Car-Following Model Using Gaussian Process Regression for Machine Learning: Hybrid Model/Data Approach, Validated with Traffic Radar Detection

Traffic engineers rely on microscopic traffic models to design, plan and operate a wide range of traffic applications. Recently, large data sets are becoming available thanks to technology improvements and governmental efforts. With this study the authors aim to gain new empirical insights into longitudinal driving behaviour and to formulate a model which can benefit from these new challenging data sources. This paper proposes a new hybrid paradigm for car-following (CF) to describe individual longitudinal driving behaviour. The key of our approach is that the method integrates a parametric and a non-parametric mathematical formulation. The model predicts individual drivers acceleration given a set of variables. It uses Gaussian process regression (GPR) for machine learning formulation to make predictions when there exist correlation between new input and the training data set. The data-driven model benefits from a large training data set to capture all driver longitudinal behaviour, which would be difficult to fit in fixed parametric equation(s). The methodology allows to train models with new variables without the need of altering the model formulation. And importantly, the model also uses existing traditional parametric CF models to predict acceleration when no similar situations are found in the training data set. A case study using radar data in an urban environment shows that a hybrid model performs better than parametric model alone and suggests that traffic light status over time influence drivers acceleration. This methodology can help engineers to use large data sets and to find new variables to describe traffic behaviour.

Language

  • English

Media Info

  • Media Type: Digital/other
  • Features: Figures; References; Tables;
  • Pagination: 18p

Subject/Index Terms

Filing Info

  • Accession Number: 01764010
  • Record Type: Publication
  • Report/Paper Numbers: TRBAM-21-00809
  • Files: TRIS, TRB, ATRI
  • Created Date: Dec 23 2020 11:17AM