I am a undergraduate student at Tsinghua University, majoring in Mechanics & Interdisciplinary Engineering.
Currently, I am a Research Assistant at Tsinghua Robot Control Lab (RCL) in Department of Automation, advised by Professor Mingguo Zhao.
Learning Humanoid Ball Manipulation Through Legged Locomotion
Robot Control LabTsinghua University
Without using motion capture data and a predefined reference trajectory for dribbling, we have used reinforcement learning to train a humanoid robot to dribble in the simulator and have successfully transferred the policies into the real world.
In real-world experiments, the robot is able to receive ball's target linear velocity commands given by a joystick and dribble through an environment with obstacles.
The completed comparative experiments have preliminarily proved that our method is more efficient than that of using a joystick to directly command the linear and angular velocities of the robot’s base.
Currently, we are trying to conduct more experiments to further verify the performance of our method.
SkyRover: Air-Ground Robots for Low-Altitude Air Delivery Scenarios
DISCOVER LabTsinghua University
We have developed a simulation and physical platform, SkyRover, for air-ground robots, and we have enabled SkyRover to possess capabilities including perception, SLAM, and navigation.
Real-world experiments have demonstrated that SkyRover can efficiently complete simple delivery tasks in various scenarios, preliminarily validating the feasibility of low-altitude delivery in a micro urban community.
Peter: a Fully Automatic Fruit and Vegetable Peeling Machine Based on Arduino and Traditional Control Theory
DISCOVER LabTsinghua University
We have used Solidworks to design a fully automatic fruit and vegetable peeling machine, which covers multiple functional modules such as peeling, material transportation, and self-cleaning.
We have successfully built a prototype through 3D printing.
RoboCup Experiences
Top Eight in RoboCup 2024 Humanoid League KidSize Soccer Competition
Team LeaderEindhoven, Netherlands
As the team leader, I led the team to successfully develop the localization algorithm, powerful shooting skill for strikers and goalkeeping skills for the goalkeeper.
In a total of 5 games, we scored 20 goals and conceded only 2, demonstrating extremely high scoring efficiency and overall competitiveness.
Our team entered the top eight of the competition and has grown into a world-class team.
1st Place in RoboCup China 2024 Humanoid League KidSize Soccer Competition
Team LeaderFujian, China
As the team leader, I led all the team members to enhance the stability of robot's hardware and improve the robustness of the decision-making system.
Finally, with the winning goal scored by robot Jack, we won the championship, the first championship in the team’s 20-year history.
4th Place in RoboCup Asia-Pacific 2023 Humanoid League KidSize Soccer Competition
Team MemberPyeongchang, South Korea
We competed with and learned from other teams in the Asia-Pacific competition and won fourth place.
This competition revealed the shortcomings of our robot, which allowed us to start modifying, improving and optimizing it.
2nd Place in RoboCup China 2023 Humanoid League KidSize Soccer Competition
Team MemberFujian, China
This was my first time to represent Tsinghua University in the RoboCup competition.
In the 4v4 match, I was responsible for debugging and operating the robot named Carol.
Despite the ups and downs, we finally won the second place.