[Integrating Trajectory Optimization and Reinforcement Learning for Quadrupeda Jumping with Terrain-Adaptive Landing]

Renjie Wang, Shangke Lyu, Xin Lang, Wei Xiao, Donglin Wang,

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2025.

• This work proposes a safe landing framework that achieves adaptive landing on rough terrains by combining Trajectory Optimization (TO) and Reinforcement Learning (RL) together.

[CARP: Visuomotor Policy Learning via Coarse-to-Fine Autoregressive Prediction]

Zhefei Gong, Pengxiang Ding, Shangke Lyu, Siteng Huang, Mingyang Sun, Wei Zhao, Zhaoxin Fan, Donglin Wang,

IEEE/CVF International Conference on Computer Vision (ICCV), 2025.

• In this paper, we introduce Coarseto-Fine AutoRegressive Policy (CARP), a novel paradigm for visuomotor policy learning that redefines the autoregressive action generation process as a coarse-to-fine.

[Gevrm: Goal-expressive Video Generation Model for Robust Visual Manipulation]

Hongyin Zhang, Pengxiang Ding, Shangke Lyu, Ying Peng, Donglin Wang,

International Conference on Learning Representations (ICLR), 2025.

• We propose a novel closed-loop VLA method GEVRM that integrates the classic internal model control (IMC) principle to enhance the robustness of robot visual manipulation.

[The robotic guide dog for individuals with visual impairments]

Shangke Lyu, Zhengyu Wei, Donglin Wang,

Nature Review Electrical Engineering, vol.2, no.1, pp.9-10, 2025.

• The development of robotic guide dogs is crucial to improve the independence and mobility of people with limited vision, offering a safer and more efficient alternative to traditional aids. Here we highlight our Robotic Guide Dog system, focusing on its technological framework and discussing challenges encountered during commercialization.

[QUART-Online: Latency-Free Multimodal Large Language Model for Quadruped Robot Learning]

Xinyang Tong, Pengxiang Ding, Yiguo Fan, Donglin Wang, Wenjie Zhang, Can Cui, Mingyang Sun, Han Zhao, Hongyin Zhang, Yonghao Dang, Siteng Huang, Shangke Lyu,

The IEEE International Conference on Robotics and Automation (ICRA), 2025.

• We introduce a novel latency-free quadruped MLLM model, dubbed QUARTOnline, designed to enhance inference efficiency without degrading the performance of the language foundation model.

[Koopman-based Robust Learning Control with Extended State Observer]

Shangke Lyu†, Xin Lang†, Donglin Wang,

IEEE Robotics and Automation Letters, vol. 10, no. 3, pp. 2303-2310, 2025.

• In this paper, we propose a robust active learning (RAL) control method designed to optimize data efficiency during model learning while ensuring robust and precise control during task execution.

[Toward Air Operation Aerial Manipulator Control with a Refined AntiDisturbance Architecture]

Shangke Lyu, Yu Zhang, Jianliang Wang, Chien Chern Cheah, Xiang Yu,

IEEE Transactions on Automation Science and Engineering,vol. 22, pp. 4076-4091, 2025.

• In this paper, a refined anti-disturbance control architecture is proposed for the decentralized aerial manipulator model, where various disturbances with different mathematical properties are well explored and tackled according to their positions and effects acting on the system.

[Precise end-effector control for an aerial manipulator under composite disturbances: Theory and experiments]

Meng Wang†, Shangke Lyu†, Qianyuan Liu, Ziqi Yang, Kexin Guo, Xiang Yu,

IEEE Transactions on Automation Science and Engineering, vol. 22, pp. 4006-4021, 2025.

• In this paper,a joint velocity planner is proposed to handle the base-floating disturbance in kinematics loop, while NN assisted DOB is adopted to handel the uncertainties in dynamics level.

[RL2AC: Reinforcement Learning-based Rapid Online Adaptive Control for Legged Robot Robust Locomotion]

Shangke Lyu, Xin Lang, Han Zhao, Hongyin Zhang, Pengxiang Ding, Donglin Wang,

Robotics: Science and Systems(RSS), 2024.

• In this paper, we seek to ascertain the control mechanism behind the locomotion RL policy, from which we propose a new RL based Rapid onLine Adaptive Control (RL2AC) algorithm to complementarily combine the RL policy and the adaptive control together.

[A New Observer for Perspective Vision Systems with Partially Uncertain Linear Motion Parameters]

Shangke Lyu, Xiaoyu Ma, Jianliang Wang, Zhitao Wang, Jianzhong Qiao, Yukai Zhu

IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 54, no. 10, pp. 5936-5948, 2024.

• In this article, we consider the depth estimation problem for the perspective vision system in the case that the camera angular velocities are accurate, while partial camera linear velocities are contaminated by some disturbances.

[A Coordinated Framework of Aerial Manipulator for Safe and Compliant Physical Interaction]

Qianyuan Liu, Shangke Lyu, Kexin Guo, Jianliang Wang, Xiang Yu, Lei Guo,

Control Engineering Practice, vol.146, 2024.

• This paper proposes a coordinated interactive framework for aerial manipulators to achieve safe and compliant interaction when physically contacting the surroundings.

[Contact Force Estimation of Robot Manipulators With Imperfect Dynamic Model: On Gaussian Process Adaptive Disturbance Kalman Filter]

Yanran Wei, Shangke Lyu, Wenshuo Li, Xiang Yu, Zidong Wang, Lei Guo,

IEEE Transactions on Automation Science and Engineering, vol. 21, no. 3, pp. 3524-3537, 2024.

• In this article, we consider the depth estimation problem for the perspective vision system in the case that the camera angular velocities are accurate, while partial camera linear velocities are contaminated by some disturbances.

[GeRM: A Generalist Robotic Model with Mixture-of-experts for Quadruped Robot]

Wenxuan Song, Han Zhao, Pengxiang Ding, Can Cui, Shangke Lyu, Yaning Fan, Donglin Wang,

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2024.

• In this article, we develop the Generalist Robotic Model (GeRM) to utilizes a Mixture-of-Experts architecture to enhance multi-task learning in quadruped robots, optimizing data utilization through offline reinforcement learning.

[A Composite Control Strategy for Quadruped Robot by Integrating Reinforcement Learning and Model-Based Control]

Shangke Lyu, Han Zhao, Donglin Wang,

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2023.

• In this paper, the proposed model-based controller is able to rectify the gait commands generated by DRL based on the system dynamic model so as to enhance the robustness of the quadruped robot and alleviate the sim-to-real problem.

[Quadrotor UAV: Collision Resilience Behaviors]

Dadong Fan, Kexin Guo, Shangke Lyu, Xiang Yu, Lihua Xie, Lei Guo,

IEEE Transactions on Aerospace and Electronic Systems, vol. 59, no. 3, pp. 2092-2104, 2023.

• In this article, a safety control scheme for quadrotor is proposed to guarantee collision resilience like flying insects.

[Interaction Task Motion Learning for Human-Robot Interaction Control]

Shangke Lyu, Muthuchamy Selvaraj Nithish, Chien Chern Cheah,

IEEE Transactions on Human-Machine Systems, vol. 52, no. 5, pp.894-906, 2022.

• In this article, a task learning approach is proposed for human–robot interaction systems, where the motion behaviors demonstrated by humans can thus be acquired by the robot by seeking the appropriate task parameters of the dynamic potential energy function.

[Compensation Filtering for Spacecraft Attitude Estimation Using Error-Covariance Reconstruction]

Zhenbin Qiu, Jianchun Zhang, Shangke Lyu,

IEEE Transactions on Instrumentation and Measurement, vol. 71, pp.1-10, 2022.

• This article proposes a new scheme to reconstruct the error covariance matrix to minimize this truncation error.

[Design of an Aerial Manipulator System Applied to Capture Missions]

Wenyu Zhang, Qianyuan Liu, Meng Wang, Jindou Jia, Shangke Lyu, Kexin Guo, Xiang Yu, Lei Guo,

International Conference on Unmanned Aircraft Systems (ICUAS), 2021.

• This paper proposes the design of an aerial manipulator system for capture missions.

[Human-Robot Interaction Control Based on a General Energy Shaping Method]

Shangke Lyu, Chien Chern Cheah,

IEEE Transactions on Control Systems Technology, vol. 28, no. 6, pp. 2445-2460, 2020.

• In this article, a general HRI control framework is proposed for the scenario of the human and robot coexisting in the same workspace.

[Data-driven Learning for Robot Control With Unknown Jacobian]

Shangke Lyu, Chien Chern Cheah,

Automatica, vol. 120, pp.109-120, 2020.

• In this paper, a NN based data driven offline learning algorithm and an online learning controller are proposed, which are combined in a complementary way.