Simultaneous recovery model for aircraft and passengers

Usually some unforeseen events make airlines to reconstruct their schedules. A mathematical model for airlines schedule recovery which recovers aircrafts and disrupted passengers simultaneously is presented in this study. Aircraft recovery decisions affect on passengers but disrupted passengers and recovering them were not explicitly considered in the most previous aircraft recovery models so recovery of these two resources - aircrafts and passengers - concurrently is one of our contributions.The modeling is based on defining the recovery scope as well as employing aircraft rotations and passengers’ itineraries instead of flights. These are two of our other contributions.Our model examines possible flight re-timing, aircraft swapping, ferrying, utilization of reserve aircrafts, cancellation, and passenger reassignment to generate an efficient schedule recovery plan.Model parameters are user-specific therefore it helps airlines to apply their policies in the model. Defining the recovery scope reduces the problem size and ensures that the schedule returns to normal within a certain time. The objective is in the form of cost minimization which involves three kinds of cost—operational aircraft recovery, flight cancellation, and delay as well as disrupted passengers. A data set with two disruption scenarios is used to test the proposed model. The computational results show that it is capable of handling the simultaneous aircraft and passenger recovery problem successfully.     

基于航班纯度的鲁棒性机型指派问题研究

机型指派问题(FAP)是指根据飞机舱位能力、运营成本、潜在收益及飞机可用性,将具有不同舱位容量的飞机指派给各定期航班的问题.它是飞机、机组排班问题的基础,是整个航班计划中的最核心环节.然而,单独的机型指派最优对整个航班计划而言,可能是次优的,甚或是不可行的.本文针对国内航班计划特点,从航班的角度,定义了航班纯度概念,并建立基于航班纯度的机型指派模型.最后,以国内某航空公司干线网络为实例,表明该研究具有很大的理论意义和实践价值.     

资源约束条件下多模式项目调度问题研究

本文建立了多模式资源约束下的项目调度问题MMRCPSP的优化模型。该模型是对传统的单模式资源约束下的项目调度问题RCPSP更接近实际的拓展，设计并编写了启发式遗传算法求解MMRCPSP模型，并用数值实例说明了模型的合理性和算法的有效性。     

Online algorithms for scheduling with machine activation cost on two uniform machines

INTRODUCTION In this paper we investigate the uniform ma-chines scheduling problem with machine activationcost. This problem has application in garment pro-duction of international trade and is motivated by thefollowing scenario. Import-export company is com-pared to scheduler in this model, and orders are jobs,which arrive one by one. And garment factories canbe regarded as machines. Since jobs should be fin-ished on time, scheduler will choose a reasonablenumber of machines to make the…     

有保证调度算法的研究

有保证的调度对硬实时系统是必须的 ,这是因为系统中每个进程必须满足它们的截止时间 ,否则后果不堪设想 .本文讨论了两种有保证的调度算法 ,它们是分支限界法和回溯法 ,给出了用这两种算法寻找最优解的过程 ,提出逻辑结点和物理结点的概念 .通过对多个进程集合寻找最优解的实验 ,对两种算法在产生的结点个数、比较次数、搜索树的拓扑结构及占用的内存空间等方面进行了比较     

基于蚁群系统的网格资源分配与任务调度

为了解决基于启发式算法的资源分配和任务调度过程中由于没有考虑任务间的相互依赖关系而出现的任务死锁问题,提出了一种基于蚁群系统的改进算法.首先阐述了如何将分配调度问题映射到任务资源分配图的优化选择问题上和如何将信号量机制引入到最优任务资源分配图中来解决死锁问题.其次说明了基于蚁群系统如何利用网格信息素系统模型实现该算法,涉及任务资源分配图的构造,以及通过蚁群的正反馈和分布式并行计算机制优化任务资源分配图.最后模拟试验结果说明所提出的算法可以有效地解决网格中任务死锁问题.     

多处理机系统任务调度研究

目前大规模的并行分布多处理机系统中,调度算法好坏直接影响计算系统的高性能计算潜力能否发挥,调度的目的就是如何分配资源使系统性能最优。本文主要讨论分布式多处理机系统进行任务调度时的关键问题,包括问题模型的描述,调度策略,常用算法,评估标准,数据平台以及该问题的发展趋势。     

JBPM在调度运行系统中的应用

随着信息化建设的发展,网上协同办公日趋重要,工作流技术在协同办公系统中的应用也日益普及。首先介绍了工作流开源产品Jboss jBPM,然后给出了基于jPDL的图形化流程定义工具T-Jdpl的设计与实现思想,最后结合调度运行系统,利用JBPM和T-Jpdl实现了一个业务流程。     

基于蚁群优化和遗传操作的混合方法

尽管蚁群优化算法（ACO）在优化计算中已得到了很多应用,但在进行大规模优化时,其收敛时间过长仍是应用该算法的一个瓶颈．为了确保资源利用完成时间最小化和完成用户指定的最终期限延迟最小化,找到一个优化的调度方法,在计算网格中针对资源分配和调度提出了基于蚁群优化和遗传操作的混合方法．     

基于学习遗忘作用的人员调度实验设计及其软件开发

建立了学习遗忘作用与人员工效的关系模型。采用遗传算法进行调度求解,并在Matlab环境下完成了考虑学习遗忘作用的人员调度的软件开发,设计了相关实验,通过实验内容可以实现多任务人员的调度,并表明调度过程中的学习遗忘现象。通过对参数和约束条件的设置,模拟了不同生产类型,并产生人员在不同生产模式下的调度策略。