Multiagent systems are increasingly becoming popular among researchers spanning multiple fields of study. However, existing studies only models communication interaction between agents as either fixed or switching topologies described by crisp graphs supported by algebraic graph theories. In this paper, we propose an alternative approach to describing agent interactions using fuzzy graphs. Our approach is aimed at opening up new research avenues and defining new problems in coordination control especially in terms of dynamics between agents’ states, graph topologies and coordination objectives. This paper studies distributed coordination on fuzzy graphs where the edge-weights modeling network topologies are dependent on the states of the agents in the network. In hindsight, the network weights are adjustable based on the situational state of the agents. First, we introduce the concept of fuzzy graphs and give some distinguishing features from the crisp or fixed graphs. Next, we provide some membership functions to define the state-dependent weights and finally we use some simulations to demonstrate the convergence of the proposed consensus algorithms especially for cases where the agents are subject to system failures.

多主体系统越来越受到跨多个研究领域的研究人员的欢迎。但是，现有研究仅将代理之间的通信交互建模为固定的或交换的拓扑，这些拓扑由代数图理论支持的明快图描述。在本文中，我们提出了一种使用模糊图描述代理交互的替代方法。我们的方法旨在开辟新的研究途径并定义协调控制中的新问题，尤其是在代理状态，图形拓扑和协调目标之间的动态方面。本文研究模糊图上的分布式协调，其中边缘权重建模网络拓扑取决于网络中代理的状态。事后看来，网络权重可根据座席的情境状态进行调整。首先，我们介绍模糊图的概念，并给出清晰图或固定图的一些区别特征。接下来，我们提供一些隶属函数来定义状态相关的权重，最后我们使用一些仿真来证明所提出的共识算法的收敛性，特别是对于代理易遭受系统故障的情况。