The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Prever o momento do mau funcionamento das turbinas eólicas é essencial para manter a elevada rentabilidade do negócio de geração de energia eólica. Métodos de aprendizado de máquina foram estudados usando dados do sistema de monitoramento de condições, como dados de vibração e dados de controle de supervisão e aquisição de dados (SCADA), para detectar e prever anomalias em turbinas eólicas automaticamente. As técnicas baseadas em autoencoder atraíram um interesse significativo na detecção ou previsão de anomalias por meio de aprendizagem não supervisionada, na qual o padrão da anomalia é desconhecido. Embora tenha sido comprovado que as técnicas baseadas em autoencoder detectam anomalias de maneira eficaz usando dados SCADA relativamente estáveis, elas funcionam mal no caso de dados SCADA deteriorados. Nesta carta, propomos um método de filtragem de curva de potência, que é uma técnica de pré-processamento usada antes da aplicação de uma técnica baseada em autoencoder, para mitigar a sujeira dos dados SCADA e melhorar o desempenho de previsão da degradação de turbinas eólicas. Avaliamos seu desempenho usando dados SCADA obtidos de um parque eólico real.
Masaki TAKANASHI
Toyota Central Research and Development Laboratories Incorporated
Shu-ichi SATO
Toyota Central Research and Development Laboratories Incorporated
Kentaro INDO
Eurus Technical Service Corporation
Nozomu NISHIHARA
Eurus Technical Service Corporation
Hiroto ICHIKAWA
Eurus Energy Holdings Corporation
Hirohisa WATANABE
Toyota Tsusho Corporation
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copiar
Masaki TAKANASHI, Shu-ichi SATO, Kentaro INDO, Nozomu NISHIHARA, Hiroto ICHIKAWA, Hirohisa WATANABE, "Anomaly Prediction for Wind Turbines Using an Autoencoder Based on Power-Curve Filtering" in IEICE TRANSACTIONS on Information,
vol. E104-D, no. 9, pp. 1506-1509, September 2021, doi: 10.1587/transinf.2020EDL8127.
Abstract: Predicting the malfunction timing of wind turbines is essential for maintaining the high profitability of the wind power generation business. Machine learning methods have been studied using condition monitoring system data, such as vibration data, and supervisory control and data acquisition (SCADA) data, to detect and predict anomalies in wind turbines automatically. Autoencoder-based techniques have attracted significant interest in the detection or prediction of anomalies through unsupervised learning, in which the anomaly pattern is unknown. Although autoencoder-based techniques have been proven to detect anomalies effectively using relatively stable SCADA data, they perform poorly in the case of deteriorated SCADA data. In this letter, we propose a power-curve filtering method, which is a preprocessing technique used before the application of an autoencoder-based technique, to mitigate the dirtiness of SCADA data and improve the prediction performance of wind turbine degradation. We have evaluated its performance using SCADA data obtained from a real wind-farm.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2020EDL8127/_p
Copiar
@ARTICLE{e104-d_9_1506,
author={Masaki TAKANASHI, Shu-ichi SATO, Kentaro INDO, Nozomu NISHIHARA, Hiroto ICHIKAWA, Hirohisa WATANABE, },
journal={IEICE TRANSACTIONS on Information},
title={Anomaly Prediction for Wind Turbines Using an Autoencoder Based on Power-Curve Filtering},
year={2021},
volume={E104-D},
number={9},
pages={1506-1509},
abstract={Predicting the malfunction timing of wind turbines is essential for maintaining the high profitability of the wind power generation business. Machine learning methods have been studied using condition monitoring system data, such as vibration data, and supervisory control and data acquisition (SCADA) data, to detect and predict anomalies in wind turbines automatically. Autoencoder-based techniques have attracted significant interest in the detection or prediction of anomalies through unsupervised learning, in which the anomaly pattern is unknown. Although autoencoder-based techniques have been proven to detect anomalies effectively using relatively stable SCADA data, they perform poorly in the case of deteriorated SCADA data. In this letter, we propose a power-curve filtering method, which is a preprocessing technique used before the application of an autoencoder-based technique, to mitigate the dirtiness of SCADA data and improve the prediction performance of wind turbine degradation. We have evaluated its performance using SCADA data obtained from a real wind-farm.},
keywords={},
doi={10.1587/transinf.2020EDL8127},
ISSN={1745-1361},
month={September},}
Copiar
TY - JOUR
TI - Anomaly Prediction for Wind Turbines Using an Autoencoder Based on Power-Curve Filtering
T2 - IEICE TRANSACTIONS on Information
SP - 1506
EP - 1509
AU - Masaki TAKANASHI
AU - Shu-ichi SATO
AU - Kentaro INDO
AU - Nozomu NISHIHARA
AU - Hiroto ICHIKAWA
AU - Hirohisa WATANABE
PY - 2021
DO - 10.1587/transinf.2020EDL8127
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E104-D
IS - 9
JA - IEICE TRANSACTIONS on Information
Y1 - September 2021
AB - Predicting the malfunction timing of wind turbines is essential for maintaining the high profitability of the wind power generation business. Machine learning methods have been studied using condition monitoring system data, such as vibration data, and supervisory control and data acquisition (SCADA) data, to detect and predict anomalies in wind turbines automatically. Autoencoder-based techniques have attracted significant interest in the detection or prediction of anomalies through unsupervised learning, in which the anomaly pattern is unknown. Although autoencoder-based techniques have been proven to detect anomalies effectively using relatively stable SCADA data, they perform poorly in the case of deteriorated SCADA data. In this letter, we propose a power-curve filtering method, which is a preprocessing technique used before the application of an autoencoder-based technique, to mitigate the dirtiness of SCADA data and improve the prediction performance of wind turbine degradation. We have evaluated its performance using SCADA data obtained from a real wind-farm.
ER -