Bike-sharing systems became one of the most popular means of individual transport in many European cities. With the growth of popularity of bicycle-sharing systems, the estimation of demand for its services becomes a very significant task for transportation planners. The data from GPS tracking systems installed on bicycles can be used as a reliable source of information in this case. The paper presents an approach to identify the segment of recreational trips, implemented within the bike-sharing system, based on poplar clusterisation algorithms. The authors developed the subroutines for cleaning the raw data obtained from GPS-trackers. By using the purified data on the numeric parameters of trips in a bicycle sharing system, the clustering model identifies such a cluster that represent recreational trips. The k-means, mini-batch k-means, birch algorithms, and agglomerative clustering are examined in this study in order to allocate recreational trips. The use of the proposed approach is demonstrated on the example of data obtained from Wavelo bike-sharing system in the city of Krakow, Poland.
Using Clustering Algorithms to Identify Recreational Trips Within a Bike-Sharing System
Lect. Notes in Networks, Syst.
International Conference on Reliability and Statistics in Transportation and Communication ; 2019 ; Riga, Latvia October 16, 2019 - October 19, 2019
Reliability and Statistics in Transportation and Communication ; Chapter : 14 ; 130-138
2020-03-29
9 pages
Article/Chapter (Book)
Electronic Resource
English
Travel demand , Bike-sharing system , Recreational trips , Data cleansing , Clustering algorithms Engineering , Computational Intelligence , Transportation Technology and Traffic Engineering , Engineering Economics, Organization, Logistics, Marketing , Data Mining and Knowledge Discovery , Applied Statistics
Estimating bike-share trips using station level data
Online Contents | 2015
|Transportation Research Record | 2014
|The Impact of Dockless Bike-Sharing and Built Environment on Ride-Sourcing Trips
DOAJ | 2024
|