Rerouting UAVs in a Dynamic Environment Based on an Optimized Interfered Fluid Dynamical System
TL;DR
This work improved the IFDS algorithm with both global and local path adaptation, incorporating environmental data such as weather and terrain to enhance path accuracy, and used numerical optimization methods to address dynamic path-planning issues.
Abstract
Navigating unmanned aerial vehicles (UAVs) through areas full of obstacles is a complex task. Dynamic obstacles in real-world scenarios include not only physical moving objects but also changing environmental conditions. To ensure mission success, UAVs must be able to reroute in real time to avoid collisions and high-risk zones. This need for real-time computation makes the dynamic path-planning problem even more challenging. We developed a new dynamic path-planning program based on the Interfered Fluid Dynamical System (IFDS) framework to address these issues. We improved the IFDS algorithm with both global and local path adaptation, incorporating environmental data such as weather and terrain to enhance path accuracy, and used numerical optimization methods. Every aspect of the program is expressed in a closed form, enabling efficient real-time operation even in three-dimensional environments. Our approach offers a lightweight and transparent system that allows any part of the program to be adjusted as needed, making it well-suited for real-world UAV navigation, where the system must be tailored to specific requirements or constraints.