Research

Publications

[1] Zhang, C., Chen, J., Li, J., Peng, Y., & Mao, Z. (2023). Large language models for human-robot interaction: A review. Biomimetic Intelligence and Robotics, 100131. (Paper link)

[2] Zhang, C., Meng, X., Qi, D., & Chirikjian, G. S. (2024). RAIL: Robot Affordance Imagination with Large Language Models. arXiv preprint arXiv:2403.19369.(Under review of IROS 2024)(Arxiv link)

[3] Peng, Y., Zhang, C., Hu, C., & Mao, Z. (2024). Integration of Large Language Models in Control of EHD Pumps for Precise Color Synthesis. arXiv preprint arXiv:2401.11500.(Arxiv link)

Projects

(1) RAIL: Robot Affordance Imagination with Large Language Models

• An affordance reasoning pipeline that only requests target affordance names.
• An imagination framework simulates customized profiles for multi-class affordances.
• A real robot manipulation system for performing novel tasks on unseen objects based on affordance reasoning.

(2) Learning to kick: Non-prehensile manipulation for quadruped robots.

• Generation of kicking motion: An effective MPC-based controller applied.
• Data fitting model: Used for making predictions on motion parameters.
• Continuous navigation: A locomotion controller trained by RL for reaching goal.

(3) Investigation on reinforcement learning in hand robot manipulation.

• Exploration in RL task environment settings.
• Performance comparison between different RL algorithms.
• Investigation on hyperparameter sittings and reward functions.
• Effects of parallel training.

(4) Reinforcement learning for underwater robot control. [github repo]

• Design of RL training environment for robot.
• Implementation of multiple RL algorithms and performance comparison.

(4) SLAM and path planning for automatic nagavition of mobile vehicles. [github repo]

• Implementation of SLAM and path planning algorithms.
• Simulation environment of obstacles in physical environments