Abstract:For intelligent quadcopter UAVs, a robust and reliable autonomous planning system is crucial. Most current trajectory planning methods for UAVs are suitable for static environments but struggle to handle dynamic obstacles, which can pose challenges and even dangers to flight. To address this issue, this paper proposes a vision-based planning system that combines tracking and trajectory prediction of dynamic obstacles to achieve efficient and reliable autonomous flight. We use a lightweight object detection algorithm to identify dynamic obstacles and then use Kalman Filtering to track and estimate their motion states. During the planning phase, we not only consider static obstacles but also account for the potential movements of dynamic obstacles. For trajectory generation, we use a B-spline-based trajectory search algorithm, which is further optimized with various constraints to enhance safety and alignment with the UAV's motion characteristics. We conduct experiments in both simulation and real-world environments, and the results indicate that our approach can successfully detect and avoid obstacles in dynamic environments in real-time, offering greater reliability compared to existing approaches. Furthermore, with the advancements in Natural Language Processing (NLP) technology demonstrating exceptional zero-shot generalization capabilities, more user-friendly human-machine interactions have become feasible, and this study also explores the integration of autonomous planning systems with Large Language Models (LLMs).

What problem does this paper attempt to address?

### What problem does this paper attempt to solve? This paper aims to address the challenges faced by small quadrotor Unmanned Aerial Vehicles (UAVs) in autonomous navigation within dynamic environments. Specifically: 1. **Dynamic Obstacle Perception**: - Most current trajectory planning methods are primarily suited for static environments and face difficulties when dealing with dynamic obstacles. Since the state of dynamic obstacles is usually unknown, treating them as static obstacles may not provide enough time to replan the trajectory, increasing the risk of collisions. 2. **Efficient and Reliable Autonomous Flight System**: - A vision-based planning system is proposed, combining dynamic obstacle tracking and trajectory prediction to achieve efficient and reliable autonomous flight. This system uses lightweight object detection algorithms (such as NanoDet) to identify dynamic obstacles and employs Kalman Filtering to track and estimate their motion state. During the planning phase, it considers not only static obstacles but also the potential movement of dynamic obstacles. 3. **B-Spline Trajectory Generation Algorithm**: - A trajectory search algorithm based on B-spline curves is used, with multiple constraint optimizations to enhance safety and smoothness. Experimental results show that this method can detect and avoid obstacles in dynamic environments in real-time, making it more reliable compared to existing methods. 4. **Integration of Natural Language Processing Technology**: - The paper explores the integration of Large Language Models (LLMs) into the autonomous planning system to improve human-machine interaction friendliness. By leveraging advanced natural language processing technology, UAV operation becomes more user-friendly, providing valuable insights for future research. In summary, this paper aims to develop an autonomous planning system capable of effectively avoiding obstacles and ensuring safe flight in complex dynamic environments. It also explores the integration of the latest natural language processing technologies to enhance the overall usability and intelligence level of the system.

A Safer Vision-based Autonomous Planning System for Quadrotor UAVs with Dynamic Obstacle Trajectory Prediction and Its Application with LLMs

Autonomous Flights in Dynamic Environments with Onboard Vision

Onboard Distributed Trajectory Planning Through Intelligent Search for Multi-UAV Cooperative Flight

Integrated Path Planning and Trajectory Tracking Control for Quadrotor UAVs with Obstacle Avoidance in the Presence of Environmental and Systematic Uncertainties: Theory and Experiment

Real-time Motion Planning and Trajectory Tracking in Complex Environments Based on Bézier Curves and Nonlinear MPC Controller

A Vision-based UAV Tracker Aiming at Aerial Targets

Dynamic Trajectory Planning for Unmanned Aerial Vehicle Based on Sparse A* Search and Improved Artificial Potential Field

6-DOF UAV Path Planning and Tracking Control for Obstacle Avoidance: A Deep Learning-Based Integrated Approach

Path Planning for Fixed-Wing UAVs Based on Expert Knowledge and Improved VFH in Cluttered Environments

A Motion Camouflage-Inspired Path Planning Method for UAVs Based on Reinforcement Learning

Intent Prediction-Driven Model Predictive Control for UAV Planning and Navigation in Dynamic Environments

Perception-Aware Baesd Quadrotor Translation and Yaw Angle Planner in Unpredictable Dynamic Situations

Safety-Driven and Localization Uncertainty-Driven Perception-Aware Trajectory Planning for Quadrotor Unmanned Aerial Vehicles

A Reinforcement Learning Based Motion Planner for Quadrotor Autonomous Flight in Dense Environment

A Hybrid Human-in-the-Loop Deep Reinforcement Learning Method for UAV Motion Planning for Long Trajectories with Unpredictable Obstacles

Cooperative Motion Planning for Persistent 3D Visual Coverage with Multiple Quadrotor UAVs

DPMPC-Planner: A real-time UAV trajectory planning framework for complex static environments with dynamic obstacles

A real-time dynamic obstacle tracking and mapping system for UAV navigation and collision avoidance with an RGB-D camera

Quadrotor Trajectory Planning for Visibility-Aware Target Following

EVA-Planner: Environmental Adaptive Quadrotor Planning

A Novel Real-Time Moving Target Tracking and Path Planning System for a Quadrotor UAV in Unknown Unstructured Outdoor Scenes