Data-based structure-constrained sliding mode control scheme for interconnected multi-machine power systems

This paper is dedicated to the sliding mode control (SMC) problem for interconnected multi-machine power systems (IMMPSs) subjected to false data injection (FDI) attacks in the framework of feedback communication networks (FCNs) with structure constraints. Specifically, to cope with certain communication link disconnections within the IMMPS, we first formulate the structure-constrained FCNs (SFCNs) using a set of encoding matrices and based on this architecture, we design and solve the stabilising gain with sparse block structure. Following this, we propose a structure-constrained SMC scheme to compensate the SFCNs-dependent attack strategies and ensure that the resulting sliding mode dynamics (SMD) is consistent with the actual SFCNs within IMMPS. It is also proved that, by adopting this structure-constrained SMC scheme, the reachability of predefined ideal sliding surface and the asymptotic stability of the corresponding SMD are established. In addition, to overcome the prevalent IMMPS model uncertainty and circumvent conventional initial Hurwitz requirements, we develop an data-based extended algorithm that allows the structure-constrained SMC scheme to be obtained in a data-driven manner, while also exhibiting certain computational efficiency advantages over existing online algorithms. Finally, we separately apply our approach to a three-machine model and a three-area load frequency control (LFC) power system model to demonstrate its effectiveness.