The Study On Personalized Ridesharing Services

Ridesharing means multiple people with similar journeys travel in the same vehicle.It has many advantages,such as convenient public travel,relieving travel pressure,promoting energy conservation and emission reduction,sharing travel costs and so on.The existing Ridesharing service only considers the similarity of passengers’ paths,and seldom considers the improvement of passenger experience through personalized factors such as social relations,interests and hobbies among users.There is still much development space in personalized and customized Ridesharing service.Firstly,this paper proposes a personalized vehicle sharing service that integrates the social relationship of users.In this paper,two kinds of social relationship and price model are proposed.Passengers with similar paths and social relationships are prioritized on one vehicle.Three vehicle sharing matching algorithms are proposed,which are based on quadtree-based candidate vehicle tailoring,SR-Qtree and ESR-Qtree index structure.And other technical means to optimize vehicle matching speed.The validity of the proposed algorithm and model is validated by simulation experiments.In the process of simulation experiments,we find that some users have sparse social relations.These users have no social relations with most other users,and it is difficult for these passengers to match other passengers with similar social relations to travel together.In order to improve the Personalized Ridesharing experience of sparse social relations users,we designed a personalized Ridesharing service based on user’s interests and hobbies,interests and hobbies of users and vehicles model is proposed,and this paper proposed a method to measure the similarity of users’ and users’ interests and hobbies between users and vehicles.When matching vehicles,priority should be given to matching passengers with similar paths and interests in a car.In addition,this paper presents a hybrid integer programming-based vehicle pooling matching algorithm and a threshold-based vehicle pooling matching algorithm,which can improve the efficiency of vehicle matching through space-time constraints,candidate vehicle tailoring and other methods.Finally,the effectiveness of the proposed model and algorithm is verified in the ridesharing matching simulation system.In this paper,the simulation system of ridesharing matching is written in Java language on OpenStreetMap,and the simulation experiments are carried out using the real order data sets of DiDi and New York Taxi.Compared with other advanced research work in the world,when the proposed algorithm and model are used to match,the social relationship between users increases by 8.5%,hobby similarity increases by 31.9%,and the matching time between passengers and vehicles decreases by 78.2%.The experimental results show that the research work in this paper can improve the personalized travel experience from the perspective of user social relations and interests,improve the matching speed between passengers and vehicles,and promote the upgrade of shared travel mode represented by vehicle sharing from "functional" to "personalized".