出租车合乘模式下的智能匹配问题的研究与实现

发布时间：2019-07-01 12:29

【摘要】：随着我国经济的快速发展，出租车作为唯一能为城市居民提供个性化出行的公共交通方式，已经成为人们日常生活中不可或缺的一部分。然而，出租车在为广大寻常百姓带来了许多便利和享受的同时，其24小时寻客和“一车一人”的服务方式，造成了出租车空间资源浪费、城市交通拥堵、大气环境污染、石油能源消耗等一系列问题。针对上述情况，“车辆合乘”的概念应运而生，被许多专家学者誉为改善出租车运营问题的最佳途径。本文以出租车智能合乘匹配问题作为研究对象，对相关理论方法展开了系统的分析，从乘客的实际需求出发，设计完成了出租车智能匹配公共服务系统。该系统运用一种分段求解的智能匹配算法，将整个问题分为两部分，先通过乘客分配，将多辆出租车合乘匹配问题简化为单辆出租车合乘匹配问题，再经过车辆路线优化，最终得到合乘匹配的最优解。本文主要进行了以下几个方面的工作：（1）对乘客进行分配，确定每辆出租车搭乘的乘客。根据出租车初始路径中各站点周围的情况，运用粒子群优化算法，以匹配率作为目标优化函数，加入约束条件和乘客的个性化需求，反复迭代得出每辆出租车行驶过程中途径站点的最佳调整半径。将半径范围内的乘客划分到某辆特定的出租车上，为进行下一步的优化过程做基础。（2）对车辆路线进一步优化，得到花费最少的行驶路线。将已经划分到同一辆出租车上的乘客，运用遗传算法进行优化求解。以总花费作为目标优化函数，对个体进行遗传选择，通过反复循环地进行选择、交叉、变异操作，对每辆车上的乘客内部上下车顺序进行排序，剔除适应度低的个体，留下适应度高的个体，最后得到优化的行驶路线，满足总花费最少的目标。（3）在算法研究的基础上设计完成了出租车智能匹配公共服务系统。该系统是为满足乘客的个性化合乘需求而设计，集成了乘客合乘平台、出租车车载平台、智能匹配平台和智能监管平台四大服务平台。通过分段匹配算法，为出租车司机和乘客提供便捷的车辆合乘匹配服务，以最少的花费和代价，满足尽可能多的乘客出行需求。同时，利用出租车作为信息采集终端，配合部分乘客用车的数据，为监督管理部门提供智能交通分析服务、安全生产分析服务等。本文通过产生随机数据进行实验模拟，验证了算法的有效性，对解决出租车合乘问题具有较强的实际应用价值，可以为今后出租车行业的发展提供一定的理论支持，对优化出租车资源合理配置，改善城市交通状况，促进城市公共交通更加和谐稳定地发展提供了很大程度上的帮助。
[Abstract]:With the rapid development of economy in our country, taxi, as the only public transportation mode which can provide individualized travel for urban residents, has become an indispensable part of people's daily life. However, taxi has brought a lot of convenience and enjoyment to ordinary people, at the same time, its 24-hour passenger search and "one car, one person" service mode have caused a series of problems, such as waste of taxi space resources, urban traffic congestion, air environment pollution, oil energy consumption and so on. In view of the above situation, the concept of "vehicle ride" emerges as the times require, which is praised by many experts and scholars as the best way to improve taxi operation. In this paper, the intelligent taxi combination matching problem is taken as the research object, and the related theories and methods are systematically analyzed. According to the actual needs of passengers, the taxi intelligent matching public service system is designed and completed. The system uses a piecewise intelligent matching algorithm to divide the whole problem into two parts. first, the problem of multiple taxi combination matching is simplified to a single taxi combination matching problem through passenger assignment, and then the optimal solution of combination matching is obtained by optimizing the vehicle route. The main work of this paper is as follows: (1) distribute the passengers and determine the passengers taken by each taxi. According to the situation around each station in the initial path of taxi, the particle swarm optimization algorithm is used to take the matching rate as the objective optimization function, and the constraints and the personalized needs of passengers are added to obtain the optimal adjustment radius of each taxi during the driving process. The passengers in the radius range are divided into a specific taxi to serve as the basis for the next optimization process. (2) further optimize the vehicle route and get the least expensive driving route. The passengers who have been divided into the same taxi are optimized by genetic algorithm. Taking the total cost as the objective optimization function, the genetic selection of the individual is carried out. Through repeated cyclic selection, crossing and variation operation, the internal boarding sequence of each passenger is sorted, the individuals with low fitness are eliminated, the individuals with high fitness are left behind, and the optimized driving route is finally obtained to meet the goal of the least total cost. (3) on the basis of algorithm research, the intelligent matching public service system of taxi is designed and completed. The system is designed to meet the personalized needs of passengers. It integrates four service platforms: passenger ride platform, taxi vehicle platform, intelligent matching platform and intelligent supervision platform. Through the segmented matching algorithm, a convenient vehicle matching service is provided for taxi drivers and passengers to meet the travel needs of as many passengers as possible with the least cost and cost. At the same time, the taxi is used as the information collection terminal to cooperate with the data of some passengers to provide intelligent transportation analysis service, safety production analysis service and so on for the supervision and management department. In this paper, the effectiveness of the algorithm is verified by generating random data, which has strong practical application value for solving the taxi ride problem, can provide certain theoretical support for the development of taxi industry in the future, and provides great help for optimizing the rational allocation of taxi resources, improving urban traffic conditions and promoting the more harmonious and stable development of urban public transport.
【学位授予单位】：中国海洋大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U492.434;U495;TP18

【参考文献】