Modal decomposition of the dynamic response of wind turbine during one year of continuous monitoring

Abstract

Summary The analysis of the dynamic behavior of wind turbines is a complex task. These structures present diverse configurations, dependent on the wind conditions. Consequently, the dynamic behavior is highly dependent on their operating conditions. This paper analyzes the dynamic behavior of a 2.0 MW wind turbine during 1 year of operation. In that sense, post‐processing tools aiming to decompose the measured vibration response of structures into modal responses, originally developed to be applied with the SSI‐DATA algorithm, are implemented and applied. Initially, an original data processing that permits the application of these tools to two additional output‐only modal identification algorithms is introduced. Then, these tools are applied for the first time to data collected in a wind turbine during a period of 1 year. The obtained results evidenced the effectiveness of the proposed post‐processing procedure and the distinct dynamic response of the wind turbine according to its operating conditions.

Constrained mode decomposition method for modal parameter identification

“…By comparison of Equation (10) and Equation ( 12), if the combination coefficients α satisfy the following equation:…”

“…Time-domain methods include the widely used eigensystem realization algorithm (ERA) [1][2][3] and autoregressive moving average (ARMA) models, [4][5][6] the Ibrahim time domain (ITD) approach, 7,8 and stochastic subspace identification (SSI) method. [9][10][11] Frequency-domain methods include the simplest peak picking method, 12 maximum likelihood identification, 13 and the frequency-domain decomposition method. 14 These classical methods use the relevant knowledge of structural vibration theory to establish a mathematical model of the relationship between the vibration response and modal parameters, 15,16 which is then used to identify and possibly optimize the modal parameters.…”

“…Classical modal parameter identification methods are usually formulated and operate in time domain or in frequency domain. Time‐domain methods include the widely used eigensystem realization algorithm (ERA) 1–3 and autoregressive moving average (ARMA) models, 4–6 the Ibrahim time domain (ITD) approach, 7,8 and stochastic subspace identification (SSI) method 9–11 . Frequency‐domain methods include the simplest peak picking method, 12 maximum likelihood identification, 13 and the frequency‐domain decomposition method 14 .…”

Monopile Foundation Stiffness Estimation of an Instrumented Offshore Wind Turbine through Model Updating

“…Tese approaches span from input-state-parameter estimation methods, which enable estimation of both the properties and input excitations [3], through to operational modal analysis (OMA) methods [4], which have been used to infer damage or change from deviations in structural modal properties [5]. Te application of a number of OMA methods to OWT monitored data has been explored by diferent groups, such as the work of Oliveira et al [6], which showed that stochastic subspace identifcation (SSI) methods can provide robust estimates of OWT modal properties over a range of operational modes. Such methods can be extended to directly identify structural properties using a second stage of model updating, where numerical structural model parameters are optimised to best match the predicted and measured modal behaviour.…”