A method to design plug-and-play (PnP) distributed controllers for large-scale nonlinear systems represented by interconnected Takagi_Sugeno fuzzy models with nonlinear consequent is presented in this article. From the combination of techniques to use multiple fuzzy summations and to explore the chordal decomposition of the interconnection graph associated with the large-scale nonlinear system, sufficient conditions for distributed stabilization are derived in terms of linear matrix inequalities (LMIs). Conditions specially designed to allow seamless subsystems plugging-in and unplugging operations from the large-scale system, without requiring the redesign of all previously tuned distributed controllers, are provided. The approach can be used together with fault detection and isolation (FDI) systems, and also in the context of mixed distributed and decentralized controllers operating in a network of interconnected systems. To illustrate the effectiveness of the proposed PnP approach, a network of nonlinearly coupled and heterogeneous Van der Pol oscillators is used in the numerical experiments.

中文翻译：

---

大规模非线性系统的即插即用分布式控制


本文提出了一种针对以具有非线性结果的互连 Takagi_Sugeno 模糊模型为代表的大规模非线性系统的即插即用 (PnP) 分布式控制器设计方法。通过结合使用多个模糊求和的技术以及探索与大规模非线性系统相关的互连图的弦分解，根据线性矩阵不等式（LMI）导出了分布式稳定的充分条件。提供了专门设计的条件，允许从大型系统进行无缝子系统插入和拔出操作，而不需要重新设计所有先前调整的分布式控制器。该方法可以与故障检测和隔离（FDI）系统一起使用，也可以在互连系统网络中运行的混合分布式和分散控制器的背景下使用。为了说明所提出的 PnP 方法的有效性，在数值实验中使用了非线性耦合和异构 Van der Pol 振荡器网络。