cv

About

I am Shuyuan Wang, born in Harbin, China. I am now pursuing my phd degree at University of British Columbia, supervised by Prof.Bhushan Gopaluni. Before that, I obtained my master degree at Harbin Institute of Technology, supervised by Prof.Lixian Zhang, and bachelor degree from BeiJing JiaoTong University. I study control theory, reinforcement learning, control and planning in robotics. I have a strong foundation on mathematics, physics and control theory. I enjoy the feeling of observing that things could work in the real world after design and analysis. In my spare time, I like playing oboe, running, reading, meditation and studying philosophy.

Education:

University of British Columbia(20-present)
Research direction: Process control and Reinforcement learning
Supervisor: Bhushan Gopaluni, UBC

Harbin Institute of Technology(18-20)
Major: Control Science and Engineering
Supervisor: Lixian Zhang, HIT, Fellow of IEEE

BeiJing JiaoTong University (14-18)
Bachelor of Engineering
Major: Electrical Engineering, grade: 89.2/100
Minor: Information and Computer Science.
Thesis supervisor: Shengbo Eben Li, Tsinghua University

Publications:

  1. “Guaranteed performance control of switched linear systems: A differential-Riccati-equation-based approach”,Wang, D., Wang, S., Yuan, S. et al. Peer-to-Peer Netw. Appl. (2019) 12: 1810.[link]
  2. “Switching control of a Mecanum wheeled mobile robot for vision-based tracking with intermittent image losses”, The 2019 IEEE International Conference on Systems, Man and Cybernetics (SMC).[Link][oral presentation]
    Lixian Zhang, Shuyuan Wang, Bo Cai, Tianhe Liu, Yiming Cheng.(Supervisor serves as first author and communication author)

Selected Projects:

  1. “Ballbot platform: design and control”, BeiJing JiaoTong University, Mar 17-Jun 17.


Ballbot is an underactuated robot with 3 omnidirectional wheels driven by 3 DC motors to keep the robot standing on a ball. In this project, I debug and adjust simulation platform of the Ballbot based on Simscape and help my collaborator to verify the mathematical model.




  1. Internship in FIMI, FIMI Corporation, Jul 17.


During the internship, I worked for a project of UAV logistics. To protect the item from shaking and collision, and also protect motors and drive-circuits, my main work is designing a piecewise function which can generate acceleration which is time-derivable for drone. Under the maximum speed and acceleration constraint, the goal of the function is to make the drone achieve the given destination in the shortest time.

3. “Parallel reinforcement learning with multi-agent”, id-Lab in Tsinghua University, 17-18.

This is my Bachelor thesis. In this research, multi agents are used to improve the exploring speed of RL. A scheme is developed to allow multi agents to cooperate with each other. Conclusions are given about the connection between acceleration of the scheme and the amount of agents.[code] [slides]

4. Survey for robot-arm manipulation, Nov 18.

In this survey, I summarize some advanced techniques about robot-arm manipulation from a human-machine interaction point of view. Main attention is focused on how human instruction can improve the performance of the robot and therefore reduce the design complexity. Chinese version has gained 3000+ views. English version is going to appear.

Ongoing projects:

  1. Quadruped robot control with multi-modality perception


The initial goal is to control the robot walking on complex terrain with high performance. Both visual and tactile sensors will be used. To fully utilize the two kinds of information, an advanced machine learning technique called "cross-modalities fusion" will be used, which finds the connection and transfer between the two kinds of information. I am going to use Stanford Doggo as the platform, which is an open-source quadruped robot. To prepare for the project, I am now learning "Underactuated System" in OCM,MIT, and studying the technical documents of Stanford Doggo.

Academic activities:

Reviewer:

  • IEEE Transactions on Cybernetics
  • Optimal Control, Applications and Methods

Released news:

  • On Dec 24, 2016, our interesting design: “A device converting potential energy in magnetic feild into kinetic energy” was specially reported by the TV program “Is It True?“ on channel CCTV2 [video]!(start from 35min57s)
  • On Oct 7, 2019, I was glad to give an oral presentation on the paper “Switching control of a Mecanum wheeled mobile robot for vision-based tracking with intermittent image losses” at the IEEE International Conference on Systems, Man and Cybernatics, at Bari, Italy!

Selected Awards:

  • Outstanding undergraduate award (Rate: <10%), issued by BeiJing JiaoTong University;
  • Meritorious winner of Mathematical Contest In Modeling (Rate: <10% worldwide), conducted by The Consortium for Mathematics and Its Application(COMAP).
  • Major award on academic performance (Rate:<30%), issued by Graduate School, Harbin Institute
    of Technology.

Contact:

Email: antergravity@gmail.com;18s104163@stu.hit.edu.cn