Adaptive look-ahead driving for enhancing traffic flow performance at urban signalized intersections

Enhancing driving efficiency while ensuring high throughput at signalized intersections is challenging to achieve safe and sustainable urban mobility. This paper proposes a novel adaptive look-ahead driving (ALD) system that employs anticipatory driving decisions to improve traffic flow performance, intersection throughput, and vehicle fuel efficiency at signalized intersections for a given signal timing to address this issue. The proposed system combines a naturalistic human driving model with an adaptive cruise control (ACC) model by extending them with a look-ahead step to realize desired stopping and starting behavior at traffic signals through safe and precise following of a preceding vehicle. Considering the predictive or look-ahead states of the preceding vehicle and utilizing an adaptive desired speed in response to the red signal ahead allows the host vehicle to behave like a predictive, optimally controlled vehicle with minimal computational effort. The performance of the proposed system is analyzed through microscopic traffic simulations at signalized intersection traffic scenarios, focusing on maintaining safe time headway and optimizing intersection throughput. The comparative analysis demonstrates improvements in intersection clearance, human-like vehicle spacing, and overall traffic flow efficiency. The proposed system can be implemented as an advanced driver assistance system (ADAS) for safe and sustainable mobility at urban intersections.