R2Deep: Recharging Recommendation System for Electric Taxis based on Deep Learning

Electric taxi (eTaxi) often face an inherent long waiting and recharging time (e.g., in hours) at charging stations. Therefore, where to charge and how long to charge an eTaxi has already emerged as an urgent and crucial problem to be solved for wide deployment of eTaxis. In this project, the research team proposes to develop a recharging recommendation system based on deep learning, called R2Deep, for eTaxi drivers to improve their operational efficiency as well as increase the revenue of both eTaxi drivers and companies. The project has three tasks: 1) analyze the existing eTaxi global positioning system (GPS) trajectory data and convert them into information on the grid maps, which will then be directly fed into deep learning models; 2) utilize deep learning techniques including both Recurrent Neural Network (RNN) and Convolutional Neural Network (CNN) to learn latent patterns behind eTaxi data sets and provide real-time suggestions on recharging time and charging stations to eTaxis drivers; 3) evaluate the R2Deep model, analyze its performance (e.g., the recommendation accuracy, increase in the eTaxi drivers and companies’ revenue, average reduction of waiting time at charging stations, etc.) with real world data.

Language

  • English

Project

  • Status: Active
  • Funding: $74638
  • Contract Numbers:

    DTRT13-G-UTC48

  • Sponsor Organizations:

    Office of the Assistant Secretary for Research and Technology

    University Transportation Centers Program
    Department of Transportation
    Washington, DC  United States  20590
  • Managing Organizations:

    Office of the Assistant Secretary for Research and Technology

    University Transportation Centers Program
    Department of Transportation
    Washington, DC  United States  20590
  • Project Managers:

    Su, Lu

    Dong, Wen

    Qiao, Chunming

  • Performing Organizations:

    State University of New York, Buffalo

    212 Ketter Hall
    Buffalo, NY  United States  14260
  • Principal Investigators:

    Qiao, Chunming

    Su, Lu

    Dong, Wen

  • Start Date: 20170901
  • Expected Completion Date: 20180831
  • Actual Completion Date: 0
  • USDOT Program: University Transportation Centers

Subject/Index Terms

Filing Info

  • Accession Number: 01652095
  • Record Type: Research project
  • Source Agency: Transportation Informatics University Transportation Center
  • Contract Numbers: DTRT13-G-UTC48
  • Files: UTC, RiP
  • Created Date: Nov 28 2017 2:44PM