Output-Feedback Optimal Control Of Heterogeneous Quadrotor Swarms With Communication Delays

Abstract

This paper presents a unified framework for optimal output-feedback control of heterogeneous quadrotor uncrewed aerial vehicles (QUAVs) operating in coordinated formations under communication delays, without requiring exact system dynamics. To address partial state observability, a multi-layer neural network (MNN) observer is developed to precisely estimate unmeasured states. Reinforcement learning (RL) is employed to achieve optimal control using an MNN critic that ensures adaptability. Together, the MNN observer and RL relax the need for explicit system dynamics, enabling the framework to be applied to heterogeneous QUAVs with differing masses and parameters. A Lyapunov-Krasovskii functional is incorporated into the learning process to compensate for communication delays, which arise either in transmitting commands from the ground station to the leader or in relaying leader information to the followers. Such delays, as a non-stationary source, can otherwise cause instability and degrade formation performance. A three-dimensional leader-follower formation strategy expressed in spherical coordinates enables efficient maneuvering. Theoretical analysis establishes closed-loop stability, and simulation studies validate the effectiveness of the proposed framework in ensuring optimal, stable, and adaptive formation control of heterogeneous QUAVs under communication delays.

Department(s)

Electrical and Computer Engineering

Second Department

Computer Science

Comments

Army Research Office, Grant W911NF-22-2-0185

Keywords and Phrases

formation control; Lyapunov-Krasovskii functional; non-stationary environments; Reinforcement learning

International Standard Book Number (ISBN)

979-8-3315-9381-0; 979-8-3315-9382-7

International Standard Serial Number (ISSN)

0743-1619

Document Type

Article - Conference proceedings

Document Version

Citation

File Type

text

Language(s)

English

Rights

© 2026 Institute of Electrical and Electronics Engineers, All rights reserved.

Publication Date

01 Jan 2026

Link to Full Text

Share

 
COinS