Implementação em FPGA de método de detecção de novidades adaptativo baseado em DTW para sinais de qualidade de energia utilizando processadores Soft-Core

Loading...
Thumbnail Image

Journal Title

Journal ISSN

Volume Title

Publisher

Universidade Federal de Juiz de Fora (UFJF)

DOI

Abstract

This work presents the implementation of a novelty detector on a Field Pro- grammable Gate Array (FPGA) platform based on the Dynamic Time Warping (DTW) similarity metric. The prototype stands out by avoiding false detection alarms in the presence of frequency variations. It also features adaptive segmentation capabilities and analysis of segments with different lengths, targeting the application of detecting Power Quality Disturbances (PQD) in voltage and current signals of the Electric Power System (EPS). The topic is current, as unknown disturbances may emerge in the new context of Smart Grid (SG) consolidation. The functional prototype was developed using Soft-Core processors to optimize hardware resource usage. For the experimental tests, the Novelty Detection (ND) threshold was determined through empirical simulations, considering a Signal-to-Noise Ratio (SNR) of 35 dB. Bench tests confirm that the design meets the re- quirements for real-time operation and proves effective in detecting novelties. Additionally, the implementation used approximately 4% of the FPGA memory and less than 3% of other available resources.

Description

Citation

Endorsement

Review

Supplemented By

Referenced By

Creative Commons license

Except where otherwise noted, this item's license is described as Acesso Aberto