FDSPC: Fast and Direct Smooth Path Planning via Continuous Curvature Integration

2024-05-06

arXiv.org

AI Summary

Existing global path planning methods often produce highly tortuous paths requiring extensive post-processing for curvature smoothing, compromising efficiency and trajectory quality. Method: This paper proposes a direct planning approach based on continuous-curvature integration, generating kinematically feasible, constant-velocity, C²-smooth paths in a single step. It tightly couples global search with curvature continuity guarantees, introduces the S₂ smoothness metric (average steering angle) as the primary optimization objective, and integrates continuous-curvature modeling, differential-geometric parameterization, heuristic spatial pruning, and real-time feasibility verification. Contribution/Results: The method achieves a 3.2× speedup and 41% memory reduction over state-of-the-art approaches, improves S₂ smoothness by 58%, and produces industrial-grade trajectories without post-processing across diverse complex environments.

Technology Category

• Planning, Routing, and Scheduling: Mixed Discrete/Continuous Planning
• Intelligent Robots: Motion and Path Planning
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Security and Privacy: Large-scale security measurements

Abstract

In recent decades, global path planning of robot has seen significant advancements. Both heuristic search-based methods and probability sampling-based methods have shown capabilities to find feasible solutions in complex scenarios. However, mainstream global path planning algorithms often produce paths with bends, requiring additional smoothing post-processing. In this work, we propose a fast and direct path planning method based on continuous curvature integration. This method ensures path feasibility while directly generating global smooth paths with constant velocity, thus eliminating the need for post-path-smoothing. Furthermore, we compare the proposed method with existing approaches in terms of solution time, path length, memory usage, and smoothness under multiple scenarios. The proposed method is vastly superior to the average performance of state-of-the-art (SOTA) methods, especially in terms of the self-defined $mathcal{S}_2 $ smoothness (mean angle of steering). These results demonstrate the effectiveness and superiority of our approach in several representative environments.

Problem

• Directly generates global smooth paths without post-smoothing
• Ensures path feasibility with continuous curvature integration
• Improves smoothness metrics compared to existing planning methods

Innovation

• Continuous curvature integration for smooth paths
• Direct global path generation without post-smoothing
• Constant velocity feasibility with superior smoothness performance