Towards Data-Driven Vehicle Estimation for Signalised Intersections in a Partially Connected Environment.

Roozbeh Mohammadi,Claudio Roncoli

doi:10.3390/s21248477

Abstract

Connected vehicles (CVs) have the potential to collect and share information that, if appropriately processed, can be employed for advanced traffic control strategies, rendering infrastructure-based sensing obsolete. However, before we reach a fully connected environment, where all vehicles are CVs, we have to deal with the challenge of incomplete data. In this paper, we develop data-driven methods for the estimation of vehicles approaching a signalised intersection, based on the availability of partial information stemming from an unknown penetration rate of CVs. In particular, we build machine learning models with the aim of capturing the nonlinear relations between the inputs (CV data) and the output (number of non-connected vehicles), which are characterised by highly complex interactions and may be affected by a large number of factors. We show that, in order to train these models, we may use data that can be easily collected with modern technologies. Moreover, we demonstrate that, if the available real data is not deemed sufficient, training can be performed using synthetic data, produced via microscopic simulations calibrated with real data, without a significant loss of performance. Numerical experiments, where the estimation methods are tested using real vehicle data simulating the presence of various penetration rates of CVs, show very good performance of the estimators, making them promising candidates for applications in the near future.

Highlights

Urban traffic congestion is a widespread phenomenon appearing in most cities worldwide, bringing a range of negative impacts on the quality of citizens’ lives and environment
root-mean-square error (RMSE) and mean absolute error (MAE) calculated for the aggregated estimation method decrease as the number of Connected vehicles (CVs) increases
When the number of CVs is higher than six, RMSE and MAE are less than 1 veh, indicating a very good estimation performance; we see that the two metrics converge, implying that the absolute values of the errors are similar for all data points

Summary

Introduction

Urban traffic congestion is a widespread phenomenon appearing in most cities worldwide, bringing a range of negative impacts on the quality of citizens’ lives and environment. Various adaptive traffic control strategies have been proposed during the last decades to facilitate traffic movement in urban signalised intersections, including, e.g., [2,3,4,5,6,7,8]. The availability of accurate and reliable real-time information is a prerequisite for running efficient adaptive signal control strategies. For this purpose, various infrastructurebased sensors, such as, e.g., loop detectors, radars, cameras, and magnetometers, have been employed to collect the necessary measurements [9]. Various infrastructurebased sensors, such as, e.g., loop detectors, radars, cameras, and magnetometers, have been employed to collect the necessary measurements [9] These data collection tools have several deficiencies. In order to deal with incomplete measurements retrieved from infrastructure-based sensors, various studies proposed the usage of estimation techniques, aimed at supplementing missing data (see, e.g., [13,14,15])

Methods

Results

Conclusion