Integrating Learning-Based Manipulation and Physics-Based Locomotion for Whole-Body Badminton Robot Control

2025-04-24

AI Summary

To address the challenge of simultaneously achieving training efficiency, stability, and safety in agile badminton rallies—where learning-based control and physics-based modeling are difficult to reconcile—this paper proposes a hybrid imitation learning (IL) and reinforcement learning (RL) framework grounded in a model-based foundation with physics augmentation. Specifically, a model predictive controller (MPC) governs base locomotion, while a physics-informed learned policy controls the arm. During IL, a critic network is pre-trained to mitigate policy degradation during transfer; privileged information is further incorporated to accelerate convergence and enhance robustness. Experiments on a custom-built robot demonstrate 94.5% rally success rate against a ball-serving machine and 90.7% against human players—substantially outperforming purely learning-based baselines. Moreover, the framework exhibits strong cross-task generalization, successfully transferring to agile grasping and table-tennis manipulation tasks.

Technology Category

• Machine Learning: Imitation Learning & Inverse Reinforcement Learning
• Intelligent Robots: Behavior Learning & Control
• Multiagent Systems: Adversarial Agents

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Learning-based methods, such as imitation learning (IL) and reinforcement learning (RL), can produce excel control policies over challenging agile robot tasks, such as sports robot. However, no existing work has harmonized learning-based policy with model-based methods to reduce training complexity and ensure the safety and stability for agile badminton robot control. In this paper, we introduce ourmethod, a novel hybrid control system for agile badminton robots. Specifically, we propose a model-based strategy for chassis locomotion which provides a base for arm policy. We introduce a physics-informed ``IL+RL'' training framework for learning-based arm policy. In this train framework, a model-based strategy with privileged information is used to guide arm policy training during both IL and RL phases. In addition, we train the critic model during IL phase to alleviate the performance drop issue when transitioning from IL to RL. We present results on our self-engineered badminton robot, achieving 94.5% success rate against the serving machine and 90.7% success rate against human players. Our system can be easily generalized to other agile mobile manipulation tasks such as agile catching and table tennis. Our project website: https://dreamstarring.github.io/HAMLET/.

Problem

• Harmonize learning-based and model-based methods for robot control
• Reduce training complexity while ensuring safety and stability
• Develop hybrid control for agile badminton robot tasks

Innovation

• Hybrid control system combining learning and model-based methods
• Physics-informed IL+RL training framework for arm policy
• Model-based chassis locomotion supporting arm policy