复杂多智能体系统的分布式优化分析与设计

复杂多智能体系统的分布式优化研究是当前复杂大型系统研究的核心问题之一。本项目将基于我们近年来在此方面已取得的成果，主要研究在复杂环境下具有复杂动态的多智能体系统进行分布式优化收敛性分析与有效算法设计。研究主要包括以下几个方面：（1）给出在外部环境干扰下的分布式抗干扰优化设计方法，以及针对量测误差给出基于近似梯度的优化方法并提供收敛性分析；（2）考虑多智能体具有复杂动态和复杂任务时的协同优化和控制设计问题，特别是对多层次的非线性智能体分布式优化控制进行研究；（3）考虑某些通讯约束或不确定性并解决基于量化的或处理具有随机因素（噪声等）情况下的分布式优化收敛性以及定量的优化误差分析。总之，我们将利用多种优化与控制方法有机的结合，给出一套处理相当一大类复杂多智能体系统的分布式优化分析和设计方法。

Distributed optimization of complex multi-agent systems is one of the key problems in the study of complex large-scale systems. Based on the results obtained on the topic by us in recent years, this project mainly investigates the convergence analysis and effective algorithm design of distributed optimization for multi-agent systems with complex dynamics in complex environments. The main research topics include: (1) to propose distributed disturbance-rejection optimization method in the presence of external environmental disturbance, and propose approximate-gradient-based optimization method along to deal with measurement uncertainties, along with convergence analysis; (2) to consider the cooperative optimization and control design when the multi-agent system has complex dynamics or completes complex tasks, especially to study the optimization and conrol of hierarchical and nonlinear multi-agent systems; (3) to consider the communication constraints and uncertainties and solve the convergence of distributed optimization algorithms to deal with the quantization and random factors (e.g., noises) and the quantitative optimization error analysis. In a sum, we will effectively integrate some good methods of optimization and control, in order to estabilish a package of methods to handle distributed optimization analysis and design for quite a large class of complex multi-agent systems.