In this paper a novel distributed adaptive dual-layer super-twisting sliding mode observer-based scheme is designed to isolate, reconstruct and mitigate the effects of disturbances and a class of communication attacks affecting both generator nodes and load nodes in power networks. Voltage phase angles are measured at each node by means of Phasor Measurement Units (PMUs). Based on this information, an interconnection of adaptive dual-layer super-twisting sliding mode observers is designed both to estimate frequency deviation in each generator node, and to perform robust detection and reconstruction of both disturbances and a class of communication attacks. The proposed estimation scheme exhibits a distributed structure, since it requires only information received from neighbouring nodes and measurements taken locally in the power network. The novelty of the proposed scheme is its capability to reconstruct simultaneous disturbances affecting the generator nodes and load nodes, automatically adjusting the values of the gains of the observers. More precisely, the adaptive gains of the observer obey a recently proposed dual-layer adaptation law for the super-twisting sliding mode architecture. A disturbance mitigation strategy is also proposed at each generator node utilising the disturbance estimates. Numerical simulations are discussed to assess the proposed distributed scheme.

本文设计了一种新颖的基于分布式自适应双层超扭曲滑模观测器的方案，以隔离，重建和减轻干扰的影响以及一类同时影响发电机节点和负载节点的通信攻击在电力网络中。电压相角是通过相量测量单元（PMU）在每个节点处测量的。基于此信息，设计了自适应双层超扭曲滑模观测器的互连，以估计每个发电机节点中的频率偏差，并执行对干扰和一类通信攻击的鲁棒检测和重建。所提出的估计方案表现出分布式结构，因为它仅需要从相邻节点接收的信息以及在电力网络中本地进行的测量。提出的方案的新颖性在于它能够重建同时影响发电机节点和负载节点的干扰，并自动调整观测者的增益值的能力。更确切地说，观察者的自适应增益服从最近提出的针对超扭曲滑模架构的双层自适应定律。骚动还利用干扰估计在每个发电机节点处提出了缓解策略。讨论了数值模拟，以评估所提出的分布式方案。