Reinforcement learning of cooperative behaviors for multi-robot tracking of multiple moving targets

Traditional reinforcement learning algorithms learn based on discrete/finite states and actions, thus limit the learned behaviors to discrete/finite space. To address this problem, this paper introduces a distributed reinforcement learning controller that integrates reinforcement learning with behavior based control networks. This learning controller can enable the robot to generate appropriate control policy which combines different elementary behaviors. In addition, to address the problems in concurrent learning, a distributed learning control algorithm is proposed to coordinate concurrent learning processes. The distributed reinforcement learning controller and learning control algorithm are applied to multi-robot tracking of multiple moving targets. The efficacy is demonstrated by simulations.

I. Introduction

For multi-robot systems, one of the key research problems is to achieve cooperation among robots by decentralized (distributed) control methodology [1]. Normally, the desired cooperation is in the task level [2], which means that the common mission is broken down into tasks, and robots choose different tasks (roles) according to the state and behave differently. However, the design for the task level controller is quite difficult. Therefore, in recent years, machine learning techniques have been proposed and studied for multi-robot systems that aim to enable the robots to learn how to cooperate without the need for human design or coding.