Algorithms for vendor managed inventory routing and machine scheduling models

1.简介/概述 这项提议旨在促进物流、运输和运筹学领域知识的发展，目的是向学术界和实践者传播这方面的知识。本研究方案将重点研究供应链管理中的两个重要问题：1)供应商管理库存问题；2)基于两个智能体的调度问题。供应商管理库存(VMI)是供应链通过降低成本和改善服务来提高其绩效的方法之一。这是沃尔玛、壳牌化工等零售商采用的成功商业模式之一。具有两个代理的作业调度问题在电信、综合多媒体服务、铁路轨道、制造业、维修计划等许多商业领域都存在。 2.建议的模式： I.VMI无限时间范围模型：该模型假定每个客户的平均需求是已知的且恒定的。文献中的大多数论文都将基准周期时间作为决策变量。然而，出于交通规划的目的，从业者更喜欢方便的基本周期时间，如一周、两周或一个月。我们的模型将假设一个固定的基本周期时间，这将由经理指定。 VMI有限时间范围模型：在假设产品需求与时间相关的情况下，这类模型最小化总库存和运输成本。我们将重点研究单仓库、多客户、时间依赖需求和订单至上补货策略模型。 基于两个智能体的时间切换调度问题：现有文献中基于两个智能体的调度模型没有考虑不同智能体之间的切换。每当在处理一个代理的作业之后处理另一个代理的作业时，就会出现切换时间。切换时间的概念类似于调度文献中使用的建立时间。 3.建议的方法： 我们将为集成模型开发基于启发式和元启发式的求解技术。这些解决技术将用C或C++编程语言编写，并将用基准问题实例进行实验。 4.方案的预期结果： 所提出的问题属于NP-Hard问题，求解该问题的最优解实际上非常困难。该程序将设计求解方法，为不同的模型找到接近最优的解。 5.通过提案进行培训 该项目将招收一名博士生和三名硕士生。博士生和硕士生将接受针对该问题建立数学和计算机模型的培训。 6.工作的预期意义： 拟议的研究有可能通过为管理者提供一个最大限度地减少生产、运输和库存成本的框架来改善供应链的运作。

1. Introduction/Overview This proposal aims to contribute towards the advancement of knowledge in the area of logistics, transportation and operations research with a purpose to disseminate this knowledge to both academics and practitioners. This research proposal will particularly focus on two important topics of supply chain management 1) vendor managed inventory problem and 2) two agents based scheduling problem. Vendor Managed Inventory (VMI) is one of the ways a supply chain can improve its performance by reducing cost and improving service. This is one of the successful business models used by Wal-Mart, Shell Chemicals and other retailers. The scheduling of jobs with two agents arises in many business areas such as telecommunications, integrated multimedia services, rail road tracks, manufacturing industry, maintenance planning etc. 2. Proposed Models: I. VMI Infinite time horizon model: This model would assume that mean demand for each customer is known and constant. Most of the papers in literature consider the base cycle time to be a decision variable. However, practitioners prefer a convenient base cycle time such as one week, two weeks or a month for transportation planning purposes. Our model would assume a fixed base cycle time which will be specified by the manager. II. VMI Finite time horizon model: This type of model minimizes the total inventory and transportation cost assuming a time dependent product demand. We will focus upon the single depot, multiple customers, time dependent demand and order-up-to-level replenishment policy model. III. Two agents based scheduling problem with switch over time: The existing literature on two agents based scheduling model does not consider switch over time between different agents. A switchover time arises whenever a job of one agent is processed after a job of another agent. The concept of switchover time is similar to the setup time used in the literature of scheduling. 3. Proposed Methods: We will develop heuristics and meta-heuristics based solution techniques for the integrated models. These solution techniques will be coded in C or C++ programming language and an experiment will be performed with benchmark problem instances. 4. Anticipated outcome of the program: The proposed problem falls under category of NP-hard problems for which getting optimal solution is practically very difficult. The program will design solution methods for finding near optimal solution for the different models. 5. Training to take place through the proposal One PhD student and three Master students will be recruited for this program. PhD and Master's students will be trained in developing mathematical and computer models for the problem. 6. Anticipated significance of the work: The proposed research has the potential to improve supply chain operations by providing managers with a framework to minimize the production, transportation and inventory cost.