Prediction of Transfer Probability of Traffic Flow for Shared Bikes based on Hybrid Optimization and Deep Learning Algorithms

Shared Bikes refers to the bike-sharing services across the residential, the commercial, and the public service areas. Predicting the destinations of the cycling trips and the traffic flow transfer of shared bikes between the traffic zones can improve the dispatching efficiency of shared bikes, recommendations of possible locations for the users, and improve the navigation efficiency. This paper proposes a stacked Restricted Boltzmann Machine (RBM)-Support Vector Regression (SVR) deep learning algorithm for predicting the transfer probability of traffic flow of Shared Bikes. To further improve the accuracy of prediction, the parameters in the stacked RBM-SVR algorithm were optimized by proposing a hybrid optimization algorithm named DEGWO, which obtained by merging the Differential Evolution (DE) algorithm and the Grey Wolf Optimization (GWO) algorithm. In an experimental case, the destinations of the cycling trips and the probability of traffic flow transfer for shared bikes between traffic zones were predicted by computing 2.46 million trajectory points recorded by shared bikes in Beijing. By making comparisons, it revealed that the stacked RBM-SVR algorithm, with the help of the DEGWO algorithm, outperformed the other two conventionals. As a result, the prediction results were much more accurate.

• Supplemental Notes:
  • This paper was sponsored by TRB committee ABJ35 Standing Committee on Highway Traffic Monitoring.
• Corporate Authors:

  Transportation Research Board

  ,    
• Authors:
  • Tu, Wenwen
  • Liu, Hengyi
• Conference:
  • Transportation Research Board 98th Annual Meeting
  • Location: Washington DC, United States
  • Date: 2019-1-13 to 2019-1-17
• Date: 2019

Language

• English

Media Info

• Media Type: Digital/other
• Features: Figures; References; Tables;
• Pagination: 7p

Subject/Index Terms

• TRT Terms: Algorithms; Bicycle travel; Machine learning; Optimization; Traffic flow; Vehicle sharing
• Uncontrolled Terms: Deep learning
• Subject Areas: Data and Information Technology; Highways; Operations and Traffic Management; Pedestrians and Bicyclists;

Filing Info

• Accession Number: 01697763
• Record Type: Publication
• Report/Paper Numbers: 19-04772
• Files: TRIS, TRB, ATRI
• Created Date: Mar 1 2019 3:51PM