Title: Bionic experiments based on autonomous operant conditioning automata

Authors: Lizhen Dai; Xiaogang Ruan; Jing Chen

Addresses: Institute of Artificial Intelligence and Robots, College of Electronic and Control Engineering, Beijing University of Technology, Beijing 100124, China. ' Institute of Artificial Intelligence and Robots, College of Electronic and Control Engineering, Beijing University of Technology, Beijing 100124, China. ' Institute of Artificial Intelligence and Robots, College of Electronic and Control Engineering, Beijing University of Technology, Beijing 100124, China

Abstract: This paper proposes an autonomous operant conditioning (AOC) automaton that, designs a bionic autonomous learning control method, and constructs an operant conditioning learning system. AOC provides a recursive operation programme. AOC simulates biological operant conditioning mechanism, which has a bionic self-organising feature, including self-learning and adaptive capabilities, and can be used for description, simulation, and design of a variety of self-organising systems. The bionic autonomous learning control method can be used to describe and simulate a bionic autonomous learning process. The learning system can learn on line by interaction with environment, and achieve the best consequences. Based on the above, we apply the model to simulate Skinner-rat experiment to prove that AOC can simulate the learning mechanism of operant conditioning, and apply it to achieving the balancing control of two-wheeled self-balancing robots which shows that AOC can be used to design a variety of intelligent behaviour for robotic systems.

Keywords: autonomy; operant conditioning; bionic autonomous learning; automatons; bionic self-organising; self-learning; learning control; balancing control; two-wheeled robots; self-balancing robots; robot control; intelligent robotics.

DOI: 10.1504/IJMIC.2011.043152

International Journal of Modelling, Identification and Control, 2011 Vol.14 No.4, pp.286 - 293

Published online: 21 Mar 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article