Autonomous Quadrotor Navigation Using LiDAR SLAM and Adaptive Tree-Based Path Planning in Simulation

Autonomous quadrotor navigation is vital for applications like search and rescue, environmental monitoring, and infrastructure inspection, where human intervention is limited or unsafe, requiring quadrotors to navigate complex environments independently. In this paper, we propose a novel approach that integrates accurate localization, efficient path planning, and real-time obstacle avoidance to enhance autonomous navigation. Our system, developed and tested through the ROS2 framework and Gazebo simulator, utilizes a binary occupancy grid generated manually via LiDAR or sourced externally. The quadrotor then localizes itself using Adaptive Monte Carlo Localization (AMCL) and plans 2D paths with a Rapidly-exploring Random Tree (RRT) algorithm enhanced by a Dynamic RRT (DRRT) algorithm for real-time obstacle adaptation. Path outputs are translated into precise rotor speed commands, enabling stable flight and effective obstacle avoidance, with altitude and stability maintained through PID controllers. Experimental results in simulation demonstrate our system's effectiveness in localization, path planning, and obstacle avoidance, highlighting its potential for real-world deployment in dynamic environments.

I. Introduction

The autonomous navigation of quadrotors has become a focal point of research due to its significant applications in diverse fields such as search and rescue, environmental monitoring, and surveillance. The ability of quadrotors to maneuver in three-dimensional space, access difficult terrains, and perform precise tasks makes them invaluable in scenarios where human intervention is impractical or risky.