CP-WTG Analytics: Uma ferramenta para análise e detecção de pontos de mudança em sinais de dados de turbinas eólicas

Caio D. Gilo; Marcos V. A. Costa; Luiz A. M. Lima; Leiva C. Oliveira

doi:10.20906/SBAI-SBSE-2023/3992

Caio D. Gilo Departamento de Computação, Universidade Federal Rural do Semi-Árido, Mossoró-RN
Marcos V. A. Costa Programa de Pós-graduação em Engenharia Elétrica, Universidade Federal Rural do Semi-Árido, Mossoró-RN
Luiz A. M. Lima Voltalia Energia do Brasil, Rio de Janeiro-RJ
Leiva C. Oliveira Departamento de Computação / Programa de Pós-graduação em Engenharia Elétrica, Universidade Federal Rural do Semi-Árido, Mossoró-RN

DOI: https://doi.org/10.20906/SBAI-SBSE-2023/3992

Keywords: Wind Turbine Generator, Change Points Detection, Python, Ruptures, Time Series, Prediction

Abstract

The wind turbine generator components’ monitore task is useful to detect non-normal operating behavior, and enhances the operation and maintenance processes. The change points indicates when an unusual patterns in the sensor responses time series happened. This work presents a tool for detection and analysis change points in measurement data generated by wind turbine sensors. With simple and intuitive interface, containing the main functionalities for selection, configuration, adjustments, display and storage of the results for the detected change points, and thus facilitates communication between the different professionals involved in the operation. Experiments were carried out from datas of a wind turbine that generated alarm of possible problems in generator. The detected change points indicated non-normal operation behavior with 60 days in advance, compared to generated alarm. Furthermore, different configurations and adjustments were analyzed to demonstrate the tool feaseability.