Transfer station reckoning method based on multivariate data

Abstract

The invention discloses a transfer station reckoning method based on multivariate data, and particularly relates to the technical field of urban intelligent bus data mining, comprising the steps of S1, acquiring main transfer station pairs and selecting station pairs which may have multi-station transfer, S2, filling in multi-station transfer lines and station lists according to transfer habits and transfer psychological factors of passengers; S3, according to the obtained algorithm transfer distance, carrying out transfer proportion calculation on the number of passengers getting off the bus at adjacent stations; and S4, according to the actually measured number of transferred passengers and the types of the transfer station pairs, training the above algorithm based on various bus peak and low peak time periods, outputting a corresponding training model. According to the method, the actual passenger flow transfer information of the high and low mean peaks can be accurately determined, corresponding transfer facilities and equipment can be conveniently matched by urban construction, traffic management and other departments, and more accurate transport capacity and transport volume matching can be conveniently carried out by public transport operation and management departments, so that important public transport urban indexes such as public transport transfer efficiency and passenger satisfaction can be improved.

Description

technical field [0001] The present invention relates to the technical field of urban intelligent public transport data mining, and more specifically, the present invention relates to a transfer station estimation method based on multivariate data. Background technique [0002] For traffic passenger flow analysis, it is very important to obtain the alighting and disembarking stations of public transportation passengers (passenger flow OD analysis). At present, domestic buses generally only swipe the IC card when getting on the bus, and do not swipe the card when getting off, resulting in the IC card passenger flow data only having the boarding station information but not the getting off station information. For this reason, the industry calculates the alighting station by combining the traffic travel rules of passengers within a period of time combined with the information of the bus GPS. For the specific technical background, please refer to the author's paper - "Bus passeng...

AI Technical Summary

Problems solved by technology

[0003] In the calculation of the specific alighting station, especially when the bus transfers to the track or other lines of the bus, the following problems are usually encountered: the track/bus station that the passenger enters can be known through the swiping data of the track/bus IC card that the passenger transfers , but when the bus line where the passenger got off had two or more stations close to the track/bus line to be transferred, how to judge the exact bus stop of the passenger? If the calculation of the transfer station is wrong, on the one hand, it will lead to unreasonable guarantee and design of transfer facilities and passages at different stations; and so on

But according to the actual test, it is found that when there are two or more adjacent bus stops beside the track site, due to the actual transfer distance, the arrival sequence of the bus stops (in the case of road congestion, passengers will choose the distance but the The bus s

Images