Nonlinear MPC for Full-Pose Manipulation of a Cable-Suspended Load using Multiple UAVs

TL;DR

This paper introduces a nonlinear model predictive control approach for multiple UAVs to manipulate a load's full pose, considering system nonlinearities and constraints for safe, fast operation.

Contribution

It presents the first control method that accounts for the full nonlinear load-UAV system model with comprehensive feasibility constraints.

Findings

Simulation results show fast computation times (<100 ms) for up to 10 UAVs.

The method enables safe and efficient manipulation of a payload's full pose.

Demonstrates potential for industrial applications like construction and heavy-load transport.

Abstract

In this work, we propose a centralized control method based on nonlinear model predictive control to let multiple UAVs manipulate the full pose of an object via cables. At the best of the authors knowledge this is the first method that takes into account the full nonlinear model of the load-UAV system, and ensures all the feasibility constraints concerning the UAV maximumum and minimum thrusts, the collision avoidance between the UAVs, cables and load, and the tautness and maximum tension of the cables. By taking into account the above factors, the proposed control algorithm can fully exploit the performance of UAVs and facilitate the speed of operation. Simulations are conducted to validate the algorithm to achieve fast and safe manipulation of the pose of a rigid-body payload using multiple UAVs. We demonstrate that the computational time of the proposed method is sufficiently small (<100 ms) for UAV teams composed by up to 10 units, which makes it suitable for a huge variety of future industrial applications, such as autonomous building construction and heavy-load transportation.