多智能体系统的自适应群集分布式优化

Distributed optimization for adaptive flocking of multi-agent systems

中文关键词  代价函数优化  分布式优化  自适应多智能体系统  群集

英文关键词  Cost Function Optimization  Distributed Optimization  Adaptive Approach Multi-Agent System  Flocking

中文摘要

      本文对具有非线性函数群集行为的连续时间多智能体系统的分布式优化问题进行了研究。本文的目的是 使局部代价函数之和最小。每个智能体只知道与其对应的代价函数。为了解决这一问题，本文设计了一个分布式 控制律，在这个研究中该控制律仅仅依赖于自己和邻居的速度。通过李雅普诺夫稳定性证明了多智能体系统的收 敛性，而且在最小化局部代价函数之和的同时所有智能体可以避免碰撞。最后，通过一个仿真案例来说明所获得 的分析结果。

英文摘要

      A distributed optimization problem is investigated for continuous-time multi-agent systems with flocking behavior of a nonlinear continuous function. This paper aims at minimizing the sum of local cost functions. Each cost function can only be known to its corresponding agent. To solve this problem, a distributed control law is designed, in which each agent depends only on its own velocities and neighbor’s velocities in this study. Then, the convergence of the multi-agent systems is proved by Lyapunov stability. Furthermore, the agents can avoid inter-agent collision while minimizing the sum of local cost functions. Finally, using a simulation case to illustrate the obtained analytical results.