Integrated Decision Control Approach for Cooperative Safety-Critical Payload Transport in a Cluttered Environment

In this article, the problem of coordinated transportation of heavy payload by a team of unmanned aerial vehicles (UAVs) in a cluttered environment is addressed. The payload is modeled as a rigid body and is assumed to track a precomputed global flight trajectory from a start point to a goal point. Due to the presence of local dynamic obstacles in the environment, the UAVs must ensure that there is no collision between the payload and these obstacles while ensuring that the payload oscillations are kept minimum. An integrated decision controller (IDC) is proposed that integrates the optimal tracking control law given by a centralized model predictive controller with safety-critical constraints provided by the exponential control barrier functions. The entire payload-UAV system is enclosed by a safe convex hull boundary, and the IDC ensures that no obstacle enters this boundary. To evaluate the performance of the IDC, the results for a numerical simulation as well as a high-fidelity Gazebo simulation are presented. An ablation study is conducted to analyze the robustness of the proposed IDC against practical dubieties like noisy state values, relative obstacle safety margin, and payload mass uncertainty. The proposed method is then compared with a baseline controller qualitatively, emphasizing the use of barrier functions in safety-critical applications. The results clearly show that the IDC achieves both trajectory tracking and obstacle avoidance successfully while restricting the payload oscillations within a safe limit.

I. Introduction

Recent advancements in unmanned aerial vehicles (UAVs) and sensor technology has created a widespread interest in their ability to solve payload transportation problems in the logistics sector. Collaborative interaction of UAVs can efficiently solve many transportation problems, especially when there are requirements for delivering versatile payloads that vary in size. This is because utilizing multiple UAVs to lift payloads are economically feasible and efficient, as opposed to utilizing a single UAV of a bigger form factor. Industrial missions [1], military expeditions [2], and medical search-rescue operations [3] demand that the UAVs have to ensure proper delivery of the payload along a set preplanned global trajectory, while at the same time, preventing collisions with the local dynamic obstacles present around it and maintaining minimal oscillations of the payload throughout the course. However, when there are multiple UAVs interacting with one another to transport a payload, existing path planning algorithms [4] fall short, as they do not consider the kinodynamic-maneuverability of the system and it is highly coupled complex dynamics.