Open Access

ARTICLE

Extraction of Strain Characteristic Signals from Wind Turbine Blades Based on EEMD-WT

Jin Wang¹, Zhen Liu^1,*, Ying Wang¹, Caifeng Wen^2,3, Jianwen Wang^2,3

1 School of Mechanical Engineering, Inner Mongolia University of Technology, Hohhot, 010051, China
2 School of Energy and Power Engineering, Inner Mongolia University of Technology, Hohhot, 010051, China
3 Inner Mongolia Key Laboratory of Wind energy and Solar Energy Utilization Mechanism and Optimization, Hohhot, 010051, China

* Corresponding Author: Zhen Liu. Email:

(This article belongs to the Special Issue: Wind Energy Development and Utilization)

Energy Engineering 2023, 120(5), 1149-1162. https://doi.org/10.32604/ee.2023.025209

Received 28 June 2022; Accepted 14 November 2022; Issue published 20 February 2023

Abstract

Analyzing the strain signal of wind turbine blade is the key to studying the load of wind turbine blade, so as to ensure the safe and stable operation of wind turbine in natural environment. The strain signal of the wind turbine blade under continuous crosswind state has typical non-stationary and unsteady characteristics. The strain signal contains a lot of noise, which makes the analysis error. Therefore, it is very important to denoise and extract features of measured signals before signal analysis. In this paper, the joint algorithm of ensemble empirical mode decomposition (EEMD) and wavelet transform (WT) is used for the first time to achieve sufficient noise reduction and effectively extract the feature signals of non-stationary strain signals. The application process of EEMD-WT is optimized. This optimization can avoid the repeated selection of wavelet basis function and the number of decomposition layers due to different crosswind conditions. EEMD adaptively decomposes the strain signal into intrinsic mode functions, to judge the frequency of IMFs, remove the high-frequency noise components, retain the useful components. The useful components are denoised twice by the wavelet transform, the components and residual terms after the secondary denoising are reconstructed to obtain the characteristic signal. The EEMD-WT was applied to process the simulating signals and measured the strain signals. The results were compared with the results of the EEMD. The results showed that the EEMD-WT method has better noise reduction performance, and can effectively extract the characteristics of strain signals, which lays a solid foundation for accurate analysis of wind turbine blade strain signals under crosswind conditions.

---

Keywords

---