论文总字数：47632字

摘 要

本文以我国公路货运由于信息的不对称导致的返程空车和货运配载率不高等问题，带来了居高不下的物流成本和社会资源的极大浪费为研究背景，选取车货匹配信息平台中的典型平台A平台作为研究对象。针对A平台自动搜索为主的车货匹配方式下返回的查询列表只以距离和发布时间倒序排列，带来的匹配结果不够精准和有效的现状，对A平台进行车货匹配模型方面的研究。

首先对国内外有关双边匹配和车货匹配的研究进行文献综述，从算法、决策方法与模型以及双边匹配在不同领域和市场的应用方面对双边匹配的研究进行回顾，从平台的运营管理和设计和车货匹配方法与模型方面对车货匹配领域的研究进行整理。其次，梳理了多目标决策问题和整数规划的相关理论。然后，对A平台的基本情况和平台主要的车货匹配模式进行了介绍，分析了目前模式存在的不足。接着，根据A平台的具体情况和收集到的数据情况，结合前人对于车货匹配指标方面的研究，进行车货匹配的指标设计，重点构建了司机的信誉值、司机活跃度以及货主信誉值模型，司机方面主要考虑司机的自身行为和历史订单交易情况，货主方面主要考虑货主的历史订单交易情况和发货行为。对A平台的匹配采用固定时间间隔进行匹配的方式，并对司机给予随着等待匹配时间增加而提升的优先级参数，在此基础上建立车货双方的单独效用函数，进一步构建一对一多目标车货匹配模型，然后采用线性加权和的方式，将多目标转化为单目标进行求解。然后，选取货主和司机数据进行实例检验，并采用LINGO12.0求解。最后，从在车货匹配中加入司机信誉值、司机活跃度和货主信誉值的必要性和方法的角度，给出了对于A平台优化车货匹配后对接实现交易的成功率的建议，给其他平台及行业提供一点借鉴。

文章的末尾给出了本次研究中存在的不足和未来可以继续改进的研究方向。本研究对于提高匹配后成功交易的可能性，降低空驶率，降低物流成本，从而节约社会资源具有一定的参考意义。

关键词：车货匹配，多目标决策，效用函数

ABSTRACT

In this thesis, under the background of the high logistics cost and great waste of social resources caused by the return empty vehicles and low freight carrying rate caused by asymmetric information of highway freight in China, platform A, a typical platform of vehicle and cargo matching information platform, is selected as the research object. In view of the fact that the query list returned under the auto-searching mode based on platform A is only in reverse order of distance and release time, and the matching results are not accurate and effective enough, the research on the car-cargo matching model of platform A is carried out.

Firstly, the domestic and foreign researches on bilateral matching and vehicle-cargo matching are reviewed. The research on bilateral matching is reviewed from the aspects of algorithms, decision-making methods and models, and the application of bilateral matching in different fields and markets. This thesis summarizes the research in the field of vehicle and cargo matching from the aspects of platform operation management and design, vehicle and cargo matching methods and models. Secondly, the theory of multi-objective decision problem and integer programming is introduced. Thirdly, the basic situation of platform A and the main matching mode of vehicles and goods on the platform are introduced, and the deficiencies of the current mode are analyzed.

Then, according to the specific situation of platform A and the collected data, combining with previous studies on the matching index of vehicle and cargo, the index design of vehicle and cargo matching is carried out. The driver's credit value, the driver's activity and the shipper's credit value model are mainly constructed. The driver mainly considers the driver's own behavior and the historical order transaction situation, while the shipper mainly considers the shipper's historical order transaction situation and delivery behavior. The matching of platform A is carried out at a fixed time interval, and priority parameters are given to the driver as the waiting time for matching increases. On this basis, the individual utility function of the vehicle and cargo is established, and the one-to-one multi-objective vehicle and cargo matching model is further constructed. Then the multi-objective is transformed into a single objective by linear weighting method. Then, some owner and driver data are selected for instance test, and LINGO12.0 is adopted to solve the problem. Finally, the paper gives some Suggestions for platform A to improve the success rate of the matching of vehicles and goods, and provides some references for other platforms and industries.

At the end of the thesis, the deficiencies in this research and the research direction that can be further improved in the future are given. This study has certain reference significance for improving the possibility of successful transaction after matching, reducing the no-load rate, reducing the logistics cost, and thus saving social resources.

KEY WORDS: vehicle and cargo matching, multiple objective decision, the utility function

目 录

摘 要 I

ABSTRACT II

第一章 绪论 1

1.1研究背景和研究意义 1

1.1.1研究背景 1

1.1.2研究意义 2

1.2国内外研究现状 2

1.2.1双边匹配的算法研究 3

1.2.2双边匹配问题决策方法及模型 4

1.2.3双边匹配在不同领域和市场的研究 5

1.2.4车货匹配平台的运营管理和设计研究 5

1.2.5车货匹配方法与模型研究 6

1.3研究方法 7

1.4研究内容 7

第二章 相关理论概述 9

2.1多目标决策理论 9

2.2整数规划 10

第三章 A平台车货匹配现状分析 12

3.1 A平台的基本情况介绍 12

3.2 A平台的匹配模式 12

第四章 A平台的车货匹配模型构建 14

4.1 A平台的车货匹配问题描述 14

4.2 一对一车货匹配模型基础 15

4.2.1车货匹配模型假设 15

4.2.2车货匹配模型参数 15

4.2.3车货匹配指标设计 16

4.3 一对一车货匹配模型构建 21

4.3.1车货匹配数学模型构建 21

4.3.2模型转换 22

4.4 实例分析 22

4.4 相关优化建议 40

第五章 总结与展望 42

5.1本文主要工作的总结 42

5.2研究的不足与展望 42

参考文献 44

致 谢 48

第一章 绪论

1.1研究背景和研究意义

1.1.1研究背景

目前，根据中国国家统计局发布的《2018年国民经济和社会发展统计公报》可以看到，我国2018年一年的货物运输总量就达到了514.6亿吨，比上年增长7.1%，而货物运输周转量为205452亿吨公里，增长4.1%，其中由公路运输方式完成的货物运输量为395.9亿吨，比上年增长7.4%，2018年由公路完成的货物运输量要占到总体货物运输量的八成左右，这表明公路运输是我国货物运输最主要的方式也是最常用的方式^[1] 。根据中国国家统计局的年度数据显示，早在2017年，我国全国公路营运载货汽车保有量就达到了1368.62万辆，由此可见，运输行业尤其是公路运输行业正处于蓬勃发展的时期，互联网和电商的快速发展带来了许多强劲的公路运输需求。但是目前我国公路货运由于信息的不对称导致的返程空车和货运配载率不高等问题，带来了居高不下的物流成本和社会资源的极大浪费。

剩余内容已隐藏，请支付后下载全文，论文总字数：47632字