| 基于云模型理论的蚁群算法改进研究 | |
| Alternative Title | Improvement of ant colony algorithm based on cloud models theory |
| 段海滨; 王道波; 于秀芬; 朱家强; 北京8701信箱 | |
| Department | 国家863计划微波遥感技术实验室 |
| Source Publication | 哈尔滨工业大学学报
 |
| 2005 | |
| Volume | 37Issue:1Pages:115-119 |
| ISSN | 0367-6234 |
| Language | 中文 |
| Keyword | 云模型理论 蚁群算法 信息素 定性关联规则 |
| Abstract | 近几年优化领域中新出现的蚁群算法采用分布式并行计算机制,易于与其它方法结合,具有较强的鲁棒性,但易限于局部最优解是其最突出的缺点.云模型是一种新的实现定性概念和定量数值之间转换的有力工具,本文在介绍云模型理论的基础上,提出了一种利用云模型来有效限制蚁群算法陷入局部最优解的方法,最后将基于云模型理论的改进蚁群算法与未改进的蚁群算法分别应用于著名的CHC144TSP进行实验.改进后的蚁群算法采用升半正态云规则进行控制,并选取了500个云滴,仿真计算结果证明了该方法的有效性和可行性. |
| Other Abstract | Ant colony algorithm is a new category of parallelized bionic algorithm in optimization fields. It has strong robustness and is easy to combine with other methods in optimization, but it is easy to fall in local best. Cloud models theory is a powerful tool to convert numerical quantitative analysis to conceptual qualitative analysis. On the basis of introduction of cloud models, a novel qualitative strategy for improving the global optimization properties by use of cloud models is proposed. Finally, the computational experiments on CHC144 TSP have been performed. In the experiments，the rule of increasing half normal cloud is adopted in the improved ant colony algorithm，and the optimal number of cloud drops is 500. Simulation results show that this novel method has certain validity and feasibility. |
| Funding Project | 中国科学院空间科学与应用研究中心 |
| Document Type | 期刊论文 |
| Identifier | http://ir.nssc.ac.cn/handle/122/1848 |
| Collection | 微波遥感部 |
| Corresponding Author | 北京8701信箱 |
| Recommended Citation GB/T 7714 | 段海滨,王道波,于秀芬,等. 基于云模型理论的蚁群算法改进研究[J]. 哈尔滨工业大学学报,2005,37(1):115-119. |
| APA | 段海滨,王道波,于秀芬,朱家强,&北京8701信箱.(2005).基于云模型理论的蚁群算法改进研究.哈尔滨工业大学学报,37(1),115-119. |
| MLA | 段海滨,et al."基于云模型理论的蚁群算法改进研究".哈尔滨工业大学学报 37.1(2005):115-119. |
| Files in This Item: | ||||||
| File Name/Size | DocType | Version | Access | License | ||
| 2005371115.pdf（448KB） | 开放获取 | CC BY-NC-SA | Application Full Text | |||
Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.
Edit Comment