Artificial wrestling: A dynamical formulation of autonomous agents fighting in a coupled inverted pendula framework

Abstract

We develop autonomous agents that fight with each other, inspired by human wrestling. For this purpose, we propose a coupled inverted pendula (CIP) framework in which: (a) the tips of two inverted pendula are linked by a connecting rod, (b) each pendulum is primarily stabilized by a proportional-derivative (PD) controller, (c) and each is additionally equipped with an intelligent controller. Based on this framework, we dynamically formulate an intelligent controller designed to store dynamical correspondence from initial states to final states of the CIP model, to receive state vectors of the model, and to output impulsive control forces to produce the desired final states of the model. By developing a quantized and reduced-order design for this controller, we obtain a practical control procedure based on an off-line learning method. We then conduct numerical simulations to investigate the individual performance of the intelligent controller, and we show that the performance can be improved by adding a delay element. The results show that the performance depends not only on the quantization resolution of the learning data but also on the delay time of the introduced element. Finally, we install the intelligent controllers into both pendula in the proposed framework in order to demonstrate autonomous competitive behavior between inverted pendula.

Keywords

• multiagent system
• competitive problem
• intelligent control
• nonlinear dynamics
• reachable set