到邻居位置的概率。随后为了评估该算法性能,将该算法在一组测试函数上进行了测试并进一步应用于经典的0/1背包问题。实验将BABC算法与BPSO和BGA进行了比较,结果表明BABC算法在离散二进制问题上同样有良好的优化效果。     (2) 基于外部档案的多目标人工蜜蜂群优化算法     许多现实世界的问题需要同时优化多个目标。而这些目标可能是相互矛盾的。传统的经典多目标求解方法如线性规划法、加权求和法、目标规划法等有很多限制。随着进化算法的发展,过去二十年中许多多目标进化算法相继被提出并取得了良好的求解效果。人工蜜蜂群优化算法在单目标问题上已经表现出了良好的优化性能。但其在多目标优化领域还未有较为成熟的发展。因此针对人工蜜蜂群算法在多目标问题上的不足,提出了一种外部档案的多目标人工蜜蜂群算法(Multi-Objective Arti?cial Bee Colony, MOABC)。该算法将外部档案和拥挤距离等概念融入ABC算法中,用以保存找到的非支配解。并采用了广泛学习的方式使得ABC算法种群能够保持较好的多样性。实验结果表明MOABC算法在多目标问题求解的多样性、收敛距离、鲁棒性方面都要优于其它比较算法。     (3) 多群体协同人工蜜蜂群优化算法     传统的蜜蜂优化算法中更多的是模拟单一种群内蜜蜂个体之间的信息交互。这使得种群内蜜蜂易产生群体“趋同”的问题,不利于保持种群的多样性。其实在现实世界中真实蜜蜂的群体之间也有一定的交流,也会从其他蜜蜂群体获得一定的信息。因此受启发于蜜蜂种群间的信息交流现象,提出了一种多群体协同人工蜜蜂群优化算法(Cooperative Artificial Bee Colony Algorithm, CABC)。该算法通过种群间的协作不但避免了单一种群难以保持种群多样性的缺点,同时还大大加快了求解速度与准确性。实验结果表明CABC算法无论对单峰函数还是多峰函数,在求解精度、收敛速度等方面都要优于其他算法。为了进一步验证本算法的性能将该算法用于聚类分析问题,在多个数据集上进行了测试。结果显示CABC获得了令人满意的聚类效果。     (4) 基于不同拓扑结构的人工蜜蜂群优化算法     驱动群体型生物工作的的本质是社会交流。种群里的个体根据自己得到的知识相互学习,个体的移动总是越来越接近它们“更好”的邻居。群体型生物的社会结构影响着种群内个体间信息的流通与传播。在高度连接的社会网络中大多数个体彼此之间能够通信,因此信息能够快速通过网络传递。在优化求解中这也意味着能够更快地收敛到一个解,也容易陷入局部最优值。而弱连接网络收敛速度慢,但会保持较好的解的多样性,不容易陷入局部最优值。由此可见基于群体型生物行为的智能算法的性能非常依赖于社会网络的结构。因此我们将多种不同类型的拓扑结构应用于人工蜜蜂群优化算法中以试图寻找一种较好的拓扑网络结构能提高ABC算法的优化性能。经实验结果验证,确实存在某些网络使得ABC算法在大多数问题上的优化效果要好于原始的ABC优化算法。为了验证所得结论,将算法用于神经网络训练问题。在多个测试实例上进行了测试,结果显示基于冯诺依曼结构的人工蜜蜂群优化算法表现出较强的优势。.
英文摘要: Through the research of social biology and the group behaviors, scholars have developed many swarm intelligence algorithms. The so-called swarm intelligence algorithm is that each individual in the swarm is very simple, but the swarm consist of many simple individuals shows very complicated intelligent behaviors. Researchers have successfully simulated the complicated behaviors, and many algorithms are developed and applied to resolve lots of complicated practical problems.     In the 1990s, based on the ant, bird, and fish, researchers developed Ant Colony Optimization Algorithm, and Particle Swarm Optimization Algorithm, which are widely applied to resolve the optimization problems in many areas. However, in the recent decade, scholars show great interest in the swarm intelligent behaviors. And through wide and deep research, scholars have developed many new algorithms.     In 2005, Karaboga presented a new swarm intelligence algorithm, called Arti?cial Bee Colony (ABC) algorithm, which is based on the behaviors of real bees in ?nding food sources and sharing the information with other bees. Since ABC algorithm is simple in concept, easy to implement, and has fewer control parameters, it has attracted great attention of scholars from domestics and overseas, and it has been widely used in many fields. However, there are also some disadvantages of the ABC algorithm. It is not that satisficatory in the aspect of high dimensional problem, discrete problem, and multiobjective problem. There are some space to develop in the individual structure in the population.     Therefore, the purpose of this paper is to aim at the shortcomings of the original Arti?cial Bee Colony algorithm to give their improvements or new optimization models that make them more effective and reliable. On the other hand, the new algorithms will be applied to the some practical engineering fields to expand the application areas of Arti?cial Bee Colony optimization algorithms     The content of the thesis includes: Discrete Binary Arti?cial Bee Colony algorithm, Multi-Objective Arti?cial Bee Colony algorithm Based on External Archive, Multi-Swarm Cooperative Artificial Bee Colony Algorithm, Artificial Bee Colony Algorithm Based On Different Topology Structure; and their applications of the algorithms. The specific content and innovative research results can be summarized as follows:     (1) Though the original ABC algorithm performs very well in continuous space; for discrete search space, it cannot be applied directly. This paper presents a binary version of the ABC algorithm, namely the Binary Article Bee Colony (BABC), for solving discrete problems. BABC improved the method of producing new solution by the original ABC algorithm, which enabled all dimensions update at the same time. The meaning of the distance between the new solution and the original solution in each dimension is probalilistic. That means the probalilistic of a individual fly to its neighbor. So as to evaluate the performance of the BABC, we tested it on a set of benchmark functions and then applied BABC for solving classical 0/1 Knapsack Problem. We compared the performance of the BABC algorithm with that of discrete version of PSO and GA. The simulation results show that the proposed BABC has a signi?cant optimization performance for discrete binary problem.     (2) In the real world, many optimization problems have to deal with the simultaneous optimization of two or more objectives. In some cases, however, these objectives are in contradiction with each other. Some classical methods such as linear programming, the weighted sum method and the goal programming method have many limitations. Over the past two decades, with the development of the evolutionary algorithm, a number of multi-objective evolutionary algorithms (MOEAs) have been proposed and achieved good solution effect. Though ABC algorithm has been successful in solving single-objective optimization problems, it is not well developed in the field of multiobjective optimization. Therefore, as for the shortcomings of ABC algorithm solving multi-objective problems, we present a Multi-Objective Arti?cial Bee Colony (MOABC) algorithm based on external archive. The concept of “External Archive” and “Crowding Distance” are integrated into ABC to keep the nondominated solution vectors which has been found. And we use comprehensive learning strategy to ensure the diversity of population and to find a good spread of solutions. The experimental results illustated that MOABC is superior to some existing algorithms for multi-objective problems in the terms of solving diversity metric, convergence metric and result robustness.     (3) Traditional ABC algorithm imitates information transformation in a singel swarm. It will bring the problem of convergence phenomena in the bee colony, and it is harmful for diversity of the swarm. However, in the real world, there is information transformation between different bee colonies. Therefore, we proposed a multi-swarm Cooperative Artificial Bee Colony (CABC) algorithm based on information transformation in different bee colonies. The CABC algorithm divides the swarm into several subgroup. Each subgroup not only evolve itself, but also cooperate with other subgroup to evolve together. The CABC algorithm has overcome single swarm shortcomings of less diversity, and improved the solving precision. The experimental results showed that CABC is superior to some existing algorithms for multivariable, multimodal function optimization in the terms of convergence speed and solving precision. To futher prove the performance of the algorithm,we apply the CABC algorithm to solve clustering problem. The algorithm has been tested on several well-known real data sets. And the result shows that CABC can perform quite satisfactory in the convergence effect.     (4) The essence of driving colonial creature evolving is social communication。Individuals in swarm learn from each other, and it move towards the better neighbor. The social structure of colonial creature influences the information circulation and transmission in swarm. In the social network with high connectivity, most individuals can communicate with each other, therefore information can be fast transmitted through network. In optimization algorithm, highly interconnected populations would be converged fast and easily fall into local optimum. However, populations with fewer connections might perform better on diversity, but be converged slow. It can be seen that the performance of swarm intelligence algorithm depends on social network of populations. Therefore, we apply many different topology for Artificial Bee Colony to find a kind of structure which could improve the optimization performance of ABC algorithm. The experimental results illustated that there are some network structure which made the performance of ABC superior to the original ABC algorithm in most benchmarks. To verify the results, we apply Von Neumann ABC (VABC) for neural network training. The results showed that VABC de?nitely outperformed the original ABC.
语种: 中文
产权排序: 1
内容类型: 学位论文
URI标识: http://ir.sia.cn/handle/173321/9417
Appears in Collections:信息服务与智能控制技术研究室_学位论文

Files in This Item:
File Name/ File Size Content Type Version Access License
人工蜜蜂群优化算法研究及应用.pdf(6118KB)----限制开放 联系获取全文

Recommended Citation:
邹文平.人工蜜蜂群优化算法研究及应用.[博士学位论文].中国科学院沈阳自动化研究所.2012
Service
Recommend this item
Sava as my favorate item
Show this item's statistics
Export Endnote File
Google Scholar
Similar articles in Google Scholar
[邹文平]'s Articles
CSDL cross search
Similar articles in CSDL Cross Search
[邹文平]‘s Articles
Related Copyright Policies
Null
Social Bookmarking
Add to CiteULike Add to Connotea Add to Del.icio.us Add to Digg Add to Reddit
所有评论 (0)
暂无评论
 
评注功能仅针对注册用户开放,请您登录
您对该条目有什么异议,请填写以下表单,管理员会尽快联系您。
内 容:
Email:  *
单位:
验证码:   刷新
您在IR的使用过程中有什么好的想法或者建议可以反馈给我们。
标 题:
 *
内 容:
Email:  *
验证码:   刷新

Items in IR are protected by copyright, with all rights reserved, unless otherwise indicated.

 

 

Valid XHTML 1.0!
Copyright © 2007-2016  中国科学院沈阳自动化研究所 - Feedback
Powered by CSpace