Reachable Sets-based Trajectory Planning Combining Reinforcement Learning and iLQR

2025-03-19

AI Summary

To address insufficient trajectory safety in complex driving scenarios, this paper proposes a risk-aware hierarchical trajectory planning framework. First, a driving risk field is constructed and integrated into reachable set modeling, yielding a feasible region bounded by risk constraints. Second, an initial trajectory is generated via safety-enhanced reinforcement learning and projected onto the risk-constrained reachable set. Finally, a constrained iterative Linear Quadratic Regulator (iLQR) performs multi-objective co-optimization for safety, comfort, and efficiency. This work is the first to explicitly incorporate the risk field into reachable set construction, enabling “planning outside risk boundaries” and joint multi-objective optimization. In high-speed lane-change simulations, the generated trajectories consistently avoid high-risk regions while satisfying acceleration/curvature constraints and time-to-completion requirements, demonstrating significant safety improvement.

Technology Category

• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Intelligent Robots: Motion and Path Planning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

The driving risk field is applicable to more complex driving scenarios, providing new approaches for safety decision-making and active vehicle control in intricate environments. However, existing research often overlooks the driving risk field and fails to consider the impact of risk distribution within drivable areas on trajectory planning, which poses challenges for enhancing safety. This paper proposes a trajectory planning method for intelligent vehicles based on the risk reachable set to further improve the safety of trajectory planning. First, we construct the reachable set incorporating the driving risk field to more accurately assess and avoid potential risks in drivable areas. Then, the initial trajectory is generated based on safe reinforcement learning and projected onto the reachable set. Finally, we introduce a trajectory planning method based on a constrained iterative quadratic regulator to optimize the initial solution, ensuring that the planned trajectory achieves optimal comfort, safety, and efficiency. We conduct simulation tests of trajectory planning in high-speed lane-changing scenarios. The results indicate that the proposed method can guarantee trajectory comfort and driving efficiency, with the generated trajectory situated outside high-risk boundaries, thereby ensuring vehicle safety during operation.

Problem

• Enhances safety in complex driving scenarios using risk reachable sets.
• Integrates driving risk field for accurate risk assessment and avoidance.
• Optimizes trajectory for comfort, safety, and efficiency via reinforcement learning.

Innovation

• Reachable set integrates driving risk field
• Safe reinforcement learning initializes trajectory
• Constrained iLQR optimizes safety and comfort