共查询到20条相似文献,搜索用时 15 毫秒
1.
《国际计算机数学杂志》2012,89(6):693-701
Vehicle Routing Problems (VRPs) considered in this paper involve a set of engineers operating from one base and a set of geographically distributed jobs or services to be performed by them under the given technological and temporal constraints. All engineers have different time-windows for their working and the technological constraints specify their competence to do jobs which can be performed within the given time-windows. A solution to a VRP is the scheduling of jobs among different engineers so that these jobs can be performed in minimum cost with all the temporal and technological constraints satisfied. The objective is to maximize the work done measured in terms of total number of jobs completed and minimize the total distance travelled by all the engineers. Three types of VRPs considered are under, critically and over resourced VRPs. VRPs belong to the class of NP-Complete problems. This paper investigates experimentally the application of an iterative method Tabu Search (TS) [5] in solving all the three types of randomly generated VRPs. The performance of TS is measured in terms of the number of unallocated jobs left and the cost of the solution measured in terms of total distance travelled by all the engineers. Starting with a given initial solution, TS attempts to obtain a near optimal solution containing the minimum number of unallocated jobs and the minimum cost. It is compared with the Hill-Climbing (HC) algorithm also used to solve same VRPs. In almost all cases, TS performs better than HC. However, in some cases, HC is found to give better results in terms of number of unallocated jobs. 相似文献
2.
3.
4.
5.
混合遗传算法与模拟退火法 总被引:10,自引:0,他引:10
论文将适合全局搜索的遗传算法(GA)和适合局部搜索的模拟退火算法(SA)相结合,提出了混合GA-SA计算方法。一方面,算法采用混沌初始化,提高了初始群体的质量;另一方面,算法采用Gray编码以及动态自适应调节交叉概率和变异概率,提高了收敛速度,并有效防止种群早熟现象。实例验证了该算法的可行性和有效性。 相似文献
6.
基于模拟退火的混合遗传算法研究 总被引:17,自引:2,他引:17
针对常规遗传算法会出现早熟现象、局部寻优能力较差等不足,在遗传算法运行中融入模拟退火算法算子,实现了模拟退火的良好局部搜索能力与遗传算法的全局搜索能力的结合。经验证,该混合算法可以显著提高遗传算法的运行效率和优化性能。 相似文献
7.
8.
基于模拟退火遗传算法的控制系统优化设计 总被引:2,自引:0,他引:2
提出了一种基于模拟退火遗传算法的线性系统优化设计方法。该方法以控制系统的性能指标,包括瞬态指标和稳态指标及其组合为目标函数,实现了由传递函数描述的控制器的自动设计,而不必预选择特定的控制方案。遗传算法使用十进制数编码,配合使用模拟退火技术来得到更精细的调整。使用这种方法,不需要手工计算,就可以获得控制系统的最优性能。该设计方法还可以应用于非线性对象。 相似文献
9.
有别于传统的单目标方法,将带时间窗约束的车辆路径问题描述成为一个多目标最优化问题,并为之提出了一种多目标遗传算法。在算法中设计了擂台法则作为构造非支配集的方法,提出了可变爬山率的局部爬山法,并通过将组合种群分成多层非支配集来实现精英保留策略。实验结果表明,该算法能有效地求解车辆路径问题并且为决策者提供了强有力的决策支持。 相似文献
10.
11.
12.
并行遗行/模拟退火混合算法及其应用 总被引:4,自引:0,他引:4
1 引言人们常常应用随机优化方法,例如:遗传算法GA(Genetic Algorithms),模拟退火算法SA(Simulated Annealing),爬山算法HC(Hill Climbing),Tabu算法等,解决复杂的非线性函数优化问题。这些方法通常需要大量的计算,从而导致运行时间开销较大。随着计算机及网络技术的高速发展,在高性能计算平台上并行化随机优化方法成为当今研究领域的热门。特别是Beowulf PCs Cluster技术的成熟,为研究人员提供了 相似文献
13.
14.
15.
《国际计算机数学杂志》2012,89(3):305-324
It is well-known that the problem of MEG source localization can be cast as an optimization problem. So far, there have been many works in which various optimization methods were adopted for source localization. In this paper, we compare the performance of three typical and widely used optimization techniques for a specific MEG source localization problem. We first introduce a hybrid algorithm by combining genetic and local search strategies to overcome disadvantages of conventional genetic algorithms. Second, we apply the tabu search, a widely used optimization method in combinational optimization and discrete mathematics, to source localization. To the best of our knowledge, this is the first attempt in the literature to apply tabu search to MEG/EEG source localization. Third, in order to further compare the performance of the above algorithms, simulated annealing is also applied to MEG source localization problem. The computer simulation results show that our local genetic algorithm is the most effective approach to dipole localization, and the tabu search method is also a very good strategy for this problem. 相似文献
16.
基于扩张矩阵和GA的入侵检测新方法 总被引:1,自引:1,他引:0
提出了一种新的基于扩张矩阵和遗传算法理论产生最优检测规则的方法。该方法产生的规则简单、能够反映问题的本质。实验结果表明,它的检测效果优于同类的其它方法。 相似文献
17.
分布式实时多媒体的大量应用对能有效支持服务质量(QoS)的组播路由算法提出了迫切的要求,由于其NP-Complete特性,只能采用启发式算法。该文提出了一种基于PBIL(Population-BasedIncrementalLearning)进化算法的时延受限组播路由算法,该算法有效结合了遗传算法的进化特性与竞争学习算法的特点,实施简单,仿真表明它不但显著提高了收敛速度,而且能以较大概率收敛到最优解。 相似文献
18.
19.
20.
新型遗传模拟退火算法求解带VRPTW问题 总被引:3,自引:0,他引:3
为了克服现有遗传算法不能有效求解时间窗车辆路径问题的缺陷,提出了一种由遗传算法结合模拟退火算法的混合算法求解该问题,并与遗传算法进行了比较。该算法利用了模拟退火算法具有较强的局部搜索能力的特性,有效地克服了传统遗传算法的“早熟收敛”问题。实验结果表明,该算法具有计算效率高、收敛速度快和求解质量优的特点,是解决车辆路径问题的有效方法。 相似文献