In order to facilitate solution, a complex problem is normally decomposed into many small sub-problems during product development process. Teams are formed to resolve each sub-problem. The original problem is resolved...In order to facilitate solution, a complex problem is normally decomposed into many small sub-problems during product development process. Teams are formed to resolve each sub-problem. The original problem is resolved from solutions of sub-problems. Ideally, sub-problems are not only mutually independent but also inherent parameters of original problem. Solution of original problem can be directly derived from the collection of solutions from simplified sub-problems. In practice, the degree of interdependency is indeed reduced, sub-problems are neither totally independent nor all inherent parameters of original problem. This paper discusses team coordination under this condition and design solution from each team, which not only satisfies total requirements but also is an optimal one. The suggested optimized constraint decomposition method will insure workable Pareto solution.展开更多
为了提高物流服务优化组合的动态性、可靠性与用户满意度,本文提出了一种基于全局服务质量（quality of service,QoS）约束分解的能够感知领域质量与资源需求的物流服务优化组合方法.该研究工作首先把学习机制引入人工蜂群算法（artific...为了提高物流服务优化组合的动态性、可靠性与用户满意度,本文提出了一种基于全局服务质量（quality of service,QoS）约束分解的能够感知领域质量与资源需求的物流服务优化组合方法.该研究工作首先把学习机制引入人工蜂群算法（artificial bee colony algorithm,ABC）,形成了具有自主学习能力的改进型人工蜂群算法（LABC）;之后,应用学习人工蜂群算法（LABC）将全局QoS约束分解成每个物流子任务需要满足的局部QoS约束,从而将QoS感知的物流服务优化组合这一全局优化问题转化成以领域质量为依据的局部最优服务选择问题;其次,在物流服务流程执行的过程中,在感知物流任务节点对资源需求的前提下,为每一个物流任务节点选择一个具有最优领域QoS的物流服务;与已有的研究工作相比,该方法能够实现物流服务动态可靠的优化组合.最后,通过模拟实验验证了本文所提出的方法是可行有效的.展开更多
基金Supportedby 86 3/CIMS (No .2 0 0 1AA4 1114 0 )andtheNationalNaturalScienceFoundationofChina (No .6 0 10 4 0 0 8)
文摘In order to facilitate solution, a complex problem is normally decomposed into many small sub-problems during product development process. Teams are formed to resolve each sub-problem. The original problem is resolved from solutions of sub-problems. Ideally, sub-problems are not only mutually independent but also inherent parameters of original problem. Solution of original problem can be directly derived from the collection of solutions from simplified sub-problems. In practice, the degree of interdependency is indeed reduced, sub-problems are neither totally independent nor all inherent parameters of original problem. This paper discusses team coordination under this condition and design solution from each team, which not only satisfies total requirements but also is an optimal one. The suggested optimized constraint decomposition method will insure workable Pareto solution.
