Quadrotor Morpho-Transition: Learning vs Model-Based Control Strategies

2025-06-16

AI Summary

This work addresses the control challenges—strong nonlinearity, aerodynamic coupling, and actuator saturation—arising in aerial dynamic morpho-transition landing of quadrotors. We systematically compare end-to-end reinforcement learning (PPO) and nonlinear model predictive control (NMPC) across hardware transferability, robustness, and disturbance recovery. Our key findings are: (i) successful real-world deployment of RL policies critically depends on accurate modeling of motor dynamics and sensor observation latency; (ii) NMPC achieves zero-tuning, plug-and-play operation without requiring prior knowledge of detailed dynamics. Experiments show that RL enables higher-agility landings but suffers from limited generalization; NMPC, while less tolerant to actuator failures, significantly improves deployment reliability and engineering practicality. This establishes a new “verifiable–deployable” paradigm for morphing-flight control.

Technology Category

• Intelligent Robots: Behavior Learning & Control
• Machine Learning: Reinforcement Learning
• Multiagent Systems: Multiagent Learning

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Accountability, Transparency, and Ethics for personalization
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

Quadrotor Morpho-Transition, or the act of transitioning from air to ground through mid-air transformation, involves complex aerodynamic interactions and a need to operate near actuator saturation, complicating controller design. In recent work, morpho-transition has been studied from a model-based control perspective, but these approaches remain limited due to unmodeled dynamics and the requirement for planning through contacts. Here, we train an end-to-end Reinforcement Learning (RL) controller to learn a morpho-transition policy and demonstrate successful transfer to hardware. We find that the RL control policy achieves agile landing, but only transfers to hardware if motor dynamics and observation delays are taken into account. On the other hand, a baseline MPC controller transfers out-of-the-box without knowledge of the actuator dynamics and delays, at the cost of reduced recovery from disturbances in the event of unknown actuator failures. Our work opens the way for more robust control of agile in-flight quadrotor maneuvers that require mid-air transformation.

Problem

• Study quadrotor air-to-ground transition with complex aerodynamics
• Compare RL and MPC controllers for morpho-transition performance
• Address hardware transfer challenges like motor dynamics and delays

Innovation

• Uses Reinforcement Learning for morpho-transition control
• Incorporates motor dynamics and observation delays
• Compares RL with baseline MPC controller performance