分布式多机器人运动控制的离散事件系统方法

传统多机器人系统的运动控制主要依赖于机器人的动力学方程或运动学方程,通过求解微分方程组来获得机器人的输入控制信号.随着系统中机器人数量的增加和运行环境的复杂化,动力学方程很难描述多机器人系统的运动行为,且无法很好地解决诸如死锁等逻辑故障.本文简略综述了国内外的研究现状,重点介绍笔者所在研究组开展的关于离散事件系统方法在多机器人运动控制方面的应用性研究工作.其动机在于:1)基于离散事件系统方法的运动控制能够有效地解决系统运行过程中产生的诸如死锁等逻辑故障.首先,利用离散事件系统模型对多机器人系统的运动进行建模,从而降低计算复杂性;其次,基于所得离散事件系统模型,设计分布式安全运动控制算法,使各个机器人可以自主地、无碰撞地、无死锁地运动;设计分布式鲁棒运动控制算法,使得失效的机器人对系统的影响最小.2)基于离散事件系统方法的运动控制策略可以结合传统的基于运动学方程的运动控制方法,从而使系统不但能够避免顶层的逻辑故障,而且能够确定机器人执行器的输入信号.

Distributed approaches to motion control of multiple robots via discrete event systems

Classical motion control of multi-robot systems is dependent on the dynamic or kinematic equations, where the input control signals for robots are obtained by solving a set of differential equations. With the increase of the number of robots in a system and the more complexity of environment, it is hard to describe the behavior of robots exactly by only applying the dynamic equations and cannot deal well with some logical problems, such as deadlocks. This paper summarizes the state-of-the-art motion control technologies, especially emphasizes our research work with regard to the application of discrete event systems (DESs) on the motion control of multi-robot systems. The motivations are as follows. 1) DES-based motion control can deal with the logical problems, such as deadlocks, during the evolution of the system. First, use the DES models to formalize the motion of a multi-robot system; this can reduce the computation complexity. Second, based on the DES models, design distributed safe control algorithms for robots to avoid collisions and deadlocks, as well as guarantee that they can move independently; design distributed robust control algorithms such that the failures of robots have the least detrimental effect on the system. 2) DES-based motion control can be combined with the kinematic equation-based motion control such that the system can not only resolve high-level logical problems but also determine the input signals of the actuators of robots.