Data-Driven Islanding Detection Using a Principal Subspace of Voltage Angle Differences

Abstract

The likelihood of an unintentional power system islanding is increased in systems with significant penetration of distributed generation. To mitigate the adverse effects of islanding, a quick and reliable islanding detection method is needed. This paper first analyzes covariance matrices of a linearized power system model, and relates them to the principal component analysis of experimentally obtained covariance matrices. Additionally, a new model-independent islanding detection method is proposed that uses measurements of voltage angle differences between multiple locations in the system. The angle differences are first preprocessed to remove the effects of nonstationarity. Thereafter, a probabilistic model of principal component analysis is trained using the acquired measurements. The principal and residual spaces extracted from the measurements are used to discriminate between islanding and other events in the system. The applicability of the proposed method is demonstrated by using real measurements gathered from several locations in a transmission grid.

Microgrid Fault Detection Method Based on Lightweight Gradient Boosting Machine–Neural Network Combined Modeling

“…There are still some shortcomings in this research, and more factors need to be considered for improvement when considering the problem of radial network structure fault detection, so future work will start with the following aspects: (1) The line operation under the networked power system structure is included in the key research direction of combined model fault detection, and the accuracy and stability of model fault detection under the multi-path propagation of faulted lines are considered. (2) The networked line structure of the simultaneous failure of multiple lines is added into the model fault detection focus of the research content. Based on this situation, the model features dual fault detection functions for both networked and radial line structures, aligning more closely with the diverse development requirements of today's power systems.…”

“…When such faults arise, the remaining distributed power sources may inadvertently sustain power in the fault area, creating an "islanding effect". This not only endangers the safety of maintenance personnel but also threatens the normal functioning of equipment [2]. Therefore, the precise detection of fault locations and types, along with the timely implementation of protective measures, has become a critical area of research.…”

Analysis and Verification of Islanding Detection Techniques for Grid-integrated PV Systems

“…The method provided a zero NDZ and a stable system operation since it is observed continuously in real-time. In [11], the authors presented a datadriven IDT with real-time operation capabilities that passes the voltage angle differences measured at different points of the power system to a developed model to detect islanding. The captured data is pre-processed to cancel nonstationarity effects, followed by a probabilistic principal component analysis model training phase.…”

A Statistical Signal Processing Approach to Islanding Detection

“…The signal processing methods usually extract features of the signals obtained in the PCC to perform islanding detection. Some of the most used techniques in these categories are the Wavelet Transform (WT) and its variants [27][28][29][30][31], Stockwell Transform (ST) [32,33], Hilbert Huyang Transform (HHT) [34], Time-time Transform (TTT) [32,35], mathematical morphology (MM) [32], and Principal Component Analysis (PCA) [36][37][38][39]. When the feature space provided by the signal processing methods does not provide a clear signature of the islanding condition, setting a threshold for the detection has been proved to be difficult.…”