Passenger Route Selection Algorithm Based On Bus IC Card Data Research

With the popularity of network technology in intelligent transportation system applications, such basic services related to transportation facilities become more perfect, and public transport as an important part of urban transport, has its unique advantages to meet the residents travel in ease urban traffic pressure and reduce urban smog, etc. to make a great contribution, therefore, in recent years, cities are seeking to develop the public transport system. With the increasing of urban public traffic network with perfect, citizens travel conditions greatly improved, for the passenger to choose the route also increase gradually So how passengers travel route information and scientific decision-making, to provide reliable information to the public reasonable dispatch, and the passenger can choose the best travel route, a new problem faced by bus companies.Factors affecting passenger route choice has many aspects, such as the number of transfers, travel time, distance and cost, etc. In this study, the Ministry of Transport key demonstration project "Xinxiang City Public Transport Intelligent application demonstration projects" based on the use Xinxiang data bus IC cards, a variety of factors to take the bus passengers are taken into account in the analysis, and the choice of estimation model and ant colony algorithm are two ways to study, so that more accurate predictions of traffic, reaching avoid traffic congestion and optimization purposes passenger routing structure. The main contents are as follows:(1) In-depth analysis of some of the core concepts of the paper needed and theory:data mining and artificial intelligence theory and composition of public transportation IC card, IC card data structures and other aspects, the ant colony algorithm is presented later and selection estimation model theoretical foundation.(2) The impact on the passenger route choice analyzes the factors, mainly including the number of transfers, travel time, distance, travel costs, traffic, etc., to determine the optimal target two algorithms model.(3) Establish the estimated passenger route choice model, using a large data bus passenger travel time algorithm to obtain the probability density function is calculated for each path between OD passenger travel time composed of elements, build a single model line bus passengers more choice OD matrix using the ordered weighted averaging operator estimated OD matrix is the result of many sorts, for each OD matrix estimation of various outcomes given different weights, draw a line to select the whole bus passengers estimated weighted value, and in accordance with the value established using a combination of large data bus passengers line selection matrix estimation model and the use of relevant data Xinxiang IC card simulation model to estimate.(4) Consideration of passengers affected by a combination of factors trip distance route selection, based on traditional paper shortest route algorithm, and analyzes some of its shortcomings when it reaches the optimal route solving algorithm to GBAS passenger travel route choice feasibility analysis and specific algorithm, and then verify the validity of the algorithm by Xinxiang City, IC card data.