bachelorThesis
Desenvolvimento de um algoritmo para detecção de restrição de potência na operação de turbinas eólicas
Fecha
2022-07-20Registro en:
RIBEIRO, Thatiana Jéssica da Silva. Desenvolvimento de um algoritmo para detecção de restrição de
potência na operação de turbinas eólicas. 2022. 56f. Trabalho de Conclusão de Curso (Graduação em Engenharia de Computação) - Centro de Tecnologia, Universidade Federal do Rio Grande do Norte, Natal, 2022.
Autor
Ribeiro, Thatiana Jéssica da Silva
Resumen
In wind turbines operation the power curve analysis play an important role for monitoring the performance of the wind turbines. In this curve, the relationship between the power generated by the turbine and the wind speed is shown. A characteristic of this curve is the presence of points of normal operation distributed in the vicinity of the power curve provided by the manufacturer and the presence of abnormal operation points, which are called outliers. These abnormal points are caused by several factors, such as failures in the measurement system, turbine failures, wind curtailments, among others. Wind curtailments consists of limiting the power generation capacity of the wind turbines in wind farms due to limitations in the energy transmission system, causing economic losses. As operationally this power restriction procedure is performed in a non-automated way, when the limitation of energy transmission capacity ceases, the wind turbine often continues to operate under power restriction. The identification of wind curtailments will allow faster action by the team on site so that the operation returns to normal standards. Therefore, in this study an algorithm was developed in order to identify wind curtailments points in the power curve of wind turbines based on operating data from different wind turbines collected by the SCADA system. The logic implemented in the algorithm is based on the identification of normal operating points based on the idea of identifying a normal operating envelope and then further analyze the points outside this envelope, which are the outliers, aiming to find the points of wind curtailments from the calculation of the distance between the points, being this distance calculated separately in relation to the horizontal axis (wind speed) and the vertical axis (generated power) and also in the idea of connectivity between these points. Satisfactory results were obtained in the identification of wind curtailments in most of the analyzed months. The adjusting the hyperparameters allows the user to make the identification process more rigorous or flexible.