An improved protection strategy based on PCC-SVM algorithm for identification of high impedance arcing fault in smart microgrids in the presence of distributed generation. (April 2021)

• Record Type: Journal Article
• Authors: Eslami, Mostafa; Jannati, Mohsen; Tabatabaei, S. Sepehr
• Is Part Of: Measurement. Volume 175(2021)
• Publication Date: 2021-04
• Subjects: Microgrid -- High impedance arcing fault -- Support vector machine -- Principal component analysis -- Pearson correlation coefficient; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2021.109149
• Languages: English
• ISSNs: 0263-2241

Abstract

Highlights: Fast and reliable detection of HIAFs in the microgrids with renewable energy sources. Analyzing a vast variety of useful features and their effect on the fault detection. Using PCC and PCA, as dimension-reduction methods. Investigating unbalanced conditions and delayed asynchrony measurements. High performance on the networks with different configurations. Abstract: High impedance Arcing faults (HIAFs) are normally caused by ruptured conductors, leaning of a tree with high impedance, and/or the presence of a high impedance object between the conductor and earth. In such cases, protections available in the microgrid may not be capable of detecting the HIAFs. Hence, to increase the safety level and reliability of the microgrid, it is essential to take action for fast and reliable detection of these types of faults. Therefore, the present study introduces an appropriate strategy to detect HIAFs using a pattern recognition approach. To this end, different scenarios are implemented in the training data extraction step considering the measurement units embedded in a 25 kV microgrid in the presence of Distributed Generations (DG) and Renewable Energy Sources (RESs) in the EMTP-RV software environment. Then, after the initial processing, the scenarios are scaled-down and compared using the Pearson Correlation Coefficient (PCC) and Principal Component Analysis (PCA) methods. Next, the processed data is classified using the Support Vector Machine (SVM) method by selecting the most appropriate kernel. Simulation results in EMTP-RV and MATLAB environments demonstrate that the proposed strategy is capable of fast detection of HIAFs in microgrids with a high level of accuracy.