Research And Design Of Personalized Bus Route Recommendation System

With the economic and social development,the number of vehicles per capita in my country has increased sharply,and the traffic congestion in the city has become more and more intense.The development of public transportation has become a good way to solve the problem of urban congestion.People often choose public transportation travel plans based on the traditional electronic map bus route recommendation system,but because it is only based on a single factor to recommend routes that cannot meet the individual needs of passengers,it limits people’s willingness to take public transportation.To this end,this paper proposes a personalized bus route recommendation system,which builds a passenger route recommendation model based on the three factors of bus congestion,travel time and travel distance,and makes personalized recommendations based on the different weights given to each factor by passengers.The front-end of the system adopts the distributed micro-service system built with the Vue framework and the back-end adopts the Spring Boot+Spring Cloud framework.Through the analysis of functional requirements,the system functions are determined including: site query,route query,personalized route recommendation,popular route recommendation and background information management,etc.,Each function can be deployed,run,and tested independently for each microservice,and is a complete function of the system through calls between services.Finally,combined with the actual data set and the API interface provided by Baidu Maps,the running effect is displayed on the map.The main research contents of this paper are as follows:(1)Research on the factors that affect the choice of bus passengers’ travel routes,including: the number of transfers,route distance,travel time,cost,and passenger flow,etc.,so as to determine the factors that need to be optimized for the model.(2)Introduce the data structure of conventional bus data and preprocess the IC card data and GPS data in it,and then cluster and group the IC card data with the set time threshold as a constraint condition and match the vehicle arrival time in the GPS data,Infer the station where the swiping passenger boarded the bus.The alighting probability model based on the travel behavior of bus passengers is improved,and the alighting probability model of this paper is constructed by combining the attraction right of the station and the transfer ability of the station.From this,we can infer the degree of congestion in the car and establish the functional relationship between the degree of congestion in the car and the value of time,and then convert the degree of congestion in the car into extra time to pay,and combine travel time and distance to construct a passenger route recommendation model.Finally,the actual data set is used to verify the judgment of getting on and off stations to illustrate the effectiveness of the model.(3)Analyze the shortcomings of traditional methods for solving multi-objective problems,propose the traditional linear weighting algorithm combined with coarse-grained parallel genetic algorithm as the solution algorithm and give detailed algorithm steps,and finally compare the experimentally recommended routes with those recommended by Baidu Maps Explain the feasibility.(4)Determine the development functions required by the system through functional and non-functional requirements analysis,and combine the idea of distributed microservices to construct a system architecture diagram while designing the database.Finally,according to the actual data and combined with Baidu map API,the operation effect of the system function module is carried out.verification.