Reduction of travel times and traffic emissions using model predictive control

In this paper we present a model-based traffic flow control approach to improve both traffic flow and emissions in a traffic network. A model predictive control (MPC) is implemented using a microscopic car-following traffic flow model and an average-speed-based emission model. We consider reduction of total time spent (TTS) and total emissions (TE) as performance measures of the control strategy. Moreover, with the help of simulations we illustrate that a traffic control strategy, particularly an MPC strategy, aiming at the reduction of the TTS does not necessarily reduce the level of emissions. In particular, when the traffic flow is congested, we demonstrate that a traffic control strategy that addresses TTS (or improvement of the traffic flow) alone can cause an increment in the level of emissions and vice versa. Therefore, in this paper we explain how to integrate both requirements so that a balanced trade-off is obtained.

I. INTRODUCTION

Despite the improvements in transportation systems, the rise of fuel prices, and the imposition of more stringent environmental policies for emission levels, the demand for mobility and transportation is continuously increasing. Consequently roads are frequently congested, creating economical, social, and ecological challenges. Moreover, in recent epidemiological studies of the effects of combustion-related (mainly traffic-generated) air pollution, NO₂ was shown to be associated with adverse health effects [23], [24]. Furthermore, road traffic exhaust emissions account for 40% of volatile organic compounds, more than 70% of , and over 90% of CO in most European cities [23], and about 45% of the pollutants released in the US [21]. Frequent and longer congested traffic conditions make this even worse.