Reinforcement Twinning for Hybrid Control of Flapping-Wing Drones

TL;DR

This paper introduces a hybrid reinforcement learning approach combining model-based and model-free methods for controlling flapping-wing drones, leveraging adaptive digital twins and transfer learning to improve flight performance.

Contribution

It proposes a novel reinforcement twinning algorithm that integrates adaptive digital twins with reinforcement learning for hybrid drone control, enhancing robustness and efficiency.

Findings

Hybrid approach outperforms pure model-free methods.

Adaptive digital twin improves control accuracy.

Method effective under various initialization scenarios.

Abstract

Controlling the flight of flapping-wing drones requires versatile controllers that handle their time-varying, nonlinear, and underactuated dynamics from incomplete and noisy sensor data. Model-based methods struggle with accurate modeling, while model-free approaches falter in efficiently navigating very high-dimensional and nonlinear control objective landscapes. This article presents a novel hybrid model-free/model-based approach to flight control based on the recently proposed reinforcement twinning algorithm. The model-based (MB) approach relies on an adjoint formulation using an adaptive digital twin, continuously identified from live trajectories, while the model-free (MF) approach relies on reinforcement learning. The two agents collaborate through transfer learning, imitation learning, and experience sharing using the real environment, the digital twin and a referee. The latter selects the best agent to interact with the real environment based on performance within the digital twin and a real-to-virtual environment consistency ratio. The algorithm is evaluated for controlling the longitudinal dynamics of a flapping-wing drone, with the environment simulated as a nonlinear, time-varying dynamical system under the influence of quasi-steady aerodynamic forces. The hybrid control learning approach is tested with three types of initialization of the adaptive model: (1) offline identification using previously available data, (2) random initialization with full online identification, and (3) offline pre-training with an estimation bias, followed by online adaptation. In all three scenarios, the proposed hybrid learning approach demonstrates superior performance compared to purely model-free and model-based methods.