Working Paper Article Version 1 This version is not peer-reviewed

A Novel Compressed Deep Stacking Neural Network Based Classifier for the Monitoring and Classification of Multiple Power Quality Disturbances

Version 1 : Received: 17 September 2019 / Approved: 18 September 2019 / Online: 18 September 2019 (05:31:52 CEST)

How to cite: Shen, Y.; Abubakar, M.; Liu, H. A Novel Compressed Deep Stacking Neural Network Based Classifier for the Monitoring and Classification of Multiple Power Quality Disturbances. Preprints 2019, 2019090199 Shen, Y.; Abubakar, M.; Liu, H. A Novel Compressed Deep Stacking Neural Network Based Classifier for the Monitoring and Classification of Multiple Power Quality Disturbances. Preprints 2019, 2019090199

Abstract

Power quality disturbances (PQDs) occur as the use of non-linear load and renewable-based micro-grids increased. This paper presented a novel algorithm that comprised of discrete orthogonal S-transform (DOST), compressive sensing (CS) and deep stacking network (DSN) for automatic monitoring and classification of single and multiple PQDs. DOST- CS based method is employed for feature extraction and reduction of power quality event data. It compressed the extracted feature matrix (orthogonal S-matrix coefficients) to minimize the computational process. Moreover, compressive measurements of 24 types of multiple and nine types of single PQDs events are fed to a proposed DSN classifier for PQD recognition. The DOST-based CS feature extraction technique achieves good robustness and time-frequency localization while retaining useful information. The DSN classifier method utilizes a Batch-mode gradient as a fine-tune which has less noise gradient and improved efficiency of PQD classification. The proposed method was tested with 6680 numbers of real and synthetic, single and multiple PQDs data. 20 dB to 50 dB noise level was considered. Moreover, the proposed method is also tested with a modified IEEE 13 bus system with the wind-grid model dataset. The recently published articles and multiclass support vector machine (SVM) classifier are also compared with the proposed algorithm to examine the performance. The high classification results show that DOST-CS feature extraction and DSN classifier have high precision to recognize multiple power quality event data even in noisy conditions.

Subject Areas

multiple power quality disturbances; discrete orthogonal S-transform; deep stacking neural network; wind-grid distribution

Comments (0)

We encourage comments and feedback from a broad range of readers. See criteria for comments and our diversity statement.

Leave a public comment
Send a private comment to the author(s)
Views 0
Downloads 0
Comments 0
Metrics 0


×
Alerts
Notify me about updates to this article or when a peer-reviewed version is published.