基于改进蚁群算法的Hanoi给水管网优化设计  

Design of Hanoi Water Distribution Network Based On Improved Ant Colony Optimization

作　　者：乔俊飞[1] 佟磊[1] 潘广源[1] 

机构地区：[1]北京工业大学电子信息与控制工程学院,北京100124

出　　处：《控制工程》2013年第3期483-488,共6页Control Engineering of China

基　　金：国家自然科学基金重点项目(61034008);北京市属市管高等学校人才强教计划资助项目PHR(IHLB)201006103;北京市教育委员会科技发展计划项目(KZ201010005005)

摘　　要：给水管网优化设计是一个多变量的离散优化问题,设计方案直接影响到工程的投资。针对蚁群算法在优化给水管网问题时易陷入局部极小难以寻求最优解的缺陷,提出一种改进蚁群算法,通过蚁群分工机制与优化信息素增加值提高了蚁群算法的全局搜索能力。利用改进蚁群算法对汉诺塔管网进行优化设计,结果表明改进蚁群算法的设计方案在满足工程要求的同时造价最低。通过与其他优化方法相比,提出的改进蚁群算法在给水管网优化问题中具有较好的全局搜索能力。Optimal designing of the water supply networks is a muhi-variable optimization problem, the design have a direct impact on project investment. For the problem of easily getting in the local minimum and finding the optimal solution difficultly when the water distribution network are optimized by the Ant Colony Optimization （ACO）the paper proposes an improved ACO. By adding the division mechanism of ACO and optimizing increment of pheromone the global searching ability of ACO has been improved. Using the improved ACO to optimize the Hanoi water distribution network we get the result that the design obtained by improved ACO can meet the engineering requirements and has lowest cost at the same time. Comparing with other optimal methods, the improved ACO has the good global searching ability in the problem of water distribution optimization.

关 键 词：改进蚁群算法 分工机制 给水管网 优化设计 

分 类 号：TP27[自动化与计算机技术—检测技术与自动化装置]