Journal
IEEE TRANSACTIONS ON INTELLIGENT VEHICLES
Volume 3, Issue 3, Pages 300-309Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TIV.2018.2843125
Keywords
Drivermodeling; trajectory prediction; optimization; intelligent vehicles
Categories
Funding
- Office of Naval Research MURI Award [ONR-N000141310341]
- DURIP Award [N000141310679]
- Berkeley DeepDrive
Ask authors/readers for more resources
In order to develop safe autonomous systems, accurate and precise models of human behavior must be developed. For intelligent vehicles, one can imagine the need for predicting driver behavior to develop minimally invasive active safety systems or to safely interact with human drivers on the road. We present an optimization-based method for approximating the stochastic reachable set for human-in-the-loop systems. This method identifies the most precise subset of states that a human driven vehicle may enter, given some data set of observed trajectories. We phrase this problem as a mixed integer linear program, solved via branch and bound methods. The resulting model uncovers the most representative set that encapsulates the likely trajectories, up to some probability threshold, by optimally rejecting outliers in the data set. This tool provides set predictions consisting of trajectories observed from the nonlinear dynamics and behaviors of the human driven car, and can account for modes of behavior, like the driver state or intent. This method is applied to predict lane changing behavior, achieving similar to 90% accuracy over long time horizons. This flexible algorithm is to handle realistic complex human driving data and provide a robust and informative prediction of driver behavior.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available