基于多时滞反馈的多智能体协同控制策略研究

项目名称： 基于多时滞反馈的多智能体协同控制策略研究

项目编号： No.61273116

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 自动化技术、计算机技术

项目作者： 欧林林

作者单位： 浙江工业大学

项目金额： 70万元

中文摘要： 随着生物网络合成、微机电系统等新兴技术的出现，迫切要求复杂大规模控制系统向着低功耗、小型化、分散化控制的方向发展。现有基于先进控制算法的高性能控制器是在系统具有较强的执行计算能力的前提条件下提出来的，无法满足这些发展要求。本项目拟采用分布自主式结构和多时滞模块，给出两种具有"慢计算特性"的多智能体反馈控制系统设计策略（包括分布式多时滞控制器设计和局部多时滞控制器设计）。首先解决多时滞控制器对具有分布时滞和复系数的被控对象的镇定这一关键性问题，然后根据全局控制目标，联合个体间的网络拓扑特性，分别构建基于分布式多时滞控制器和局部多时滞控制器的多智能体反馈控制系统。将协同控制问题进行简化，进而利用所获得的镇定性结果，分别设计既能满足全局控制目标又能满足控制性能要求的分布式多时滞控制器和局部多时滞控制器，实现多智能体系统的协同控制，为小型多智能体系统的发展奠定坚实的控制理论基础。

中文关键词： 时滞；多智能体系统；协同控制；镇定性；

英文摘要： With the appearance of new technologies such as bio-network synthesis and MEMS etc., complicated large-scale control systems are urgently required to evolve along the direction of low power-consuming, small-size, and distributed control. The present high-performance controllers based on the advanced control algorithm and proposed on the precondition that the system has a relatively strong ability of calculation, cannot satisfy these developing requirements. This proposal will adopt decentralized autonomy structure and multi-delay module, and present two designing strategies of multi-agent feedback control system having the characteristic of "slow-computing", including the design of decentralized multi-delay controller and local multi-delay controller. The key issue needed to be resolved firstly is the stabilization of multi-delay controller for the controlled plant having the distributed delays and complex coefficents. According to the total control objective and network topological characteristics of agents, the multi-agent feedback control system will be constructed on the basis of decentralized multi-delay controller and local multi-delay controller. The question of cooperative control will be simplified, and the decentralized multi-delay controller and local multi-delay controller will be designed using the

英文关键词： time delay；multi-agent system；cooperative control；stabilization；