摘要
传统的占线优化理论分析假设对未来信息一无所知,根据过去和现在的信息进行决策。然而,在现实中有些信息是可以通过统计数据来估计得出或者通过实验方式获得,因此在本文中放松该假设,知道未来的部分信息。比如在汽车更新决策中,未来可能投入市场的汽车更新的某些重要参数(如购买成本和行驶成本)可以通过上述方法获得,决策者该如何决策何时更新汽车?优化目标是使竞争比最小,即汽车更新的占线成本和离线成本的比值最小。另外,汽车类大宗设备退出市场时还具有一定的残值,可以通过二手车市场交易来回收部分成本,因此本文将折旧因子引入了模型,给出了一个占线参数策略,得出了相应的竞争比,证明了折旧因子的引入改善了竞争比。
Traditional online algorithm and competitive analysis suppose that online player must make decisions based on past events without any future information.In fact,some information can be obtained by statistical method which would be terrible waste to ignore it.Therefore,this paper assumed that the online player has partial knowledge in the future.For example,some important parameters for automobile replacement are presented in the beginning,such as the purchase costs and the running costs.Then when must the decision-maker decide to replace automobile in an online fashion?The optimal aim is to minimize the ratio of the online costs and the offline costs.In this paper,the depreciation coefficient is also introduced in the model for the replaced automobile has some salvage.Furthermore,aparameter strategy is given and the competitive ratio is obtained,which is better than the result obtained by the original strategy.
出处
《系统工程》
CSSCI
CSCD
北大核心
2015年第3期118-122,共5页
Systems Engineering
基金
国家自然科学基金资助项目(70971008)
教育部人文社会科学研究基金资助项目(09YJC630008)
关键词
占线问题
有限信息
设备更新
竞争分析
Online Problem
Limited Information
Equipment Replacement
Competitive Analysis
