Parameter Adjustments in POMDP-Based Trajectory Planning for Unsignalized Intersections
Adam Kollarčík adn Zdeněk Hanzálek
TL;DR
The paper addresses safe crossing of unsignalized intersections when the vehicle lacks the right of way, under perception and behavior uncertainty. It models trajectory planning as a Partially Observable Markov Decision Process (POMDP) and solves it approximately using the Adaptive Belief Tree (ABT) algorithm, with a discretized intersection topology, a dynamics model for predicting other vehicles' states, and an observation model linking states to noisy measurements; the method outputs a trajectory as a sequence of accelerations along the path. The approach is validated via simulations using real-world aerial traffic data from two intersections with different topologies, demonstrating collision-free planning. It also analyzes how ABT/POMDP parameter settings influence performance, providing practical guidance for parameter tuning and future deployments.
Abstract
This paper investigates the problem of trajectory planning for autonomous vehicles at unsignalized intersections, specifically focusing on scenarios where the vehicle lacks the right of way and yet must cross safely. To address this issue, we have employed a method based on the Partially Observable Markov Decision Processes (POMDPs) framework designed for planning under uncertainty. The method utilizes the Adaptive Belief Tree (ABT) algorithm as an approximate solver for the POMDPs. We outline the POMDP formulation, beginning with discretizing the intersection's topology. Additionally, we present a dynamics model for the prediction of the evolving states of vehicles, such as their position and velocity. Using an observation model, we also describe the connection of those states with the imperfect (noisy) available measurements. Our results confirmed that the method is able to plan collision-free trajectories in a series of simulations utilizing real-world traffic data from aerial footage of two distinct intersections. Furthermore, we studied the impact of parameter adjustments of the ABT algorithm on the method's performance. This provides guidance in determining reasonable parameter settings, which is valuable for future method applications.