Measuring fuel consumption in vehicle routing: new estimation models using supervised learning

In this paper we propose and access the accuracy of new fuel consumption estimation models for vehicle routing. Based on real world data of instantaneous fuel consumption, time-varying speeds observations, and high-frequency traffic data related to a large set of shipping operations we propose effective methods to estimate fuel consumption and greenhouse gas emissions. By carrying out nonlinear regression analysis using supervised learning methods we develop new models that provide better prediction accuracy than classical ones. We correctly estimate consumption for time-dependent point-to-point routing under realistic conditions taking into account freight transportation operations during peak hour traffic congestion, stop-and-go driving patterns, idle vehicle states and the variation of vehicle loads. Extensive computational experiments on real datasets show the effectiveness of the proposed machine learning emissions models. A detailed analysis of the relative importance of input variables confirms the efficiency of our models.

  • Record URL:
  • Supplemental Notes:
    • Revised version of CIRRELT-2018-15
  • Corporate Authors:

    Interuniversity Research Centre on Enterprise Networks, Logistics and Transportation (CIRRELT)

    Montreal, Quebec   
  • Authors:
    • Heni, H
    • Diop, S A
    • Coehlo, L C
    • Renaud, J
  • Publication Date: 2019-3


  • English

Media Info

  • Pagination: 27p

Subject/Index Terms

Filing Info

  • Accession Number: 01714726
  • Record Type: Publication
  • Source Agency: ARRB Group Limited
  • Report/Paper Numbers: CIRRELT-2019-08
  • Files: ATRI
  • Created Date: Aug 26 2019 11:22AM