Electrical and Computer Engineering Location: Online
Add to Calendar 2020-04-13T15:00:00 2020-04-13T15:00:00 America/New_York Master's Thesis Defense for Linyuan Guo Optimal Control of a Two-wheeled Self-balancing Robot by Reinforcement Learning Online

Join Zoom Meeting

https://virginia.zoom.us/j/7603288244

Meeting ID: 760 328 8244

One tap mobile

+16465588656,,7603288244# US (New York)

+17866351003,,7603288244# US (Miami)

Optimal Control of a Two-wheeled Self-balancing Robot by Reinforcement Learning

Abstract:

The two-wheeled self-balancing robot is a typical robot, which has potential application prospects in many areas, such as transportation and exploration. Design and control of the two-wheeled self-balancing robot have attracted substantial attention in both academia and industry over the past decades. The two-wheeled self-balancing robot is an inherently unstable, high-order, multivariable, nonlinear, strongly coupled system and this robot is an underactuated mechanical system. 

This thesis concerns optimal control of the linear motion, tilt motion, and yaw motion of a two-wheeled self-balancing robot. Traditional optimal control methods for the two-wheeled self-balancing robot usually require a precise model of the system, and other control methods exist that achieve stabilization in the face of parameter uncertainties. In practical applications, it is often desirable to realize optimal control in the absence of the precise knowledge of the system parameters. This thesis proposes to use a new feedback-based reinforcement learning method to solve the linear quadratic regulation (LQR) control problem for the two-wheeled self-balancing robot. The proposed control scheme is completely online and does not require any knowledge of the system parameters. The proposed input decoupling mechanism and pre-feedback law overcome the commonly encountered computational difficulties in implementing the learning algorithms, which the former shortens the learning transient phase and the latter improves the system performance. Both state feedback optimal control and output feedback optimal control are presented. Numerical simulation shows that the proposed optimal control scheme is capable of stabilizing the system and converging to the LQR solution obtained through solving the algebraic Riccati equation.