Fault diagnosis and identification of malfunctioning protection devices in a power system via time series similarity matching

Abstract

Alarm messages uploaded to a dispatch centre following the failure of a power system contain extensive temporal information. The accuracy and speed of fault diagnosis can be improved upon taking full advantage of such temporal information contained in these messages. From this standpoint, a power system fault diagnosis model based on time series similarity matching is proposed herein. First, a set of suspected faulty components can be determined after the occurrence of a fault or a set of faults. Then, on the basis of the specifications of protection devices, including protective relays and circuit breakers, and the defined time series model, a set of alarm hypothesis time series is generated, containing action process information of the protection devices involved in these suspicious components. Meanwhile, the alarm messages received by the dispatch centre are pre-processed, and the alarm information time series are obtained. Subsequently, the fault component is obtained by calculating the similarity between each series in the alarm hypothesis time series set and the alarm information time series. Finally, the correctness and effectiveness of the proposed fault diagnosis model are demonstrated via a real-life scenario in a local power system in China.

Anomaly detection and clustering‐based identification method for consumer–transformer relationship and associated phase in low‐voltage distribution systems

“…This study used the Davies-Bouldin index (DBI) and adjusted rand index (ARI) to evaluate the accuracy of the phase identification [30][31][32].…”

Dynamic Restoration of Active Distribution Networks by Coordinated Repair Crew Dispatch and Cold Load Pickup

“…1 depicts the typical stages of restoring such a faulted ADN. The fast damage assessment aims to identify trouble spots, evaluate the extent of damages, and estimate required restoration resources [24]. Useful information can be obtained from advanced fault diagnosis and identification algorithms [25], consumers' trouble calls, and unmanned inspection systems [26].…”

Data-Driven Anomaly Detection and Event Log Profiling of SCADA Alarms

“…Wang et al formulated a MILP model for identification of false alarms sent by remote bi-directional or unidirectional fault indicators via linearized mathematical expressions [19]. Xu et al proposed a supervised data-driven method that uses dynamic time warping distance to measure similarity between observed and hypothetical alarm sequence and detect faulty protection devices [20].…”

“…The developed method follows the same principle of analytic methods like [17], [18], but it is fully data-driven and does not require solving an optimization problem or construction of mathematical expressions that mimic protection system expected operation like in [18], [19]. Moreover, it does not require a hypothetical time series of alarms, like in the supervised learning approach from [20]. • Data-driven methodology (not based in optimization problems like in [17]) to segment and categorize historical event log data, helping human operators to identify different groups of occurrences in HV lines automatically, and exclusively based on their event profiles (i.e., without using a pre-existing class in a supervised learning fashion, e.g., [5]).…”