基于最小二乘支持向量机的车货匹配实现研究  

Research on LS-SVM Based Vehicle Cargo Matching System

作　　者：黄美华[1] 李君[1] 吴耀辉[1] 王林荣 马炜 HUANG Mei-hua;LI Jun;WuYao-hui;WANG Lin-rong;MA Wei(Zhejiang Wanli University,Ningbo Zhejiang 315000)

机构地区：[1]浙江万里学院,浙江宁波315000

出　　处：《浙江万里学院学报》2018年第2期87-92,共6页Journal of Zhejiang Wanli University

基　　金：浙江省公益项目(2017C31040);宁波自然科学基金(2015A610130;2016A610225);宁波市科技惠民(2017C50028);浙江省大学生科技创新计划及新苗计划(2017R420012);浙江省自然科学基金(LY17F010002)

摘　　要：文章针对物流信息平台的特点,运用最小二乘支持向量机来匹配车货需求信息,满足客户的需求,能够有效地降低货运空载率,降低企业成本,提高用户满意度,推动物流现代化的发展,由此具有广泛的现实意义。通过分析车货信息系统的匹配原理与匹配模式,制定相应的评价指标,并运用最小二乘支持向量机建立车货信息匹配模型。为了进一步减少计算时间,文章提出基于遗传算法的最小二乘支持向量机的车货匹配算法,能够在保持高达97%匹配精度的同时,把计算时间缩短到一半左右,从而更好地满足物流企业对车货匹配的需求,为促进物流现代化起到了一定的作用。In this paper,according to the characteristics of logistics information platform,the least square support vector machine(LS-SVM)was used to study vehicle cargo match to meet the needs of customers.It could effectively reduce the freight rate,the cost of enterprises,improve the customer satisfaction and promote the logistics modernization.Therefore,it has a wide range of practical significance.By analyzing the matching principle and pattern of vehicle cargo information system,the corresponding evaluation index is developed.Meanwhile,the vehicle cargo information matching model is established by using LS-SVM.In order to further reduce the computation time,the vehicle cargo matching algorithm was proposed by a genetic algorithm based on LS-SVM,which could cut to half of the computation time while the matching accuracy rate maintains 97%.Thus,it can better meet the demand of vehicle cargo match,and will play an important role in promoting the modernization of logistics.

关 键 词：车货匹配 最小二乘支持向量机 匹配原理 遗传算法 

分 类 号：F540.32[经济管理—产业经济]