A method for assessing and locating protection measurement loop errors based on an improved similarity algorithm

• Published in: Electrical engineering Vol. 107; no. 3; pp. 3725 - 3736
• Main Authors: Mu, Zong Jun, Liu, Deng Xin, Hu, Bin, Li, Zhen
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.03.2025; Springer Nature B.V
• Subjects: Accuracy; Algorithms; Economics and Management; Electrical Engineering; Electrical Machines and Networks; Energy Policy; Engineering; Entropy; Entropy (statistics); Error detection; Euclidean geometry; Fault diagnosis; Localization; Neural networks; Original Paper; Power Electronics; Principal components analysis; Signal processing; Similarity; Similarity measures; Weighting methods; Entropy weighting method; Protection of measuring; Error localization; EWM-EDTS
• ISSN: 0948-7921; 1432-0487
• DOI: 10.1007/s00202-024-02704-1

It is of importance to detect and locate the errors of the protection measurement loop of the relay protection device for ensuring correct and timely functioning. The complex and changeable power system environment makes error detection and localization challenging. To this end, this paper proposes an entropy weight method-Euclidean distance and Tanimoto similarity (EWM-EDTS)-based method that integrates Euclidean distance and Tanimoto similarity with the entropy weight method. The Euclidean distance algorithm and Tanimoto similarity algorithm are used to calculate and obtain the similarity values between two sequences of samples and then the entropy weight method is used to calculate the weighting coefficients to fuse the two sequences of similarity values to finally obtain the EWM-EDTS distance. By comparing the value of the EWM-EDTS distance with the distance threshold, potential errors in the measurement data can be accurately located and identified. The simulation based on PSCAD shows that the proposed method can significantly improve the accuracy of error detection and localization.