Zhang Huaguang, Teng Fei, Sun Qiuye, Shan Qihe
IEEE Trans Cybern. 2019 Aug;49(8):3209-3213. doi: 10.1109/TCYB.2018.2839912. Epub 2018 Jun 7.
A distributed optimization problem based on a continuous-time multiagent system (MAS) disturbed by general noise is considered in this paper. The general noise, under some relaxed assumptions, which may be a stationary process, is proposed to describe the disturbance among agents more accurately. The noise-to-state (NOS) stability of the concerned MAS is analyzed based on an improved theoretical result of random differential equations. Furthermore, the relative sufficient conditions in the form of linear matrix inequality are developed with less conservatism, from which the minimum estimation error between the optimal solution and the NOS stable state of the proposed MAS with general noise can be obtained by choosing some appropriate distributed optimization parameters. One example is used to verify the effectiveness of the proposed approach.
本文考虑了一个基于受一般噪声干扰的连续时间多智能体系统（MAS）的分布式优化问题。在一些放宽的假设下，提出了可能是平稳过程的一般噪声，以便更准确地描述智能体之间的干扰。基于随机微分方程的改进理论结果，分析了相关MAS的噪声到状态（NOS）稳定性。此外，以线性矩阵不等式的形式给出了相对充分条件，且保守性较小，通过选择一些合适的分布式优化参数，可以得到所提含一般噪声的MAS的最优解与NOS稳定状态之间的最小估计误差。通过一个例子验证了所提方法的有效性。
