Neural Reinforcement Learning Controllers for a Real Robot Application

— Accurate and fast control of wheel speeds in the presence of noise and nonlinearities is one of the crucial requirements for building fast mobile robots, as they are required in the MiddleSize League of RoboCup. We will describe, how highly effective speed controllers can be learned from scratch on the real robot directly. The use of our recently developped Neural Fitted Q Iteration scheme allows Reinforcement Learning of neural controllers with only a limited amount of training data seen. In the described application, less than 5 minutes of interaction with the real robot were sufficient, to learn fast and accurate control to arbitrary target speeds.

• Type: Conference
• Year: 2007
• Where: ICRA
• Authors: Roland Hafner, Martin Riedmiller
• Keywords: Effective Speed Controllers, Fast Mobile Robots, ICRA 2007, Real Robot, Robotics