Research And Implementation Of The Intelligent Matching Problem On Shared-Ride Taxi

With the fast development of economy, taxi as the only public transportation toprovide personalized travel for urban residents, it has become an indispensable part ofpeople’s daily life. However, taxis bring a lot of convenience and enjoyment for themajority of ordinary people, but at the same time, taxis run24hours a day and theway of ‘one car one person’ service, also caused a series of problems like the taxispace resources waste, the urban traffic congestion, atmospheric environmentalpollution, oil energy consumption and so on. In view of the above situation, theconcept of “vehicle carpool” is produced. Many experts believe that this is the bestway to solve the defects of taxi service.In this paper, research focuses on vehicle ride matching problem,analyzes the relative theory and method deeply. From the actual demand ofpassengers, the paper completed the design of public service system fortaxi intelligent matching. The system uses an intelligent subsection matchingalgorithm, divides the whole problem into two parts. In the part of the allocation ofpassengers, first transform multiple taxi ride matching problem into a single taxi ridematching problem. Then through the part of route optimization, finally get the optimalsolution carpools matching.Three topics are proposed as following：(1) In the process of passenger’s distribution, make sure every passenger onwhich taxi. According to the situation around each site in the taxi initial path, usingparticle swarm optimization algorithm, taking matching rate as the target optimizationfunction, adding constraints and personalized needs of passengers, iterative calculatethe Optimal adjustment radius of each site in the process of taxi running. Divide thepassengers in the radius into a particular taxi, realizing the classification of thepassengers. This will make a foundation for the next step of the optimization process.(2) In the process of vehicle route optimization, get the ultimate route withminimum cost. Use genetic algorithm to optimize the order of passengers who dividedinto the same taxi. Take the total cost as the target optimization function and makegenetic selection for every individual. Through repeated cycle the operation of selection, crossover and mutation, order the sequence of every passengers on eachtaxi. Remove lower fitness individuals, leaving higher fitness individuals. Finally getthe optimal route to meet the target of total cost minimum.(3) Based on the research of algorithm, this paper completed the design of publicservice system for taxi intelligent matching. The system is designed to meetpassengers’ individual demand and integrates the passengers ride platform, taxivehicle platform, intelligent matching platform and intelligent monitoring platform.Through the segment matching algorithm, provide convenient vehicle ride matchingservice for taxi drivers and passengers with minimal cost, meeting the demand ofpassengers as much as possible. At the same time, use the taxi as a data collectionterminal. Cooperating with some passenger car data, provide intelligent trafficanalysis services and safety production analysis service for the supervision andmanagement department.In the paper, simulation experiments verify the effectiveness of the algorithmwhich has an important referenced value in the actual application. It can provide sometheoretical support for the future development of the taxi industry. This studyprovides a largely help to optimize the allocation of taxi resources, improving urbantraffic conditions, promoting the development of urban public transportation moreharmonious and stable.