PI-WAN: A Physics-Informed Wind-Adaptive Network for Quadrotor Dynamics Prediction in Unknown Environments

TL;DR

PI-WAN is a novel physics-informed neural network that enhances quadrotor dynamics prediction in unknown environments by integrating physical constraints with data-driven learning, leading to improved accuracy and robustness.

Contribution

This work introduces PI-WAN, a hybrid modeling approach that embeds physical principles into a neural network for better generalization in unpredictable conditions.

Findings

Outperforms baseline models in prediction accuracy.

Enhances tracking precision in real-world tests.

Demonstrates robustness to wind disturbances and payload variations.

Abstract

Accurate dynamics modeling is essential for quadrotors to achieve precise trajectory tracking in various applications. Traditional physical knowledge-driven modeling methods face substantial limitations in unknown environments characterized by variable payloads, wind disturbances, and external perturbations. On the other hand, data-driven modeling methods suffer from poor generalization when handling out-of-distribution (OoD) data, restricting their effectiveness in unknown scenarios. To address these challenges, we introduce the Physics-Informed Wind-Adaptive Network (PI-WAN), which combines knowledge-driven and data-driven modeling methods by embedding physical constraints directly into the training process for robust quadrotor dynamics learning. Specifically, PI-WAN employs a Temporal Convolutional Network (TCN) architecture that efficiently captures temporal dependencies from historical flight data, while a physics-informed loss function applies physical principles to improve model generalization and robustness across previously unseen conditions. By incorporating real-time prediction results into a model predictive control (MPC) framework, we achieve improvements in closed-loop tracking performance. Comprehensive simulations and real-world flight experiments demonstrate that our approach outperforms baseline methods in terms of prediction accuracy, tracking precision, and robustness to unknown environments.