复杂非完整多自主体网络协同算法设计与性能极限分析

项目名称： 复杂非完整多自主体网络协同算法设计与性能极限分析

项目编号： No.61503133

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 自动化技术、计算机技术

项目作者： 陈超洋

作者单位： 湖南科技大学

项目金额： 22万元

中文摘要： 本项目从复杂非完整多自主体网络的自组织、非完整、异质、自治等特性以及航天飞行器、深海探测器、移动机器人等实际应用需求出发，研究节点和通信存在不完整约束的网络，构建能反映网络拓扑结构特征的复杂非完整多自主体网络的动态演化模型。从节约自主体能量和减小网络通信代价的角度入手，设计针对节点和通信非完整的基于事件驱动的混合型协同算法和针对环境非完整的事件驱动性粒子群优化协同算法。对复杂非完整多自主体网络不同的协调行为（一致性、跟踪），提出相应普适的性能指标，通过设计协同优化算法得到对应的性能极限。最后，搭建多机器人群组的实物仿真平台验证其理论成果。本项目将为复杂非完整多自主体网络的建模、协同控制和性能极限分析做出贡献，同时为实际应用提供新思想和新方法。

中文关键词： 多自主体网络；非完整性约束；协同控制；性能极限

英文摘要： Based on the characteristics such as self-organization, nonholonomic, heterogeneity, autonomy of complex nonholonomic multi-agent network, as well as the application requirements including spacecraft, deep-sea detector, mobile robots, this project establishes the dynamic models of the complex nonholonomic multi-agent networks in the presence of nonholonomic constraints of nodes and communication. Taking into consideration of energy saving and communication cost decreasing, we construct event-triggered hybrid cooperative algorithm of network of nonholonomic nodes and communications, and event-triggered particle swarm optimization cooperative algorithm under the nonholonomic environment. For different coordination behaviors of complex nonholonomic multi-agent networks such as consensus and tracking, the related performance indexes of universal significance will be refined, and corresponding performance limitations will be derived by designing collaborative optimization algorithms. At last, the theoretical results will be verified on the multi-robot group simulation platform. This project will make contributions to modeling, coordination control and analysis of performance limitation for complex nonholonomic multi-agent networks, and provide new ideas and methods for practical application.

英文关键词： Multi-agent networks;Nonholonomic constraints;Cooperative control;Performance limitations