Using a Reinforcement Learning Controller to Overcome Simulator/Environment Discrepancies

Authors

Nancy Owens, Brigham Young University
Todd Peterson, Brigham Young University

Keywords

reinforcement, learning controller, simulator, environment discrepancies

College

Physical and Mathematical Sciences

Department

Computer Science

Abstract

Robotic controllers often fail to perform well when transferred from a simulated environment to a real-world situation. Such failures are caused by discrepancies between a simulator and the real-world system which it is intended to model. Traditional approaches to this problem attempt to reduce the number and severity of simulator/environment discrepancies by using calibration, by designing more accurate simulations, or by controlling the specifications of the real-world environment. These approaches, although often effective, are limited in that they require the designer to successfully anticipate the types of discrepancies the controller is likely to encounter.

Recommended Citation

Owens, Nancy and Peterson, Todd (2014) "Using a Reinforcement Learning Controller to Overcome Simulator/Environment Discrepancies," Journal of Undergraduate Research: Vol. 2014: Iss. 1, Article 1184.
Available at: https://scholarsarchive.byu.edu/jur/vol2014/iss1/1184