Online path planning for kinematic-constrained UAVs in a dynamic environment based on a Differential Evolution algorithm

2026-07-21 · Source: cs.NE updates on arXiv.org · Field: Technology & Digital — Robotics & Autonomous Systems, Artificial Intelligence & Machine Learning · Depth: Expert, long

Summary

This research introduces an online path planner for fixed-wing Unmanned Aerial Vehicles (UAVs) designed to navigate dynamic environments while adhering to kinematic constraints like maximum curvature and desired orientations. The proposed system utilizes a Non-Uniform Rational B-spline (NURBS) curve for path representation and employs the Linear-Success-History Based Adaptive Differential Evolution for Constrained Optimization Problems (LSHADE-COP) algorithm. It integrates concepts from the Velocity Obstacle approach into its constraint function to avoid dynamic obstacles within a limited sensing range. This online planner achieves a re-planning interval of approximately 0.1 seconds, demonstrating a 12,000-fold speed improvement over a previous offline version. Simulations confirm its feasibility and ability to provide safe navigation in crowded, dynamic settings, laying groundwork for future 3D extensions.

Key takeaway

For Robotics Engineers developing UAV navigation systems for dynamic environments, this online planner offers a robust solution. You should consider integrating NURBS-based path representation with a Differential Evolution algorithm and Velocity Obstacle constraints to achieve rapid, safe re-planning. This approach allows your fixed-wing UAVs to handle kinematic limitations and dynamic obstacles effectively, significantly improving operational safety and efficiency in complex missions.

Key insights

Online path planning for kinematic-constrained UAVs in dynamic environments is achieved using NURBS and Differential Evolution with Velocity Obstacle integration.

Principles

• NURBS curves offer flexible path representation.
• Velocity Obstacle approach enables dynamic collision avoidance.
• Differential Evolution can optimize constrained path planning.

Method

The method optimizes variations of NURBS curve parameters using LSHADE-COP, minimizing path length while satisfying kinematic constraints, no-fly zones, and dynamic obstacle avoidance via Velocity Obstacle integration at each re-planning interval.

In practice

• Implement NURBS for smooth UAV trajectories.
• Integrate Velocity Obstacle for dynamic collision detection.
• Use LSHADE-COP for constrained path optimization.

Topics

• UAV Path Planning
• Differential Evolution
• NURBS Curves
• Velocity Obstacle
• Kinematic Constraints
• Dynamic Environments

Best for: Research Scientist, AI Scientist, Robotics Engineer, Machine Learning Engineer

Related on AIssential

• From Distances to Trajectories: Real-Time Signed Distance Function Mapping and Distance-Accelerated Motion Planning for UAVs — Co-designing mapping and planning with Signed Distance Functions (SDFs) enhances autonomous UAV navigation.
• Coverage Path Planning: Classical Foundations, Recent Advances, and Future Directions — Coverage Path Planning (CPP) evolves from classical single-robot methods to complex multi-robot, 3D, and learning-based systems.
• Motion Planning of Cooperative Nonholonomic Mobile Manipulators — A two-step motion planning approach enables nonholonomic MMRs to cooperatively transport objects amidst complex obstacles.
• Online Navigation Planning for Long-term Autonomous Operation of Underwater Gliders — GLAE uses MCTS and a physics-informed simulator for robust, long-term autonomous underwater glider navigation.
• New research enables a robot to chart a better course — MIGHTY is an open-source, real-time UAV trajectory planner that optimizes path and time simultaneously for faster, smoother, obstacle-avoiding flight.
• Time-to-Collision Based Dynamic Obstacle Avoidance Using Pretrained Vision Models for Robots in Unstructured Environments — Pretrained vision models enable data-efficient, interpretable dynamic obstacle avoidance for robots in unstructured environments.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by cs.NE updates on arXiv.org.