A PCA-Based Consistency and Sensitivity Approach for Assessing Linkage Methods in Voltage Sag Studies

Fabricio Alves De Almeida,Luiz Gustavo De Mello,Pedro Paulo Balestrassi,Jose Henrique De Freitas Gomes,Jacques Miranda Filho,Guilherme Ferreira Gomes,Anderson Paulo De Paiva,Estevao Luiz Romao

doi:10.1109/access.2021.3088436

Fabricio Alves De Almeida, Luiz Gustavo De Mello + Show 6 more

Open Access

https://doi.org/10.1109/access.2021.3088436

Copy DOI

Journal: IEEE Access	Publication Date: Jan 1, 2021
Citations: 2	License type: CC BY 4.0

Affiliation: Federal University of Itajubá

Abstract

In the light of Brazilian energy regulatory context, cluster strategies are required to classify groups of substations for voltage sag purposes. Tuning cluster algorithms is not a trivial task, due to the fact that these methods are sensitive to small errors. Therefore, this study proposes a new methodology based on principal components analysis (PCA), attribute agreement and analysis of covariance to verify the level of consistency and sensitivity of the linkage methods in the cluster formation for voltage sag studies. In order to prove this methodology, real data from power quality indices of distribution substations are used. Four distinct scenarios with disturbances are evaluated. PCA is applied for dimensionality reduction of the data. Then, grouping is performed for eight different linkage methods and agreement analysis is applied. Ward method was the only one that presented 100% consistency in all scenarios, considered as the most robust method whereas k-means showed consistency of 94.11%, with inversion of the clusters. However, when evaluating their groupings, it was found that k-means was unable to adequately separate the groups for this dataset. Finally, the proposed methodology is adequate for choose cluster methods for extensive data and it can be extended to applications in different areas.

Highlights

Quality improvements are widely studied in several power quality (PQ) sectors, where the quality of generation and distribution significantly influences industrial sectors [1]
From the confidence intervals by the k-means method (Fig. 6), it was found that it performed clusters with close means values and with long confidence intervals, showing the lack of precision in the estimation for voltage sag analysis
--Performing the application in PQ data for substations, the Ward linkage method showed 100% consistency, demonstrating to be a robust alternative when analyzing this data set

Summary

Introduction

Quality improvements are widely studied in several power quality (PQ) sectors, where the quality of generation and distribution significantly influences industrial sectors [1]. It is possible to verify that several studies focused on PQ, investigate the phenomenon of voltage sag applying different strategies, in which, we can highlight: the use of evolutionary algorithm to optimize the allocations of PQ monitors in distribution systems [3]; use of battery energy storage systems in the investigation of voltage sag and voltage deviation problems in distribution networks [4]; a new approach to asses equipment trip using fuzzy probabilities and possibility distribution in order to mitigate voltage sag [5]; simulations of different strategies to identify voltage sag sources [6]; the use of non-hierarchical linkage method of kmeans for PQ event recognition [7]; and the use of convolutional neural networks with weighted k-nearest neighbor classifier for identification of voltage sag events [8]; a methodology which can be applied as a voltage sag mitigation solution to distribution of utility in a group of customers installing a dynamic voltage restorer [9] These and other studies infer the importance of using modern strategies to investigate the phenomenon of voltage sag for the power quality distribution.

Methods

Results

Conclusion