Research on Condition Monitoring of Key Components in wind Turbine based on Cointegration Analysis

Abstract

Traditional models based on temperature trend can not comprehensively monitor the condition of wind turbine, besides, the threshold obtained from model can cause fluctuation when wind turbine operates under different conditions. In view of this, this paper presents a novel methodology based on cointegration analysis for continuously monitoring the operating conditions of wind turbine. The first step is to determine the optimal combination of parameters from normal SCADA sample data, by using the cointegration test. Then, perform the method of cointegration analysis to calculate the cointegration residuals and stationary threshold line under normal working space. When wind turbine component is operating out of the normal track, the data sample in cointegration residuals will exceed the stationary threshold line, and send the alert signals. The method is tested using SCADA data of wind turbine with known faults, the results demonstrate the proposed method can effectively monitor the abnormal state of generator and gear box, and provide the function of early warning.

Cointegration Modelling for Health and Condition Monitoring of Wind Turbines - An Overview

“…First of all, it should be noted that the major idea of the cointegration-based condition monitoring and fault detection techniques for wind turbines [12][13][14][15][16][17][18][19][20][21][22][23], as reviewed in Section 3, is basically relied on the well-known control chart approach, which is one of the primary techniques of statistical process control. Basically, control charts plot the quality characteristic as a function of the sample number.…”

“…Next, it should be mentioned that regression analysis can be used for condition monitoring of wind power systems, as illustrated in [24]. However, cointegration analysis has been used in [12][13][14][15][16][17][18][19][20][21][22][23] instead of other regression techniques is due to two main reasons: (1) to avoid the problem of spurious regression; and (2) to actively deal with the undesired effect of environmental and operational conditions in the analysed data. The former has been discussed broadly in the econometrics literature [25].…”

“…Following the idea of the method developed in [12][13][14], the work in [15] also used the theory of cointegration for continuously monitoring the operating conditions of wind turbines. First, the optimal combination of different parameters from SCADA data in normal operating condition was determined by using the Johansen's cointegration test and statistical test.…”

“…It should be noted here that all applications -reported in [8][9][10] for SHM systems and in [12][13][14][15][16][17][18][19][20] for wind turbine condition monitoring -have used the linear cointegration theory that was originally developed in [6,7] and intimately connected with the concept of linear error correction models. However, it is well known that response signals (e.g.…”

Influences of the Temperature Variations in the Gondola of the Goldwing S50/750 Wind

“…"The monitoring of components through SCADA is crucial in the technical condition of wind turbines. By studying parameters such as temperature, the working conditions of the main aggregates can be known" [29].…”