基于内模原理的多智能体蜂拥编队跟踪控制  

Multi-agent Flocking Formation and Trajectory Tracking Control under Internal Model Principle

作　　者：黄蓉 周军[1] 黄浩乾 HUANG Rong;ZHOU Jun;HUANG Haoqian(School of Energy and Engineering,Hohai University,Nanjing Jiangsu 211100,China)

机构地区：[1]河海大学能源与电气学院,江苏南京211100

出　　处：《计算机仿真》2022年第4期375-380,共6页Computer Simulation

基　　金：国家自然科学基金(61573001)。

摘　　要：对一类具有加权矩阵的扩展蜂拥控制算法驱动下的多智能体的稳态编队存在条件及其暂态/稳态特征进行了说明。基于虚拟领导者-平均值状态方程,给出了其内模补偿增益和镇定反馈增益的设计方法与算法步骤,实现了对多智能体的蜂拥编队形成及其编队轨迹跟踪的多目标控制。提案方法的优势有,编队控制参数和轨迹跟踪控制参数可以分别独立设计;权重矩阵对编队形态可在一定范围内调整。仿真算例验证了提案控制策略的可行性和有效性。Firstly, a class of generalized flocking control algorithms defined with the matrix-weighting Euclidean norm were recalled, and ultimate existence and main properties of the multi-agent formations driven under them were explicated carefully. Secondly, based on leader-average state-space equation, the design method and algorithm steps of the internal mode compensation gain and the stabilization feedback gain were given, and the formation of expected formation as well as its trajectory tracking of multiple agents were realized. In particular, the flocking formation and trajectory tracking control algorithms can be designed separately;Formation can be adjusted by means of the weighting matrix. Finally, numerical examples were illustrated to show efficiency and performances of the suggested approach.

关 键 词：多智能体蜂拥 编队控制 内模 无静差轨迹跟踪 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]