Research On Dynamic Taxi-Matching Model And Algorithm Based On Real-time Demand Response

As the number of cars in the country continues to increase,traffic congestion and the consequent environmental pollution and energy consumption problems have become more and more serious,restricting the rapid development of the city’s economy.Taxi,as an important public travel mode in cities,has caused the phenomenon of long empty distances,difficulty in carrying passengers during off peak hours for taxi drivers,and difficulty in taking taxis during peak hours for passengers due to its cruising and soliciting and the “one-to-one” operation mode of vehicles and passengers.Without increasing the supply of vehicles and changing the existing road conditions,the organization of online taxi-matching can make full use of vehicle vacancy resources and effectively improve the supply level of the taxi market.This paper takes the online taxi-matching problem as the research object,and focuses on how to realize the realtime response to the emerging passengers’ travel demand during the operation of the online taxi-matching so as to make the current ride-sharing plan the best.This article first discusses the ideas and principles of solving the single-vehicle ride-sharing matching problem,and proposes a method of clustering passengers based on the path fit,and orderly arranging the passenger demand nodes to determine the route of the vehicle.In order to take into account the interests of both ride-hailing drivers and passengers,based on the traditional taxi pricing model,a tariff calculation method is proposed that takes into account the passenger’s shared distance and detour distance,and uses the shared fare to replace the traditional pricing model.With a fixed percentage,it is verified through an example that the pricing method using the ridehailing rate enables the ride-hailing driver to obtain more revenue and the ride-sharing passengers to obtain price concessions.Based on the idea of single-vehicle sharing and matching,the problem of multi-vehicle sharing and matching with real-time demand response is studied,and a dual-objective optimization model with the largest passenger turnover and the shortest driving distance is established.This model is based on vehicle capacity,passenger time window and bypass a variety of conditions such as acceptable proportions,passenger costs,and driver’s income are used as constraints.Design an improved plug-in algorithm for solving the model,which can match the passenger travel demand generated in real time with the vehicle and find the optimal ride-sharing scheme.Finally,the MATLAB software is used to simulate the multi-taxi matching system.According to the matching scheme obtained by the simulation,the different operation effects of the sharing or not are compared and analyzed,and the single-factor analysis of variance method is used to determine the number of online taxi-matching and the possibility of detours for passengers.The influence of acceptance ratio and passenger arrival rate on the success rate of carpool matching.The simulation experiment results verify the effectiveness of the dynamic taximatching model and algorithm,and the research conclusions have reference value and practical significance for the operation organization and development of online taximatching strategies.