Granted Patent B2

US 12,352,238 · App. 18/364,592 · Granted Jul 8, 2025

Inventors: Efren Albisu Iso (Barcelona, ES); Mauro Grammatico (Barcelona, ES); Pedro Arroyo Beltri (Barcelona, ES)

Assignee: General Electric Renovables Espana, S.L.

F03D17/011F03D17/017F05B2270/331

Abstract

Claims (34)

1. A method for detecting a state of a wind turbine blade, the method comprising:

receiving a load signal from a sensor configured to measure loads on the wind turbine blade;

determining an energy of the load signal at a first frequency, and determining an energy of the load signal at a second frequency, wherein the first frequency corresponds to a natural frequency of the wind turbine blade in a default state and the second frequency corresponds to a natural frequency of the wind turbine blade with an accumulation of ice above an allowable threshold, wherein the natural frequency of the wind turbine blade is a frequency of a first normal mode or of a second normal mode of the blade in the default state, wherein the first normal mode refers to a lowest resonant frequency and corresponding first oscillation mode, wherein the second normal mode refers to a second lowest resonant frequency and corresponding second oscillation mode;

comparing the energy of the load signal at the first frequency to the energy of the load signals at the second frequency; and

generating a flag signal when the energy of the load signal at the first frequency is smaller than the energy of the load signal at the second frequency.

2. The method of claim 1 , wherein the load signal is representative of an edgewise load acting on the wind turbine blade.

3. The method of claim 1 , wherein the default state of the blade corresponds to an ideal state of the blade.

4. The method of claim 1 , further comprising filtering a rotor rotational speed frequency from the load signal prior to determining the energy of the load signal.

5. The method of claim 4 , wherein the filtering comprises using a notch filter.

6. The method of claim 1 , wherein the load signal is determined at the first and second frequencies using bandpass filters.

7. The method of claim 6 , wherein the energy of the load signal at the first and second frequencies is determined by calculating a Root Mean Square of the load signal at the first and second frequencies.

8. The method of claim 1 , further comprising filtering high frequency content from the load signal before determining the energy of the load signal.

9. The method of claim 8 , wherein the high frequency content is filtered from the load signal using a smoothing filter.

10. The method of claim 1 , further comprising filtering the load signal with an outlier detection filter before determining the energy of the load signal.

11. The method of claim 1 , wherein the flag signal comprises a command to activate a de-icing system.

12. The method of claim 1 , wherein the flag signal comprises a command to stop operation of the wind turbine.

13. A wind turbine, comprising:

one or more blades;

a control system comprising load data corresponding to loads occurring on the one or more blades when an accumulation of ice above an allowable threshold occurs, the control system configured to perform the following operations:

receive a load signal from one or more sensors configured to measure loads on the blades;

determine an energy of the load signal in a first frequency, and determine an energy of the load signal in a second frequency, wherein the first frequency substantially corresponds to a natural frequency of the blade in a default state and the second frequency corresponds to a natural frequency of the wind turbine blade with an accumulation of ice above an allowable threshold, wherein the natural frequency of the wind turbine blade is a frequency of a first normal mode or of a second normal mode of the blade in the default state, wherein the first normal mode refers to a lowest resonant frequency and corresponding first oscillation mode, wherein the second normal mode refers to a second lowest resonant frequency and corresponding second oscillation mode;

compare the energy of the load signal in the first frequency to the energy of the load signal in the second frequency; and

generate a flag signal when the energy of the load signal in the first frequency is smaller than the energy of the load signal in the second frequency.

14. The wind turbine of claim 13 , wherein the sensors comprise strain gauges mounted on the blades.

15. The wind turbine of claim 13 , wherein the sensors are configured to provide signals indicative of edgewise loads acting on the blades.

16. A method for detecting a state of a wind turbine blade, the method comprising:

receiving one or more edgewise load signals from one or more sensors configured to measure edgewise loads on the wind turbine blade;

filtering out of the edgewise load signals a frequency corresponding to a rotor rotational speed frequency;

determining an energy of the filtered edgewise load signals at a first frequency, wherein the first frequency corresponds to a natural frequency of the wind turbine blade in a default state and the second frequency corresponds to a natural frequency of the wind turbine blade with an accumulation of ice above an allowable threshold, wherein the natural frequency of the wind turbine blade is a frequency of a first normal mode or of a second normal mode of the blade in the default state, wherein the first normal mode refers to a lowest resonant frequency and corresponding first oscillation mode, wherein the second normal mode refers to a second lowest resonant frequency and corresponding second oscillation mode;

comparing the energy of the load signals at the first frequency to the energy of the load signals at the second frequency; and

generating a flag signal when the energy of the load signals at the first frequency is smaller than the energy of the load signals at the second frequency band.

17. The method according to claim 16 , wherein determining the energy of the filtered edgewise load signals at the first and second frequencies comprises:

applying bandpass filters for the first and second frequencies to the filtered edgewise load signals; and

calculating a Root Mean Square of the load signals at the first and second frequencies.

Assignments (1)

Priority Claims (1)

Continuity (1)

References Cited (44)