UAV Path Planning for Reconnaissance and Look-Ahead Coverage Support for Mobile Ground Vehicles

Area of Science:

• Robotics and Autonomous Systems
• Artificial Intelligence
• Aerospace Engineering

Background:

• Path planning for Unmanned Aerial Vehicles (UAVs) supporting Mobile Ground Vehicles (MGVs) is complex due to variable MGV dynamics and environmental differences.
• Existing methods often focus solely on tracking, neglecting coverage objectives and facing limitations with dynamic environments and varying velocities.
• Gimbal sensors are typically used, increasing hardware dependency and computational load.

Purpose of the Study:

• To present a novel 3D path planning technique, Enhanced Dynamic Artificial Potential Field (ED-APF), for multirotor UAVs.
• To enable UAVs to provide autonomous reconnaissance and look-ahead coverage support for MGVs using non-gimbal sensors.
• To enhance MGV reconnaissance capabilities by exploring areas beyond their sensor range.

Main Methods:

• Formulated path planning as a combined follow and cover problem using non-gimbal sensors.
• Developed a vertical sinusoidal path for the UAV that dynamically adapts to the MGV's position, velocity, and heading.
• Optimized sinusoidal path amplitude and frequency to maximize look-ahead visual coverage quality (pixel density and quantity).

Main Results:

• The ED-APF technique demonstrated superior performance compared to general artificial potential field methods in simulations.
• Validated through various scenarios using Robot Operating System (ROS) and Gazebo-supported PX4-SITL.
• Showcased suitability for dynamic and obstacle-populated environments, effectively supporting MGVs.

Conclusions:

• The ED-APF provides an effective solution for UAV path planning in reconnaissance and support missions for MGVs.
• The method enhances look-ahead capabilities and extends MGV operational range in complex environments.
• ED-APF offers a computationally efficient and hardware-flexible alternative to existing techniques.