Learning Car-Following Using Inertia-Oriented Driving
This study explores an alternative car-following strategy called 'Driving to Keep Inertia' (DI) and its effectiveness in a real-world closed circuit experiment.
Why it matters
This study offers insights into alternative car-following strategies beyond the standard 'Driving to Keep Distance' model, with potential implications for traffic flow analysis and autonomous vehicle control.
Key Points
- 1Previous studies questioned the default 'Driving to Keep Distance' strategy in car-following models
- 2Drivers can adopt alternative strategies like DI by following basic instructions
- 3This study shows DI training immediately translates to reduced acceleration, deceleration, and speed variability in real-world car-following
- 4It is the first to demonstrate the potential of DI strategy in a real circuit setting
Details
The paper investigates an alternative car-following strategy called 'Driving to Keep Inertia' (DI), which differs from the commonly assumed 'Driving to Keep Distance' (DD) strategy. Previous research has questioned whether DD is a universal traffic invariant, suggesting drivers can adopt alternative strategies like DI through simple instructions. This study extends that evidence by testing the immediate impact of DI training on a real closed circuit. Twelve drivers first participated in a car-following task, exhibiting the typical DD behavior both in the field and in simulated conditions. They then received DI training and repeated the same task, showing significantly reduced acceleration, deceleration, and speed variability - indicating they had adopted the DI strategy. This is the first study to demonstrate the potential of DI in a real-world driving circuit, suggesting it as an alternative to the default DD model in car-following and traffic flow analysis.
No comments yet
Be the first to comment