Mu Yunfei, Zhang Huaguang, Yan Yuqing, Xie Xiangpeng
IEEE Trans Cybern. 2023 Oct;53(10):6737-6747. doi: 10.1109/TCYB.2022.3232531. Epub 2023 Sep 15.
This article focuses on the distributed robust fault estimation problem for a kind of discrete-time interconnected systems with input and output disturbances. For each subsystem, by letting the fault as a special state, an augmented system is constructed. Particularly, the dimensions of system matrices after augmentation are lower than some existing related results, which may help to reduce calculation amount, especially, for linear matrix inequality-based conditions. Then, a distributed fault estimation observer design scheme that utilizes the associated information among subsystems is presented to not only reconstruct faults, but also suppress disturbances in the sense of robust H optimization. Besides, to improve the fault estimation performance, a common Lyapunov matrix-based multiconstrained design method is first given to solve the observer gain, which is further extended to the different Lyapunov matrices-based multiconstrained calculation method. Thus, the conservatism is reduced. Finally, simulation experiments are shown to verify the validity of our distributed fault estimation scheme.
本文聚焦于一类具有输入和输出干扰的离散时间互联系统的分布式鲁棒故障估计问题。对于每个子系统，通过将故障视为一种特殊状态，构建了一个增广系统。特别地，增广后系统矩阵的维数低于一些现有相关结果，这可能有助于减少计算量，尤其是对于基于线性矩阵不等式的条件。然后，提出了一种利用子系统间关联信息的分布式故障估计观测器设计方案，不仅能重构故障，还能在鲁棒H优化意义下抑制干扰。此外，为提高故障估计性能，首先给出一种基于公共Lyapunov矩阵的多约束设计方法来求解观测器增益，该方法进一步扩展为基于不同Lyapunov矩阵的多约束计算方法。从而降低了保守性。最后，通过仿真实验验证了所提出的分布式故障估计方案的有效性。
