System and method for detecting damage in rotating machinery

Abstract

A method for detecting damage in a bearing coupled to a rotating shaft of a rotating machine includes receiving one or more measurement signals from one or more first sensors monitoring movement of the rotating shaft in one or more directions over a time period. The method also includes removing an effect of one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals over the time period. After the removal, the method includes analyzing a change in the one or more measurement signals from the one or more first sensors, wherein a change in the one or more measurement signals exceeding a predetermined threshold or having a certain magnitude indicates bearing damage. Further, the method includes implementing a corrective action when the change in the one or more measurement signals exceeds the predetermined threshold.

Description

Technical Field

[0001] The present disclosure relates generally to rotating machines, and more particularly to systems and methods for detecting damage in such rotating machines. Background Art

[0002] Wind power is considered to be one of the cleanest and most environmentally friendly energy sources currently available, and wind turbines have received increasing attention in this regard. Modern wind turbines typically include a tower, a generator, a gearbox, a nacelle, and one or more rotor blades. The rotor blades utilize the known airfoil principle to capture the kinetic energy of the wind. The rotor blades transfer kinetic energy in the form of rotational energy, thereby turning a low-speed main shaft that couples the rotor blades to a gearbox, or if a gearbox is not used, directly to a generator. For example, a generator may be coupled to the low-speed main shaft so that the rotation of the shaft drives the generator. For example, the generator may include a high-speed generator shaft that is rotatably coupled to the main shaft via a gearbox. The generator then converts the mechanical energy from the rotor into electrical energy, which can be deployed to a utility grid.

[0003] In addition, modern wind turbines include multiple high-speed and low-speed bearings to provide rotation of its various components. For example, a low-speed main shaft typically includes one or more main bearings mounted at its front and rear ends to allow the low-speed main shaft to rotate around an axis.

[0004] Detecting damaged components in a wind turbine (or any rotating machine) is critical to minimizing the turbine's unplanned downtime and increasing the turbine's availability. Additionally, the main bearings are large components in the nacelle and are very expensive to replace. As such, the faster preventative or corrective action is taken against a damaged main bearing that allows the main shaft to move laterally toward the gearbox, the longer the life of the main bearing can be extended. In extreme cases, such main shaft movement can damage many other components on the wind turbine, thereby increasing costs. Some environmental and / or operating conditions can also cause the main shaft to move laterally. This movement can be confused with lateral movement due to degradation or wear of the main bearings.

[0005] For at least the foregoing reasons, there is a need in the art for improved systems and methods for earlier detection of damage in rotating machines, such as wind turbines. Summary of the Invention

[0006] Aspects and advantages of the invention will be set forth in part in the following description, or may be obvious from the description, or may be learned through practice of the invention.

[0007] In one aspect, the present disclosure relates to a method for detecting damage in a bearing of a rotating shaft coupled to a rotating machine. The method includes receiving one or more measurement signals from one or more first sensors for monitoring movement of the rotating shaft in one or more directions over a time period. The method also includes removing the effects of one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals over the time period. After removing the effects, the method includes analyzing changes in the one or more measurement signals from the one or more first sensors, wherein changes in the one or more measurement signals exceeding a predetermined threshold or having a certain magnitude indicate bearing damage. In addition, the method includes implementing corrective action when the change in the one or more measurement signals exceeds the predetermined threshold.

[0008] In an embodiment, the change in the one or more measurement signals may be a decrease in the one or more measurement signals.

[0009] In another embodiment, the method includes automatically and adaptively learning the effects of one or more environmental and / or operating conditions. Thus, in some embodiments, removing the effects of one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals during the time period may include automatically and adaptively removing the effects of the one or more environmental and / or operating conditions on the movement of the rotating shaft from the one or more measurement signals. In some embodiments, the environmental and / or operating conditions may include, for example, wind speed, wind direction, gusts, wind shear, temperature, time of day, air density, generator speed, rotor speed, power output, thrust, and / or torque.

[0010] In further embodiments, the method may include automatically and adaptively learning and eliminating the effects of environmental and / or operating conditions on the movement of the rotating shaft via a machine learning regression model.

[0011] In additional embodiments, the machine learning regression model may utilize at least one of linear regression, nonlinear regression, support vector regression, gradient boosting regression, decision tree regression, random forest regression, generalized linear model, kernel regression, or a neural network.

[0012] In another embodiment, removing the effects of one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals during the time period may include determining a predicted measurement signal of the one or more first sensors via a regression model and subtracting the predicted measurement signal from the raw measurement signal of the one or more measurement signals to obtain a corrected measurement signal. In such an embodiment, the corrected measurement signal removes the effects of lateral motion due to main bearing degradation or wear. If the corrected measurement signal exceeds a threshold or exhibits an increasing or decreasing trend, main bearing wear may be indicated.

[0013] Thus, in certain embodiments, analyzing changes in one or more measurement signals may include comparing the corrected measurement signal to a predetermined threshold or determining whether the variation in the corrected measurement signal has a certain magnitude.

[0014] In a particular embodiment, the rotating machine may be a wind turbine. As such, the rotating shaft may be a main shaft of the wind turbine, and the bearing may be a main bearing of the wind turbine.

[0015] In several embodiments, the one or more first sensors may be one or more proximity sensors.

[0016] In yet another embodiment, implementing corrective action may include generating an alarm, scheduling a maintenance and / or repair procedure, and / or corrective action other than shutting down the wind turbine.

[0017] In another aspect, the present disclosure relates to a system for detecting damage in a main bearing coupled to a main shaft of a wind turbine. The system includes: one or more first sensors for monitoring movement of the main shaft in one or more directions; and one or more second sensors for monitoring one or more environmental and / or operating conditions of the wind turbine. In addition, the system includes a controller communicatively coupled to the one or more first sensors and the second sensor. The controller is configured to perform a plurality of operations, including but not limited to: receiving one or more measurement signals from the one or more first sensors over a time period; removing the effects of one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals over the time period; analyzing a decrease in one or more measurement signals from the one or more first sensors, wherein a decrease in the one or more measurement signals exceeds a predetermined threshold or has a certain magnitude indicating damage to the main bearing; and implementing corrective action when a decrease in the one or more measurement signals exceeds a predetermined threshold. It should also be understood that the system may also include any additional features described herein.

[0018] In yet another aspect, the present disclosure relates to a method for detecting damage in a bearing of a rotating shaft coupled to a rotating machine. The method includes receiving one or more measurement signals from one or more first sensors for monitoring the movement of the rotating shaft. The method also includes automatically and adaptively learning the effects of one or more environmental and / or operating conditions on the movement of the rotating shaft via a machine learning regression model during a training period. Furthermore, during a correction period, the method includes automatically and adaptively eliminating the effects of one or more environmental and / or operating conditions on the movement of the rotating shaft. Furthermore, the method includes analyzing a decrease in the one or more measurement signals after eliminating the effects of the one or more environmental and / or operating conditions. Furthermore, the method includes implementing a corrective action when the decrease in the one or more measurement signals from the one or more sensors exceeds a predetermined threshold or has a certain magnitude. It should also be understood that the method may also include any additional features and / or steps described herein.

[0019] Technical Solution 1. A method for detecting damage in a bearing coupled to a rotating shaft of a rotating machine, the method comprising:

[0020] receiving one or more measurement signals from one or more first sensors for monitoring movement of the rotating shaft in one or more directions over a period of time;

[0021] removing effects of one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals during the time period;

[0022] analyzing, after the removal, a change in the one or more measurement signals from the one or more first sensors, wherein a change in the one or more measurement signals exceeding a predetermined threshold or having a magnitude indicates bearing damage; and

[0023] Corrective action is implemented when the change in the one or more measurement signals exceeds the predetermined threshold.

[0024] Technical Solution 2. The method according to Technical Solution 1 is characterized in that the change in the one or more measurement signals includes a decrease in the one or more measurement signals.

[0025] Technical Solution 3. The method according to Technical Solution 1, further comprising monitoring the one or more environmental and / or operating conditions of the rotating machine via one or more second sensors; and

[0026] The effect of the one or more environmental and / or operating conditions on the movement of the rotating shaft is automatically and adaptively learned from the one or more measurement signals.

[0027] Technical Solution 4. The method according to Technical Solution 3, wherein removing the influence of the one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals within the time period further comprises:

[0028] The effects of the one or more environmental and / or operating conditions on the movement of the rotating shaft are automatically and adaptively removed from the one or more measurement signals.

[0029] Technical Solution 5. The method according to Technical Solution 4 is characterized in that it also includes automatically and adaptively learning and eliminating the influence of the environment and / or operating conditions on the movement of the rotating axis via a machine learning regression model.

[0030] Technical Solution 6. The method according to Technical Solution 5 is characterized in that the machine learning regression model utilizes at least one of linear regression, nonlinear regression, support vector regression, gradient boosting regression, decision tree regression, random forest regression, generalized linear model, kernel regression or neural network.

[0031] Technical Solution 7. The method according to Technical Solution 5, characterized in that removing the influence of the one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals within the time period further comprises:

[0032] determining a predicted measurement signal of the one or more first sensors via the regression model; and

[0033] The predicted measurement signal is subtracted from an original measurement signal of the one or more measurement signals to obtain a corrected measurement signal.

[0034] Technical Solution 8. The method according to Technical Solution 7 is characterized in that analyzing the changes in the one or more measurement signals also includes comparing the corrected measurement signal with the predetermined threshold or determining whether the change in the corrected measurement signal has the certain amplitude.

[0035] Technical Solution 9. The method according to Technical Solution 3 is characterized in that the rotating machine includes a wind turbine, the rotating shaft includes a main shaft of the wind turbine, and the bearing includes a main bearing of the wind turbine.

[0036] Technical Solution 10. The method according to Technical Solution 9 is characterized in that the environmental and / or operating conditions include at least one of wind speed, wind direction, gusts, wind shear, temperature, time of day, air density, generator speed, rotor speed, power output, thrust or torque.

[0037] Technical Solution 11. The method according to Technical Solution 1 is characterized in that the one or more first sensors include one or more proximity sensors.

[0038] Technical Solution 12. The method according to Technical Solution 1 is characterized in that the one or more directions include at least a transverse direction relative to the longitudinal direction of the rotation axis.

[0039] Technical Solution 13. The method according to Technical Solution 1 is characterized in that implementing the corrective action also includes at least one of generating an alarm or arranging a maintenance and / or repair procedure.

[0040] Technical Solution 14. A system for detecting damage in a main bearing coupled to a main shaft of a wind turbine, the system comprising:

[0041] one or more first sensors for monitoring movement of the spindle in one or more directions;

[0042] one or more second sensors for monitoring one or more environmental and / or operating conditions of the wind turbine; and

[0043] a controller communicatively coupled to the one or more first sensors and the one or more second sensors, the controller configured to perform a plurality of operations comprising:

[0044] receiving one or more measurement signals from the one or more first sensors over a period of time;

[0045] removing effects of one or more environmental and / or operating conditions of the wind turbine from the one or more measurement signals during the time period;

[0046] analyzing a decrease in the one or more measurement signals from the one or more first sensors, wherein a decrease in the one or more measurement signals exceeding a predetermined threshold or having a magnitude indicates main bearing damage; and

[0047] Corrective action is implemented when the decrease in the one or more measurement signals exceeds the predetermined threshold.

[0048] Technical Solution 15. The system according to Technical Solution 14 is characterized in that the multiple operations also include automatically and adaptively learning the impact of the one or more environmental and / or operating conditions on the movement of the spindle from the one or more measurement signals.

[0049] Technical Solution 16. The system according to Technical Solution 15 is characterized in that the environmental and / or operating conditions include at least one of wind speed, wind direction, gusts, wind shear, temperature, air density, generator speed, rotor speed, power output or torque.

[0050] Technical Solution 17. The system according to Technical Solution 15, wherein removing the influence of the one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals within the time period further comprises:

[0051] The effects of the environmental and / or operating conditions on the movement of the rotating shaft are automatically and adaptively learned and eliminated via a machine learning regression model.

[0052] Technical Solution 18. The system according to Technical Solution 17, wherein removing the influence of the one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals within the time period further comprises:

[0053] determining a predicted measurement signal of the one or more first sensors via the regression model; and

[0054] The predicted measurement signal is subtracted from an original measurement signal of the one or more measurement signals to obtain a corrected measurement signal.

[0055] Technical Solution 19. The system according to Technical Solution 18 is characterized in that analyzing the reduction in the one or more measurement signals from the one or more first sensors also includes comparing the corrected measurement signal with the predetermined threshold or determining whether the change in the corrected measurement signal has the certain amplitude.

[0056] Technical Solution 20. A method for detecting damage in a bearing coupled to a rotating shaft of a rotating machine, the method comprising:

[0057] receiving one or more measurement signals from one or more first sensors for monitoring movement of the rotating shaft;

[0058] automatically and adaptively learning, via a machine learning regression model, during a training period, an effect of one or more environmental and / or operating conditions on the movement of the rotational axis;

[0059] automatically and adaptively canceling the effects of said one or more environmental and / or operating conditions on said movement of said rotating shaft during a correction period; and

[0060] analyzing a decrease in the one or more measurement signals after eliminating the effects of the one or more environmental and / or operating conditions; and

[0061] Corrective action is implemented when the decrease in the one or more measurement signals from the one or more sensors exceeds a predetermined threshold or has a certain magnitude.

[0062] These and other features, aspects and advantages of the present invention will become better understood with reference to the following description and appended claims.The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the present invention and, together with the description, serve to explain the principles of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] A full and enabling disclosure of the invention, including the best mode thereof, to one of ordinary skill in the art is set forth in the specification which proceeds with reference to the accompanying drawings in which:

[0064] Figure 1 A perspective view showing a wind turbine according to one embodiment of the present disclosure;

[0065] Figure 2 showing a perspective view of the interior of a nacelle of a wind turbine according to one embodiment of the present disclosure;

[0066] Figure 3 A cross-sectional view showing one embodiment of a drive train of a wind turbine according to the present disclosure;

[0067] Figure 4 A block diagram illustrating one embodiment of suitable components that may be included in a wind turbine controller according to the present disclosure;

[0068] Figure 5 A flow chart illustrating one embodiment of a method for detecting damage in a bearing of a rotating machine according to the present disclosure;

[0069] Figure 6 a timeline diagram illustrating one embodiment of a rolling training period and a calibration period for a controller according to the present disclosure;

[0070] Figure 7 A detailed schematic diagram illustrating one embodiment of a training period and a calibration period of a controller according to the present disclosure;

[0071] Figure 8A a graph illustrating one embodiment of raw proximity sensor measurement data (y-axis) versus time (x-axis) according to the present disclosure; and

[0072] Figure 8B A graph illustrating one embodiment of corrected proximity sensor measurement data (y-axis) versus time (x-axis) in accordance with the present disclosure. DETAILED DESCRIPTION

[0073] Reference will now be made in detail to embodiments of the present invention, one or more examples of which are illustrated in the accompanying drawings. Each example is provided by explaining the present invention rather than by limiting the present invention. In fact, it will be apparent to those skilled in the art that various modifications and variations may be made to the present invention without departing from the scope or spirit of the present invention. For example, the features shown or described as part of one embodiment may be used together with another embodiment to produce another embodiment. Therefore, it is intended that the present invention covers such modifications and variations and their equivalents within the scope of the appended claims.

[0074] Generally speaking, the present disclosure relates to systems and methods for detecting damaged components in rotating machines such as wind turbines. More specifically, the present disclosure provides early detection of wind turbine main bearing failures using main shaft proximity sensors. Slow drift in the measurement signal from the main shaft proximity sensor can indicate a problem in the main bearing. Slightly worn main bearings allow the main shaft to move laterally, thereby reducing the proximity sensor measurement signal. However, variations in the proximity sensor measurement signal can also be affected by multiple additional factors, such as wind speed and temperature. As such, the systems and methods of the present disclosure extract and monitor only the component of the proximity sensor measurement signal that is due to an impending main bearing failure. More specifically, the systems and methods of the present disclosure automatically and adaptively learn from the raw measurement signal and eliminate the effects of known and systematic causes. The method can therefore detect drift much earlier than existing methods and provide earlier warning of main bearing failures.

[0075] Referring now to the accompanying drawings, Figure 1 A perspective view of one embodiment of a wind turbine 10 according to the present disclosure is shown. Although the present disclosure is described with reference to a wind turbine, it should be understood that the systems and methods of the present disclosure may be applicable to any rotating machine (e.g., a gas turbine, a steam turbine, or any other turbine system for power generation). As shown, the wind turbine 10 includes a tower 12 extending from a support surface 14, a nacelle 16 mounted on the tower 12, and a rotor 18 coupled to the nacelle 16. The rotor 18 includes a rotatable hub 20 and at least one rotor blade 22 coupled to and extending outwardly from the hub 20. For example, in the illustrated embodiment, the rotor 18 includes three rotor blades 22. However, in alternative embodiments, the rotor 18 may include more or less than three rotor blades 22. Each rotor blade 22 may be spaced about the hub 20 to facilitate rotating the rotor 18, thereby enabling kinetic energy to be converted from wind energy into usable mechanical energy and subsequently into electrical energy. For example, the hub 20 may be rotatably coupled to a drive system 28 ( Figure 2 ), to allow the generation of electrical energy.

[0076] Now refer to Figure 2 , shows a simplified interior view of one embodiment of a nacelle 16 of a wind turbine 10 housing a drive train 28. As shown, the drive train 28 includes at least a generator 24 disposed within the nacelle 16. Generally speaking, the generator 24 may be coupled to the rotor 18 of the wind turbine 10 for generating electrical energy from the rotational energy generated by the rotor 18. For example, the rotor 18 may include a main shaft 30 coupled to the hub 20 for rotation therewith. The generator 24 may then be coupled to the main shaft 30 such that rotation of the main shaft 30 drives the generator 24. For example, in the illustrated embodiment, the generator 24 includes a generator shaft 29 rotatably coupled to the main shaft 30 via a gearbox 34. However, in other embodiments, it should be understood that the generator shaft 29 may be directly rotatably coupled to the main shaft 30. Alternatively, the generator 24 may be directly rotatably coupled to the main shaft 30. It should be understood that the main shaft 30 may generally be supported within the nacelle 16 by a support frame or bedplate 36 positioned atop the wind turbine tower 12.

[0077] like Figure 1 and Figure 2 As shown, the wind turbine 10 may also include a turbine control system or turbine controller 26 within the nacelle 16. For example, Figure 2 As shown, the turbine controller 26 is disposed within a control cabinet 38 mounted to a portion of the nacelle 16. However, it should be understood that the turbine controller 26 may be disposed at any location on or within the wind turbine 10, at any location on the support surface 14, or at substantially any other location. The turbine controller 26 may generally be configured to control various operating modes (e.g., startup or shutdown sequences) and / or components of the wind turbine 10.

[0078] Each rotor blade 22 may also include a pitch adjustment mechanism 40 configured to rotate each rotor blade 22 about its pitch axis 42 via a pitch bearing 44. Similarly, wind turbine 10 may include one or more yaw drive mechanisms 46 communicatively coupled to controller 26, wherein each yaw drive mechanism 46 is configured to change the angle of nacelle 16 relative to the wind (e.g., by engaging a yaw bearing 48 of wind turbine 10 to rotate nacelle 16 about a yaw axis).

[0079] In addition, if Figure 2As shown, wind turbine 10 may also include one or more sensors 52, 53, 54, and 55. For example, as shown, one or more first sensors 52, 54 may be configured to monitor movement of main shaft 30 in one or more directions. Such first sensors may be, for example, proximity sensors positioned near main shaft 30 to monitor movement of shaft 30 in one or more directions in order to detect damage to main bearing 39 as described herein. Furthermore, as shown, one or more second sensors 53, 55 may be configured to monitor or measure any suitable environmental and / or operating conditions that may affect lateral movement of shaft 30, such as, for example, wind sensor 53 and / or additional sensors 55 positioned within nacelle 16 to monitor various operating conditions of drive train 28. Thus, it should be understood that the first and second sensors 52, 54, 53, 55 described herein may be any suitable sensors capable of monitoring movement of main shaft 30, as well as any other shaft within wind turbine 10. Furthermore, it should be understood that wind turbine 10 may include any suitable number of sensors for monitoring such movement.

[0080] Now refer to Figure 3 , a detailed cross-sectional view of the drive train 28 of wind turbine 10 is shown to further illustrate its various components. As mentioned, drive train 28 includes at least generator 24 and gearbox 34. Furthermore, as shown, generator 24 includes a generator rotor 25 and a generator stator 27. As is generally known in the art, generator rotor 25 is the rotating component of generator 24, while stator 27 is the stationary component of generator 24. Furthermore, in certain embodiments, generator 24 may be a doubly-fed induction generator (DFIG). However, it should be understood that generator 24 according to the present disclosure is not limited to a DFIG generator and may include any generator suitable for providing power to wind turbine 10 according to the present disclosure. Generally speaking, rotor blades 16 rotate the generator rotor 25 of generator 24. As such, generator rotor 25 may be operably connected to hub 18. Thus, operation of rotor blades 16 rotates rotor hub 18, which in turn rotates generator rotor 25, thereby operating generator 24.

[0081] Furthermore, as shown, the low-speed main shaft 30 is configured to provide an input rotational speed to the gearbox 34. For example, the hub 18 may be mounted to the main shaft 30. As shown, the main shaft 30 may include a main flange 41 configured to engage a mating flange (not shown) on the hub 18 to mount the hub 18 to the main shaft 30. Thus, during operation of the wind turbine 10, the rotational speed of the rotor blades 16 may be directly transmitted to the main shaft 30 through the hub 18 as the input rotational speed.

[0082] The main shaft 30 may extend through and be supported by at least one support housing 35 or a plurality of support housings 35. For example, a first housing 35 and, in some embodiments, a second housing (not shown) may be provided to support the main shaft 30. Furthermore, the housing(s) 35 may include one or more main bearings 39 configured to interact with the main shaft 30. For example, as shown, the housing(s) 35 may include a locating bearing 39 (also referred to herein as a main shaft bearing 39) disposed therein, while the second housing may include a floating bearing (not shown) disposed therein. It should be understood that the present disclosure is not limited to locating bearings and floating bearings positioned in the housings as described above, and that the figures are provided for illustrative purposes only. Furthermore, as shown, the main shaft bearing(s) 39 may include an inner race 31, an outer race 32, and a plurality of roller elements 33 disposed therebetween.

[0083] Still refer to Figure 3 As described herein, the gearbox 34 may be a planetary gearbox 34. As such, the gearbox 34 may be configured to convert an input rotational speed from the main shaft 30 into an output rotational speed. In one embodiment, the output rotational speed may be faster than the input rotational speed. Alternatively, however, the output rotational speed may be slower than the input rotational speed. In one embodiment, the gearbox 34 may be a single-stage gearbox. Thus, as discussed below, the input rotational speed may be converted to the output rotational speed via various mating gears in a single stage. Alternatively, however, the gearbox 34 may be a multi-stage gearbox, and the input rotational speed may be converted to the output rotational speed via various mating gears in multiple stages.

[0084] More specifically, the illustrated embodiment of the planetary gearbox 34 includes a fixed annular gear 45 and a plurality of rotatable gears. As such, the fixed annular gear 45 supports the various rotatable gears configured therein. Furthermore, the fixed annular gear 45 includes various axes around which the rotatable gears rotate. In certain embodiments, the planetary gearbox 34 may also include a fixed annular gear 45, one or more rotatable planetary gears 47, and a rotatable sun gear 49. For example, in one embodiment, the planetary gearbox 34 may include four planetary gears 47. However, it should be understood that more or less than four planetary gears 47 are within the scope and spirit of the present disclosure. Furthermore, each rotatable gear in the planetary gearbox 34 includes a plurality of gear teeth (not shown). As such, the teeth may mesh together so that the various gears 45, 47, 49 mesh with each other.

[0085] In some embodiments, carrier 43 can drive planetary gearbox 34. Thus, carrier 43 and main shaft 30 can be coupled such that the input rotational speed of main shaft 30 is provided to carrier 43. For example, a gearbox plate can connect carrier 43 and main shaft 30, or carrier 43 and main shaft 30 can be appropriately connected in other ways. Alternatively, however, ring gear 45 or sun gear 49 can drive planetary gearbox 34.

[0086] Still refer to Figure 3 And as mentioned, the transmission system 28 of the present disclosure may also include an output shaft or generator shaft 29. More specifically, as shown, the generator shaft 29 may be coupled to the gearbox 34 and configured to rotate at the output rotational speed. In certain embodiments, for example, the generator shaft 29 may be a sun gear 49. Thus, the sun gear 49 may engage the planetary gears 47 and may further extend from the planetary gearbox 34 toward the generator 24. In other embodiments, the generator shaft 29 may be coupled to the sun gear 49 or other output gear of the planetary gearbox 34 or other suitable gearbox so that the generator shaft 29 can rotate at the output rotational speed.

[0087] Furthermore, various bearings 39, 70, 72 may surround various rotatable components of the transmission system 28 to facilitate relatively efficient rotation of such rotatable components. For example, as shown, a plurality of carrier bearings 70 may surround the planet gear carrier 43, and a plurality of planet bearings 72 may surround the planet gears 47 and / or additional bearings supporting the sun gear or sun gear shaft (not shown). Such bearings 70, 72 may be roller bearings and include various roller elements arranged in a generally annular array, or may be journal bearings or any other suitable bearings. Furthermore, the bearings 39, 70, 72 described herein may also be referred to as low-speed bearings.

[0088] Now refer to Figure 4 , shows a block diagram of one embodiment of suitable components that may be included within the turbine controller 26 (or a separate controller) in accordance with the present disclosure. As shown, the controller 26 may include one or more processors 56 and associated memory devices 58 that are configured to perform various computer-implemented functions (e.g., perform the methods, steps, calculations, etc. disclosed herein and store related data). Additionally, the controller 26 may also include a communication module 60 to facilitate communication between the controller 26 and the various components of the wind turbine 10. Furthermore, the communication module 60 may include a sensor interface 62 (e.g., one or more analog-to-digital converters) to allow signals transmitted from the sensors 52, 53, 54, 55 to be converted into signals that can be understood and processed by the processor 56. It should be understood that the sensors 52, 53, 54, 55 may be communicatively coupled to the communication module 64 using any suitable means. For example, Figure 4As shown, sensors 52, 53, 54, 55 are coupled to sensor interface 62 via a wired connection. However, in other embodiments, sensors 52, 53, 54, 55 may be coupled to sensor interface 62 via a wireless connection, such as by using any suitable wireless communication protocol known in the art.

[0089] In additional embodiments, the sensors 52, 53, 54, 55 may also be coupled to a separate controller, which may or may not be located in the control cabinet 38. As such, the sensors 52, 53, 54, 55 may provide relevant information to the turbine controller 26 and / or to a separate controller. It should also be understood that, as used herein, the term "monitoring" and variations thereof indicate that the various sensors of the wind turbine 10 may be configured to provide direct measurements of monitored parameters and / or indirect measurements of such parameters. Thus, the sensors 52, 53, 54, 55 described herein may, for example, be used to generate signals related to the monitored parameters, which signals may then be utilized by the controller 26 to determine a condition.

[0090] As used herein, the term "processor" refers not only to what are known in the art as integrated circuits included in computers, but also to controllers, microcontrollers, microcomputers, programmable logic controllers (PLCs), application-specific integrated circuits, and other programmable circuits. Additionally, the memory device(s) 58 may generally include one or more memory elements, including, but not limited to, computer-readable media (e.g., random access memory (RAM)), computer-readable non-volatile media (e.g., flash memory), floppy disks, compact disk read-only memory (CD-ROMs), magneto-optical disks (MODs), digital versatile disks (DVDs), and / or other suitable memory elements. Such memory device(s) 58 may generally be configured to store suitable computer-readable instructions that, when executed by the processor(s) 56, configure the controller 26 to perform various functions, including, but not limited to, transmitting suitable control signals to implement corrective action(s) in response to a distance signal exceeding a predetermined threshold, as described herein, and various other suitable computer-implemented functions.

[0091] Now refer to Figure 5 , the turbine controller 26 (or any separate controller or computing system integral to or remote from the turbine controller 26) is also configured to implement an algorithm for detecting damage in a bearing coupled to a rotating shaft of a rotating machine, such as the wind turbine 10. In general, the method 100 described herein is generally applicable to the wind turbine 10 described above. However, it should be understood that the disclosed method 100 may be implemented using any other suitable rotating machine having a rotating shaft and associated bearings. Furthermore, for purposes of illustration and discussion, Figure 5The steps are depicted as being performed in a specific order. Using the disclosure provided herein, one of ordinary skill in the art will understand that the various steps of any method disclosed herein may be modified, omitted, rearranged, or expanded in various ways without departing from the scope of the present disclosure.

[0092] As shown at (102), the method 100 includes receiving one or more measurement signals from the first sensor(s) 52, 54 for monitoring movement of the main shaft 30 in one or more directions over a period of time. In one embodiment, for example, the directions may include transverse directions relative to the longitudinal direction of the main shaft 30. It should be understood that the method 100 may also include monitoring various environmental and / or operating conditions of the wind turbine 10 via the second sensor(s) 53, 55.

[0093] As shown at (104), the method 100 includes removing the effects of one or more environmental and / or operating conditions of the wind turbine 10 from the one or more measurement signals during the time period. For example, in an embodiment, Figure 6 During the illustrated training period, the controller can automatically and adaptively learn the effects of one or more environmental and / or operating conditions on the movement of the main shaft 30, for example, by monitoring the environmental and / or operating conditions via the second sensors 53, 55. Furthermore, in embodiments, the training period can include any suitable time period, for example, ranging from 1 day to 50 days, such as 28, 29, 30, or 31 days as illustrated. Furthermore, in certain embodiments, the environmental and / or operating conditions can include, for example, wind speed, wind direction, gusts, wind shear, temperature, time of day, air density, generator speed, rotor speed, power output, thrust, and / or torque.

[0094] Therefore, during the subsequent calibration period, the controller is configured to remove the effects of the one or more environmental and / or operating conditions of the wind turbine 10 from the one or more measurement signals by automatically and adaptively eliminating the effects of the one or more environmental and / or operating conditions on the movement of the rotating shaft 30 from the one or more measurement signals (i.e., data from the sensors 53, 55). For example, Figure 6 As shown, during the correction period, the controller may automatically and adaptively remove the effects of one or more environmental and / or operating conditions on the movement of the rotating shaft 30 from the measurement signal(s) (ie, by calculating a corrected measurement signal).

[0095] Special reference Figure 7, a schematic diagram of one embodiment of a training period and a calibration period for a controller according to the present disclosure is shown. During the training period, the controller may assume that the shaft 30 has not moved; however, during the calibration period, movement is detected or expected. More specifically, as shown, the method 100 may include automatically and adaptively learning and eliminating the effects of environmental and / or operating conditions (e.g., input variables 150) on the movement of the main shaft 30 via a machine learning regression model 160. In particular embodiments, the machine learning regression model 160 may utilize at least one of linear regression, nonlinear regression, support vector regression, gradient boosting regression, decision tree regression, random forest regression, generalized linear model, kernel regression, or a neural network. For example, in one embodiment, the machine learning regression model may be represented by the following equation (1):

[0096] Equation (1)

[0097] Where Y is the sensor measurement signal;

[0098] f) variations in the sensor's measurement signal due to environmental and / or operating conditions;

[0099] g is the change in the sensor measurement signal caused by bearing wear or damage;

[0100] X represents factors that contribute to environmental and / or operating conditions;

[0101] Z is bearing wear or damage; and

[0102] e is the random noise in the sensor measurement signal.

[0103] Therefore, to understand bearing wear or damage (ie, Z), the controller can be configured to estimate g(Z) using equation (1), where and are the estimators of f and g respectively:

[0104] Equation (2)

[0105] Using equation (2), the controller can derive equation (3):

[0106] Equation (3)

[0107] in, is the error estimate of f. In addition, as shown in equation (3), is the variation in the measured signal due to main bearing wear plus random noise and errors in the estimate of system variation. Therefore, instead of working with Y (as described in equation (2)), the controller uses .

[0108] Still refer to Figure 7 As shown, the controller is configured to remove the effects of the environmental and / or operating condition(s) of wind turbine 10 from the one or more measurement signals over a time period by determining a predicted measurement signal 154 for the sensor(s) via a regression model 160. Furthermore, as shown at 158, the controller is configured to subtract the predicted measurement signal 154 from the raw measurement signal 152 of the measurement signal(s) using, for example, equation (2) to obtain a corrected measurement signal 156.

[0109] Return Reference Figure 5 , as shown at (106), method 100 includes analyzing a change in the measurement signal(s), wherein the change in the measurement signal(s) exceeding a predetermined threshold or having a certain magnitude indicates bearing damage (e.g., damage to the main bearing 39). In an embodiment, for example, the change or variation in the measurement signal(s) may be a decrease in the measurement signal(s). Thus, in certain embodiments, the controller is configured to analyze the change / variation in the measurement signal(s) by comparing the corrected measurement signal to a predetermined threshold or by determining whether the variation in the corrected measurement signal has a certain magnitude. In the latter example, a high-degree variation in the corrected measurement signal may indicate that the measurement signal is no longer accurately predicted by the operational and environmental signals.

[0110] Thus, as shown at (108), method 100 includes implementing a corrective action when the change / variation in the (one or more) measurement signals exceeds a predetermined threshold or has a certain magnitude. For example, in one embodiment, the corrective action may include any suitable action other than shutting down wind turbine 10, such as, for example, generating an alarm, scheduling maintenance and / or repair procedures.

[0111] refer to Figure 8A and Figure 8B The advantages of the present disclosure may be better understood. For example, as shown, Figure 8A A graph 200 showing raw proximity sensor measurement data (y-axis) versus time (x-axis) is shown, and Figure 8B Graph 300 shows corrected proximity sensor measurement data (y-axis) versus time (x-axis). As shown particularly by circled areas 202 and 302, there is a significant reduction in RMS deviation after eliminating system variations. Thus, as shown by circled areas 204 and 304, the data provides clearer and earlier detection of bearing faults.

[0112] Various aspects and embodiments of the present invention are defined by the following numbered clauses:

[0113] Clause 1. A method for detecting damage in a bearing coupled to a rotating shaft of a rotating machine, the method comprising:

[0114] receiving one or more measurement signals from one or more first sensors for monitoring movement of a rotating shaft in one or more directions over a period of time;

[0115] removing effects of one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals during the time period;

[0116] After the removal, analyzing a change in the one or more measurement signals from the one or more first sensors, wherein the change in the one or more measurement signals exceeds a predetermined threshold or has a magnitude indicative of bearing damage; and

[0117] Corrective action is implemented when a change in one or more measurement signals exceeds a predetermined threshold.

[0118] Clause 2. The method of clause 1, wherein the change in the one or more measurement signals comprises a decrease in the one or more measurement signals.

[0119] Clause 3. The method of any preceding clause, further comprising monitoring one or more environmental and / or operating conditions of the rotating machine via one or more second sensors; and

[0120] The effects of one or more environmental and / or operating conditions on the movement of the rotating shaft are automatically and adaptively learned from the one or more measurement signals.

[0121] Clause 4. The method of clause 3, wherein removing the effects of one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals during the time period further comprises:

[0122] Effects of one or more environmental and / or operating conditions on the movement of the rotating shaft are automatically and adaptively removed from the one or more measurement signals.

[0123] Clause 5. The method of clause 4, further comprising automatically and adaptively learning and eliminating the effects of environmental and / or operating conditions on the movement of the rotating axis via a machine learning regression model.

[0124] Clause 6. The method of clause 5, wherein the machine learning regression model utilizes at least one of linear regression, nonlinear regression, support vector regression, gradient boosting regression, decision tree regression, random forest regression, generalized linear model, kernel regression, or a neural network.

[0125] Clause 7. The method of clause 5, wherein removing the effects of one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals during the time period further comprises:

[0126] determining a predicted measurement signal of the one or more first sensors via a regression model; and

[0127] The predicted measurement signal is subtracted from the original measurement signal of the one or more measurement signals to obtain a corrected measurement signal.

[0128] Clause 8. The method of clause 7, wherein analyzing changes in the one or more measurement signals further comprises comparing the corrected measurement signal to a predetermined threshold or determining whether the variation in the corrected measurement signal has a certain magnitude.

[0129] Clause 9. The method of Clause 3, wherein the rotating machine comprises a wind turbine, the rotating shaft comprises a main shaft of the wind turbine, and the bearing comprises a main bearing of the wind turbine.

[0130] Clause 10. The method of clause 9, wherein the environmental and / or operational conditions include at least one of wind speed, wind direction, gusts, wind shear, temperature, time of day, air density, generator speed, rotor speed, power output, thrust, or torque.

[0131] Clause 11. The method of any preceding clause, wherein the one or more first sensors include one or more proximity sensors.

[0132] Clause 12. The method of any preceding clause, wherein the one or more directions include at least a transverse direction relative to a longitudinal direction of the axis of rotation.

[0133] Clause 13. The method of any preceding clause, wherein implementing corrective action further comprises at least one of generating an alarm or scheduling a maintenance and / or repair procedure.

[0134] Clause 14. A system for detecting damage in a main bearing coupled to a main shaft of a wind turbine, the system comprising:

[0135] one or more first sensors for monitoring movement of the spindle in one or more directions;

[0136] one or more second sensors for monitoring one or more environmental and / or operating conditions of the wind turbine; and

[0137] a controller communicatively coupled to the one or more first and second sensors, the controller configured to perform a plurality of operations comprising:

[0138] receiving one or more measurement signals from one or more first sensors over a period of time;

[0139] removing effects of one or more environmental and / or operating conditions of the wind turbine from the one or more measurement signals during the time period;

[0140] analyzing a decrease in one or more measurement signals from the one or more first sensors, wherein a decrease in the one or more measurement signals exceeding a predetermined threshold or having a magnitude indicates damage to the main bearing; and

[0141] Corrective action is implemented when a decrease in one or more measurement signals exceeds a predetermined threshold.

[0142] Clause 15. The system of clause 14, wherein the plurality of operations further comprises automatically and adaptively learning an effect of one or more environmental and / or operating conditions on movement of the spindle from the one or more measurement signals.

[0143] Clause 16. The system of clause 15, wherein the environmental and / or operational conditions include at least one of wind speed, wind direction, gusts, wind shear, temperature, air density, generator speed, rotor speed, power output, or torque.

[0144] Clause 17. The system of clause 15, wherein removing the effects of one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals during the time period further comprises:

[0145] The effects of environmental and / or operating conditions on the movement of a rotating axis are automatically and adaptively learned and eliminated via a machine learning regression model.

[0146] Clause 18. The system of clause 17, wherein removing the effects of one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals during the time period further comprises:

[0147] determining a predicted measurement signal of the one or more first sensors via a regression model; and

[0148] The predicted measurement signal is subtracted from the original measurement signal of the one or more measurement signals to obtain a corrected measurement signal.

[0149] Clause 19. The system of clause 18, wherein analyzing the decrease in the one or more measurement signals from the one or more first sensors further comprises comparing the corrected measurement signal to a predetermined threshold or determining whether the variation in the corrected measurement signal has a certain magnitude.

[0150] Clause 20. A method for detecting damage in a bearing coupled to a rotating shaft of a rotating machine, the method comprising:

[0151] receiving one or more measurement signals from one or more first sensors for monitoring movement of a rotating shaft;

[0152] automatically and adaptively learning, via the machine learning regression model, during a training period, an effect of one or more environmental and / or operating conditions on the movement of the rotating axis;

[0153] automatically and adaptively canceling the effects of one or more environmental and / or operating conditions on the movement of the rotating shaft during the calibration period; and

[0154] analyzing a decrease in the one or more measurement signals after eliminating the effects of one or more environmental and / or operating conditions; and

[0155] Corrective action is implemented when a decrease in one or more measurement signals from one or more sensors exceeds a predetermined threshold or has a certain magnitude.

[0156] This written description uses examples to disclose the invention, including the best mode, and also to enable any person skilled in the art to practice the invention, including making and using any devices or systems and performing any combined methods. The patentable scope of the invention is defined by the claims and may include other examples that occur to those skilled in the art. These other examples are intended to be within the scope of the claims if they include structural elements that do not differ from the literal language of the claims, or if they include equivalent structural elements with insubstantial differences from the literal language of the claims.

Claims

1. A method for detecting damage in a bearing coupled to a rotating shaft of a rotating machine, the method comprising: receiving one or more measurement signals from one or more first sensors for monitoring movement of the rotating shaft in one or more directions over a period of time; Removing the effects of one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals during the time period, removing the effects of the one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals during the time period further comprising: determining a predicted measurement signal of the one or more first sensors via a regression model; and subtracting the predicted measurement signal from an original measurement signal of the one or more measurement signals to obtain a corrected measurement signal; analyzing, after the removal, a change in the one or more measurement signals from the one or more first sensors, wherein a change in the one or more measurement signals exceeding a predetermined threshold or having a magnitude indicates bearing damage; and Corrective action is implemented when the change in the one or more measurement signals exceeds the predetermined threshold.

2. The method according to claim 1, characterized in that The change in the one or more measurement signals comprises a decrease in the one or more measurement signals.

3. The method according to claim 1, characterized in that Also comprising monitoring the one or more environmental and / or operating conditions of the rotating machine via one or more second sensors; and The effect of the one or more environmental and / or operating conditions on the movement of the rotating shaft is automatically and adaptively learned from the one or more measurement signals.

4. The method according to claim 3, characterized in that Removing the effects of the one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals during the time period further comprises: The effects of the one or more environmental and / or operating conditions on the movement of the rotating shaft are automatically and adaptively removed from the one or more measurement signals.

5. The method according to claim 4, characterized in that Also included is automatically and adaptively learning and eliminating the effects of the environmental and / or operating conditions on the movement of the rotating shaft via a machine learning regression model.

6. The method according to claim 5, characterized in that The machine learning regression model utilizes at least one of linear regression, nonlinear regression, support vector regression, gradient boosting regression, decision tree regression, random forest regression, generalized linear model, kernel regression or neural network.

7. The method according to claim 1, characterized in that Analyzing the changes in the one or more measurement signals further comprises comparing the corrected measurement signal with the predetermined threshold value or determining whether the variation in the corrected measurement signal has the certain magnitude.

8. The method according to claim 3, characterized in that The rotating machine comprises a wind turbine, the rotating shaft comprises a main shaft of the wind turbine, and the bearing comprises a main bearing of the wind turbine.

9. The method according to claim 8, characterized in that The environmental and / or operating conditions include at least one of wind speed, wind direction, gusts, wind shear, temperature, time of day, air density, generator speed, rotor speed, power output, thrust, or torque.

10. The method according to claim 1, characterized in that The one or more first sensors include one or more proximity sensors.

11. The method according to claim 1, wherein The one or more directions include at least a transverse direction relative to the longitudinal direction of the rotation axis.

12. The method according to claim 1, characterized in that Implementing the corrective action also includes at least one of generating an alarm or scheduling a maintenance and / or repair procedure.

13. A system for detecting damage in a main bearing coupled to a main shaft of a wind turbine, the system comprising: one or more first sensors for monitoring movement of the spindle in one or more directions; one or more second sensors for monitoring one or more environmental and / or operating conditions of the wind turbine; and a controller communicatively coupled to the one or more first sensors and the one or more second sensors, the controller configured to perform a plurality of operations comprising: receiving one or more measurement signals from the one or more first sensors over a period of time; Removing the influence of one or more environmental and / or operating conditions of the wind turbine from the one or more measurement signals during the time period, removing the influence of the one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals during the time period further comprises: determining a predicted measurement signal of the one or more first sensors via a regression model; and subtracting the predicted measurement signal from an original measurement signal of the one or more measurement signals to obtain a corrected measurement signal; analyzing a decrease in the one or more measurement signals from the one or more first sensors, wherein a decrease in the one or more measurement signals exceeding a predetermined threshold or having a magnitude indicates main bearing damage; and Corrective action is implemented when the decrease in the one or more measurement signals exceeds the predetermined threshold.

14. The system according to claim 13, wherein: The plurality of operations further includes automatically and adaptively learning from the one or more measurement signals an effect of the one or more environmental and / or operating conditions on the movement of the spindle.

15. The system according to claim 14, wherein: The environmental and / or operating conditions include at least one of wind speed, wind direction, gusts, wind shear, temperature, air density, generator speed, rotor speed, power output, or torque.

16. The system according to claim 14, wherein: Removing the effects of the one or more environmental and / or operating conditions of the rotating machine from the one or more measurement signals during the time period further comprises: The effects of the environmental and / or operating conditions on the movement of the rotating shaft are automatically and adaptively learned and eliminated via a machine learning regression model.

17. The system according to claim 13, wherein: Analyzing the decrease in the one or more measurement signals from the one or more first sensors further comprises comparing the corrected measurement signal with the predetermined threshold value or determining whether a variation in the corrected measurement signal has the certain magnitude.

Images