A Novel Swarm-UAV Formation Tracking Control Based on Networked-MPC and Dual Quaternion Algebra
Main Article Content
Abstract
Notions of natural systems inspired the robotics and communication communities to develop a new path of research named "The Robotics Swarm Networked Control" to cooperatively handle vast areas of tasks well beyond a single robot's abilities. The cornerstone research topic in such systems is the formation tracking control. This article investigates a newly developed formation tracking control for a swarm of crewless aerial vehicles (swarm-UAV). The proposed controller is based on a model predictive control algorithm (MPC). The primary motivation for considering such a solution is the ability to handle constrained nonlinear multi-input multi-output systems (MIMO), such as swarm-UAV. To enable swarm-UAV to accurately track time-varying shapes in 3D space, dual quaternion algebra was proposed to model the swarm-UAV system, which provides a compact formula to represent the UAV's position and orientation without singularities. The proposed approach adopts graph theory to simulate the effects of the communication links between the UAVs. Simulation experiments for three scenarios using the CasADi package integrated with the MATLAB environment were conducted and discussed to highlight the effectiveness of the proposed controller for real-time implementation. Finally, some suggestions for future improvements and research lines are listed.