SYSTEMS AND METHODS FOR MONITORING A WIND TURBINE

BACKGROUND

[0001] The invention relates generally to a condition monitoring system and more specifically to a system for monitoring a condition of a wind turbine.

[0002] In the last few decades, the field of monitoring and diagnostics of mechanical systems and electrical systems has grown tremendously. Also, monitoring the condition of mechanical systems and electrical systems aids in the early detection of failures and enhanced planning of maintenance and repair cycles, thereby extending the life of the mechanical and electrical systems.

[0003] Conventional monitoring techniques employ additional sensors for identifying faulty conditions. Currently, faulty conditions in a wind turbine are monitored and diagnosed by monitoring systems that employ additional physical sensors such as a torque sensor, a visual sensor, a vibration sensor, a fiber optic sensor, and the like. Furthermore, the visual sensors such as unmanned aerial vehicles (UAVs), telescopes, and the like, are employed to inspect the wind turbines. In particular, visual images of different wind turbine components such as blades and nacelle are acquired and analyzed to monitor the wind turbine. While the visual sensor can effectively detect cracks that appear on the surface of the wind turbine components, the visual sensor fails to effectively determine internal cracks. Also, the use of torque sensors or other equivalent sensors for monitoring wind turbine entails analyzing a mechanical torque to identify the faulty condition in the wind turbine. However, while the mechanical torque can be efficiently obtained in steady state operating condition, it is difficult to efficiently acquire die mechanical torque in a transient operating condition. Consequently, equivalent parameters like electromagnetic torque are employed for identification of faults.

[0004] Moreover, use of vibration sensors and fiber optic sensors aid in identifying faulty conditions of the wind turbine at an advanced stage, but fail to identify the faulty conditions at an incipient stage. Also, use of additional sensors may increase the complexity and die cost of the monitoring systems.

BRIEF DESCRIPTION

[0005] In accordance with aspects of the present technique a method for monitoring a wind turbine is presented. The method includes acquiring one or more line parameters at an output terminal of a generator, where the generator is operatively coupled to the wind turbine. Further, the method includes estimating a derived parameter based on the one or more acquired line parameters. In addition, the method includes determining a frequency spectrum of the estimated derived parameter. The method further includes determining a magnitude of the determined frequency spectrum at a tower pass frequency, and identifying a condition of the wind turbine based on a comparison of the magnitude of the determined frequency spectrum at the tower pass frequency with a threshold value.

[0006] In accordance with another aspect of the present technique, a non-transitory computer readable medium including one or more tangible media is presented. The one or more tangible media include code adapted to perform the method for monitoring a wind turbine.

[0007] In accordance with yet another aspect of the present technique, a device is presented. The device includes a processing unit configured to acquire one or more line parameters at an output terminal of a generator operatively coupled to a wind turbine. The processing unit includes a derived parameter estimating unit configured to estimate a derived parameter based on the one or more acquired line parameters. Further, the processing unit includes a mathematical modeling unit configured to identify a non-stationary sub-signal corresponding to the derived parameter. Moreover, the processing unit includes a signal processing unit configured to process the derived parameter to minimize an error introduced in the derived parameter due to the non-stationary sub-signal. In addition, the processing unit includes an adaptive learning unit configured to adaptively learn at least one of a trend of progress of faults and a correlation of one or more of the one or more acquired line parameters, the derived parameter, a magnitude of a determined frequency spectrum at a tower pass frequency, an identified condition of the wind turbine, a recommendation for the wind turbine, or combinations thereof.

DRAWINGS

[0008] These and other features, aspects, and advantages of the present disclosure will become better understood when the following detailed description is read with reference to the accompanying drawings in which like characters represent like parts throughout the drawings, wherein:

[0009] FIG. 1 is a diagrammatical representation of an exemplary system for monitoring a condition of a wind turbine, according to aspects of the present disclosure;

[0010] FIG. 2 is a diagrammatical representation of an exemplary processing unit for use in the exemplary system of FIG. 1, according to aspects of the present disclosure;

[0011] FIG. 3 is a flow chart representing an exemplary method for monitoring a condition of a wind turbine, according to aspects of the present disclosure;

[0012] FIG. 4 is a flow chart representing a step of determining presence of non-stationary sub-signals corresponding to a derived parameter of the method of FIG. 3, according to aspects of the present disclosure; and

[0013] FIG. 5 is a diagrammatical representation of die identification of a faulty condition of a wind turbine, according to aspects of the present disclosure.

DETAILED DESCRIPTION

[0014] Unless defined otherwise, technical and scientific terms used herein have the same meaning as is commonly understood by one of ordinary skill in the art to which this disclosure belongs. The terms "first", "second", and the like, as used herein do not denote any order, quantity, or importance, but rather are used to distinguish one element from another. Also, the terms "a" and "an" do not denote a limitation of quantity, but rather denote the presence of at least one of the referenced items. The term "or" is meant to be inclusive and mean one, some, or all of the listed items. The use of "including," "comprising" or "having" and variations thereof herein are meant to encompass the items listed thereafter and equivalents thereof as well as additional items. The terms "connected" and "coupled" are not restricted to physical or mechanical connections or couplings, and can include electrical connections or couplings, whether direct or indirect. Furthermore, terms "circuit" and "circuitry" and "controller" may include either a single component or a plurality of components, which are either active and/or passive and are connected or otherwise coupled together to provide the described function.

[0015] As will be described in detail hereinafter, various embodiments of an exemplary system and method for monitoring a condition of the wind turbine are presented. By employing the system and method for monitoring the condition of the wind turbine described hereinafter, the condition of the wind turbine can be monitored at an incipient stage of the fault.

[0016] Embodiments disclosed herein relate generally to a system and method for monitoring a condition of the wind turbine. In one example, the system may include a wind turbine, a generator, a processing unit, a controller, a display unit, and an alarm generation unit. The wind turbine may include a tower, a nacelle, a blade, a hub, a wind turbine rotor, and other equivalent components. Moreover, the wind turbine may also include auxiliary components such as a drive-train, a shaft, mechanical coupling, a bearing, a gearbox, a gear, and the like.

[0017] Turning now to the drawings, by way of example in FIG. 1, a diagrammatical representation of an exemplary system 100 for monitoring a condition of a wind turbine, according to aspects of the present disclosure, is depicted. FIG. 1 illustrates a wind turbine 102 operatively coupled to a generator 104. The term generator as used herein may include an electrical generator, an electrical alternator, a synchronous generator a synchronous singly fed generator, an induction singly fed generator, a doubly fed generator, a brushless wound-rotor doubly fed generator, a magneto hydrodynamic generator, and the like.

[0018] In accordance with further aspects of the present disclosure, the monitoring system 100 may include a processing unit 106, a controller 108, a display unit 110, and an alarm generation unit 112. In one embodiment, the display unit 110 and the alarm generation unit 112 may be packaged as a single unit. As noted hereinabove, the wind turbine 102 may be operatively coupled to the generator 104. In turn, the generator 104 may be operatively coupled to the processing unit 106, while the processing unit 106 may be operatively coupled to me controller 108. Furthermore, the term operatively coupled as used herein includes wired coupling, wireless coupling, electrical coupling, magnetic coupling, radio communication, software based communication, or combinations thereof.

[0019] The controller 108 may be configured to regulate the operation of the wind turbine 102. In one non-limiting example, the controller 108 may be configured co control a blade pitch, the torque, and the like, of the wind turbine 102. Moreover, the processing unit 106 may also be operatively coupled to a display unit 110. The display unit 110, as used herein, may include a computer monitor, a light emitting diode, a liquid crystal display, or equivalents thereof. Also, the display unit 110 may be configured to display any findings generated by the processing unit 106. In one embodiment, the display unit 110 may be disposed at a location mat is remote from the system 100. However, in another embodiment, the display unit 110 may be located in the vicinity of the wind turbine 102. The monitoring system 100 may also include an alarm generation unit 112 operatively coupled to the processing unit 106. The alarm generation unit 112 may be configured to transmit alarm signals when a faulty condition is identified, in one example.

[0020] Additionally, as will be described in greater detail with respect to FIG. 2, in one embodiment, the processing unit 106 may include an adaptive learning unit, an electromagnetic parameter estimating unit, a mechanical parameter estimating unit, a mathematical modeling unit, and a signal processing unit. In one example, the processing unit 106 may include a computer, a processor, and equivalents thereof.

[0021] In a presently contemplated configuration, me processing unit 106 may be configured to acquire one or more line parameters from an output terminal of the generator 104. The one or more line parameters may include a voltage (V), a current (I), or a combination thereof. In accordance with one embodiment, the processing unit 106 may acquire the one or more line parameters at an output terminal of the generator 104 via a sensing device (not shown). The sensing device may include a voltage sensing device, a current sensing device, a power sensing device, and equivalents thereof. In one embodiment, the processing unit 106 may employ existing sensing devices in the wind turbine 102 that are used to monitor the generator parameters, thereby circumventing the need for any additional sensors. However, in another embodiment, the sensing device may be packaged in the processing unit 106. Alternatively, the sensing device may be packaged as a separate unit. Furthermore, me processing unit 106 may also be configured to acquire a wind turbine rotor rotational speed (co) and a moment of inertia (J) from the wind turbine 102.

[0022] In addition, the processing unit 106 may be configured to estimate a derived parameter based on the one or more acquired line parameters. The term derived
parameter as used herein may be used to refer to parameters that are derived using the acquired line parameters. By way of example, the derived parameter may include a mechanical torque, a mechanical power, an electromagnetic power, an electromagnetic torque, and equivalents thereof.

[0023] Moreover, in accordance with aspects of the present disclosure, the processing unit 106 may be configured to process the derived parameter to determine presence of one or more non-stationary sub-signals corresponding to the derived parameter. The non-stationary sub-signal corresponding to the derived parameter may be representative of an undesirable component which may result in erroneous identification of the condition of wind turbine. Hence, it may be desirable to minimize the error introduced in the derived parameter due to the one or more non-stationary sub-signal. To that end, the processing unit 106 may be configured to process the derived parameter to minimize the error introduced in the derived parameter due to the non-stationary sub-signal.

[0024] Subsequent to minimizing the error introduced in the derived parameter, a frequency spectrum of the derived parameter may be determined. In one example a fast Fourier transform (FFT) may be employed to determine the frequency spectrum of the derived parameter. However, use of other spectral analysis techniques is also envisaged, in a non-limiting example. However, if it is determined that the non-stationary sub-signal corresponding to the derived parameter is absent, the frequency spectrum of the derived parameter may be determined immediately after thi derived pjjameter is obtained from the acquired line parameters.

[0025] As will be appreciated, when a turbine blade passes in front of the wind turbine tower, distribution of wind may be altered due to the tower. This change in the distribution of wind results in a fluctuation in one or more line parameters. For example, when the turbine blade passes in front of the tower the wind in front of the tower is redirected, thereby reducing the torque at each turbine blade when the turbine blade is disposed in front of the tower. This results in a pulsation in one or more line parameters, a wind turbine rotor rotational speed, and/or the derived parameter. As previously noted, the derived parameter includes the electromagnetic power, the electromagnetic torque, the mechanical torque, the mechanical power, and equivalents thereof. This occurrence of pulsation is generally referred to as a tower shadow effect. In another example, the
pulsation may be due to a wind shear and/or an inconsistency in the assembly and/or manufacture of blades. Also, the frequency at which the pulsation is reflected in th.2 frequency spectrum of the one or more line parameters, the derived parameters, and the wind turbine rotor rotational speed is referred to as a tower pass frequency. The term tower pass frequency as used herein may include a fundamental component and corresponding harmonics.

[0026] In accordance with aspects of the present disclosure, a tower pass frequency component in die frequency spectrum of the derived parameter may be identified and analyzed to identify the condition of the wind turbine 102. The identified condition of wind turbine 102, as used herein, may include a healthy condition of the wind turbine 102, a faulty condition of die wind turbine 102, and the like. In one embodiment, the faulty condition may include a fault in one or more blades of wind turbine 102. By way of example, the fault in one or more blades of wind turbine 102 may include an internal crack in a blade, an external crack in a blade, a blade delamination, a blade core breakdown, a blade icing condition, and the like. As noted hereinabove, the tower pass frequency of the frequency spectrum may be analyzed to determine the condition of the wind turbine. In one embodiment, a magnitude of the frequency spectrum at die tower pass frequency may be determined. In one example, the magnitude of me frequency spectrum may be normalized with respect to the magnitude of die fundamental component of the tower pass frequency. Also, the magnitude of the frequency spectrum at the tower pass frequency may then be compared wifli a threshold value to identify a condition of the wind turbine 102.

[0027] In one non-limiting example, when a crack appears on the wind turbine blade, the stiffness and the natural frequency of die wind turbine blade changes accordingly. As noted hereinabove, when die wind turbine blade passes in front of the tower, the tower shadow effect leads to a pulsation in die derived parameter and wind turbine rotor rotational speed at a fundamental frequency. Moreover, during a faulty condition, the amplitude of the pulsation in one of the blades will be different from diat of the otiier blades. Accordingly, die magnitude of the frequency spectrum at die tower pass frequency may be compared witii me threshold value to identify a faulty condition of a wind turbine blade. In one example, if the magnitude of the frequency spectrum at the
tower pass frequency is greater than the threshold value, this may be indicative of a faulty condition.

[0028] In accordance with further aspects of the present disclosure, a recommendation for the wind turbine 102 may be modified based on the identified faulty condition. In particular, the identified condition of the wind turbine 102 may be employed to determine the recommendation for the wind turbine 102. The recommendation for the wind turbine 102 may include operating the wind turbine 102 in a partial capacity mode, predicting life of the wind turbine 102, estimating maintenance and repair cycle of the wind turbine 102, correcting a pitch demand from a controller, and the like. The term partial capacity mode as used herein may include operating the wild turbine 102 in a curtailed mode. In one non-limiting example, the curtailed mode may include curtailment of rotor rated speed and/or rotor rated power. Steps for identification of the condition of the wind turbine in the processing unit 106 will be explained in greater detail with reference to FIGs. 3-5.

[0029] Additionally, the processing unit 106 may be configured to regulate the operation of the controller 108. Specifically, the processing unit 106 may be configured to regulate the operation of the controller 108 during a measurement stage. The term measurement stage as used herein is used to refer to the acquisition of die one or more line parameters at die output terminal of the generator 104 and the determination of the recommendation for the wind turbine 102 and any intermediate steps. By way of example, the processing unit 106 may be configured to regulate the operation of the controller 108 to operate the wind turbine 102 in a partial capacity mode upon determination of the faulty condition of the wind turbine 102.

[0030] Typically, in conventional systems a controller is configured to suppress die tower pass frequency, since the tower pass frequency is undesirable. However, in accordance with aspects of the present disclosure, the tower pass frequency is analyzed to identify the condition of the wind turbine 102. Accordingly, the controller 108 may be configured to not suppress the tower pass frequency component during the measurement stage. This tower pass frequency component of the derived parameter may be employed to aid in the effective identification of die condition of the wind turbine 102.

[0031] Referring now to FIG. 2, a diagrammatical representation 200 of an exemplary processing unit such as the processing unit 106 for use in the system 100 of FIG. 1 is depicted. In the presently contemplated configuration, the processing unit 200 may include an electromagnetic parameter estimating unit 202, a mechanical parameter estimating unit 204, a mathematical modeling unit 206, a signal processing unit 208, and an adaptive learning unit 210. The processing unit 200 may include a data repository 212. In the embodiment of FIG. 2, the data repository 212 may be included in the adaptive learning unit 210. As previously noted, the processing unit 200 is configured to acquire one or more line parameters from an output terminal of a generator such as the generator 104 of FIG. 1. Also, the processing unit 200 may acquire a wind turbine rotor rotational speed (co) and a moment of inertia (J) of a wind turbine such as the wind turbine 102 of FIG. 1.

[0032] The electromagnetic parameter estimating unit 202 may be configured to determine an electromagnetic parameter (Tem) based on the one or more acquired line parameters. In one non-limiting example, the electromagnetic parameter may include an electromagnetic power, an electromagnetic torque, and the like. Furthermore, in the example of FIG. 2, the electromagnetic parameter may be determined based on a voltage (V), a current (I), and a wind turbine rotor rotational speed (co). In addition, in one embodiment, the processing unit 200 may be configured to acquire an electrical flux (\|/) of the generator which may be employed to determine the electromagnetic parameter.

[0033] In addition, the mechanical parameter estimating unit 204 may be configured to determine a mechanical parameter (T,„) based on corresponding electromagnetic parameter (Ten,) determined by the electromagnetic parameter estimating unit 202. The mechanical parameter (T,„) as used herein may include a mechanical power, a mechanical torque, and the like. In one non-limiting example, in addition to the electromagnetic parameter (Ten,), the moment of inertia (J) and a coupling constant (k) may also be employed to determine the mechanical parameter. The coupling constant as used herein is representative of an effect of a mechanical coupling of the wind turbine 102 to the generator 104. Also, in one embodiment, the mechanical parameter estimating unit 204 may include mechanical simulation models for estimating the mechanical parameters. The electromagnetic parameter estimating unit 202 and the mechanical parameter estimating unit 204 in combination may form a derived parameter estimating unit 214.

[0034] The mathematical modeling unit 206 may be configured to identify presence of a non-stationary sub-signal corresponding to the derived parameter. It may be noted that the non-stationary sub-signal corresponding to the derived parameter may be representative of a signal that has varying amplitude and/or frequency over a determined time interval/period. The non-stationary sub-signals corresponding to the derived parameter may be generated due to a blade pitch mechanism, a behavior of controller (such as controller 108 of FIG. 1), a variation in wind speed, a variation in wind turbine rotor rotational speed, or equivalents thereof.

[0035] The signal processing unit 208 may be configured to determine the tower pass frequency from the frequency spectrum of the derived parameter. Moreover, the signal processing unit 208 may also be configured to determine a magnitude of the frequency spectrum at the tower pass frequency. In one example, the magnitude of the frequency spectrum at the tower pass frequency may be normalized with respect to the magnitude of the fundamental component of tower pass frequency. Additionally, in one embodiment, the signal processing unit 208 may be configured to process the derived parameter in order to minimize me error introduced in the derived parameter due to a non-stationary sub-signal. In one embodiment, the error introduced in the derived parameter due to the non-stationary sub-signal may be minimized using a non-stationary signal processing technique. By way of example, the error introduced in the derived parameter due to the non-stationary sub-signal may be minimized using the non-stationary signal processing technique such as a windowed spectral analysis. The windowed spectral analysis of the derived parameter may include partitioning the derived parameter into a plurality of windows. Subsequently, each of the plurality of windows is analyzed employing a spectral analysis technique. The error introduced by the non-stationary sub-signal is minimal in the individual windows of the plurality of windows of the derived parameter, when compared to the error introduced in the derived parameter as a whole. Thus, the spectral analysis of each individual window aids in minimizing the error introduced in the derived parameter due to a non-stationary sub-signal. Moreover, other non-stationary signal processing technique such as, but not limited to, a wavelet transform technique, an empirical mode decomposition (EMD) technique, or combinations thereof may also be employed to minimize error introduced in the derived parameter due to the non-stationary sub-signal. In other embodiment, a moving average filter may be employed as a non-stationary signal processing technique.

[0036] Also, the adaptive learning unit 210 may be configured to 'learn' from the one or more acquired line parameters, the derived parameter, the determined frequency spectrum of die derived parameter, the magnitude of the determined frequency spectrum at die tower pass frequency the identified conditions of the wind turbine, and/or the recommendations for the wind turbine identified in previous measurement stages.

[0037] Also, the one or more acquired line parameters, the derived parameter, the determined frequency spectrum of die derived parameter, die magnitude of die determined frequency spectrum at the tower pass frequency, the identified conditions of die wind turbine, and die recommendations for die wind turbine identified in die previous measurement stages may be stored in die adaptive learning unit 210 and particularly, in die data repository 212. The stored one or more acquired line parameters, die derived parameter, die determined frequency spectrum of die derived parameter, die magnitude of me determined frequency spectrum at die tower pass frequency obtained in previous measurement stages may generally be referred to as historical data. In die example of FIG. 2, die adaptive learning unit 210 and die data repository 212, in particular, may store a voltage (V), a current (I), a rotational speed (o), a moment of inertia (J), an electromagnetic parameter (Tem), and a mechanical parameter (Tm).

[0038] The term learn as used herein is used to refer to die processing and/or analyzing of die historical data to correlate die historical data with an identified condition of wind turbine and/or recommendation for die wind turbine. Subsequently, this correlation may be employed to identify die condition of die wind turbine corresponding to subsequent measurements. In a non-limiting example, any correlation between die one or more acquired line parameters, die derived parameter, die determined frequency spectrum of die derived parameter, die magnitude of die determined frequency spectrum at die tower pass frequency, die identified conditions of me wind turbine, and the recommendations for the wind turbine may be learned by the adaptive learning unit 210.

[0039] Moreover, die adaptive learning unit 210 may also be configured to analyze die magnitude of die determined frequency spectrum at die tower pass frequency corresponding to subsequent measurement stages in order tc determine a trend of progress of a fault. The trend of progress of die fault may include a change in deviation of die magnitude of die frequency at die tower pass frequency from die threshold value
over subsequent measurement stages. This study of the trend of progress of the fault may be referred to as a trending of faults.

[0040] In accordance with aspects of the present disclosure, based on correlation of data obtained in the subsequent measurement stage and the historical data stored in the adaptive learning unit 210, the condition of the wind turbine such as an internal crack in a blade or an external crack in a blade can be determined. The use of the adaptive learning unit 210 aids in the easy identification of faulty condition in the wind turbine and in tracking of progress of faults in the wind turbine. Furthermore, in one example, the adaptive learning unit 210 may include an artificial neural network (ANN) based unit, an artificial intelligence (AT) based unit, a physics based learning unit, and equivalents thereof.

[0041] FIG. 3 is a flow chart 300 depicting an exemplary method for monitoring a condition of a wind turbine. Additionally, in one embodiment, monitoring the condition of the wind turbine may be coordinated by the processing unit 106 of FIG. 1 or 200 of FIG. 2. For ease of understanding, the method of FIG. 3 will be described with respect to the components of FIGs. 1-2. The method begins at a step 302 where, one or more line parameters are acquired. As previously noted, the one or more line parameters may include a voltage, a current, and die like. In addition, the line parameters may be acquired at an output terminal of the generator, such as generator 104 of FIG. 1. Also, the line parameters may include a three phase current, a three phase voltage, and the like. The one or more line parameters may be acquired when the generator 104 is operating near its rated power. Furthermore, the one or more line parameters may be acquired in real time.

[0042] Subsequently, at step 304, a derived parameter may be estimated based on the one or more acquired line parameters. The derived parameter may include aa electromagnetic parameter, a mechanical parameter, and the like, as previously noted. In addition, step 304 may be performed by the processing unit 200. Also, in one embodiment, if the derived parameter is an electromagnetic parameter, step 304 may be executed by the electromagnetic parameter estimating unit 202 of FIG. 2, based on voltage (V), current (I) and a wind turbine rotor rotational speed (co). However, if the derived parameter is a mechanical parameter, step 304 may be executed in rne mechanical parameter estimating unit 204 of FIG. 2, based on the electromagnetic torque (Ten,) and the moment of inertia (J).

[0043] At step 306, the derived parameter may be processed to determine a presence of a non-stationary sub-signal. To that end, a check may be carried out to determine the presence of any non-stationary sub-signal corresponding to the derived parameter. As previously noted, non-stationary sub-signal corresponding to the derived parameter is representative of a signal that has varying amplitude and/or frequency over a determined period of time. The processing unit 200 and the mathematical modeling unit 206, in particular, may be employed to verify the presence of the non-stationary sub-signal corresponding to the derived parameter. Step 306 will be explained in greater detail with reference to FIG. 4.

[0044] Furthermore at step 306, if it is determined that there exists a non-stationary sub-signal corresponding to the derived parameter, the derived parameter may be processed to minimize the error introduced in the derived parameter due to non-stationary sub-signal as indicated by step 308. The error introduced in the derived parameter due to the non-stationary sub-signal may be minimized by using a non-stationary signal processing technique. In one example, non-stationary signal processing techniques, such as, but not limited to a wavelet transform technique, an empirical mode decomposition (EMD) technique, a windowed spectral analysis technique, or combinations thereof may be used to process the derived parameter. Control may then be passed on to step 310.

[0045] With returning reference to step 306, if it is verified that the derived parameter does not have any non-stationary sub-signal, a frequency spectrum of the estimated derived parameter may be determined, as indicated by step 310. At step 310, the frequency spectrum of the derived parameter may be determined using a FFT technique. Steps 308 and 310 may be performed by the processing unit 200 and in particular, by the signal processing unit 208 of FIG. 2.

[0046] Furthermore, at step 312, a magnitude of the frequency spectrum at a tower pass frequency may be determined. As previously noted, the tower pass frequency is the frequency at which the pulsation is reflected in a frequency spectrum of the one or more line parameters, the derived parameter, and/or the wind turbine rotor rotational speed due to the tower shadow effect, the wind shear, and the like. At step 314, the magnitude of
the frequency spectrum at the tower pass frequency may be normalized with respect to the magnitude of the fundamental component of the tower pass frequency. For example, the step 312 and 314 may be performed by the signal processing unit 208 of the processing unit 200.

[0047] Also, at step 316, the magnitude of the frequency spectrum at the tower pass frequency may be compared with a threshold value to identify a condition of the wind turbine. The comparison with the threshold value will be explained in greater detail with reference to FIG. 5. In one embodiment, the threshold value may be determined based on a field trial, an experimental simulation, and the like. In another embodiment, the threshold value may be determined based on learnings of the adaptive learning unit 210. Furthermore, the threshold value may be a variable quantity, in one example. Also, the threshold value may be a function of a wind speed, a pitch angle, and/or a wind turbine rotor speed. In one example, threshold value may be indicative of a healthy condition of the wind turbine. Steps 316 may be performed in the processing unit 200. In one embodiment, the step 316 may be performed by the signal processing unit 208, in particular.

[0048] Moreover, at step 318, a condition of the wind turbine may be identified based on the comparison of magnitude of the frequency spectrum at the tower pass frequency with die threshold value. As previously noted, the identified condition of wind turbine may include a healthy condition of the wind turbine, a faulty condition of the wind turbine, and the like. Step 318 may be performed by the processing unit 200.

[0049] Also, based on the identified condition, a recommendation for the wind turbine may be determined as indicated by step 320. The recommendation for the wind turbine may include operating the wind turbine in a partial capacity mode, predicting life of the wind turbine, estimating a maintenance cycle of the wind turbine, correcting a pitch demand from a controller, and the like. The, step 320 may be performed by the processing unit 200.

[0050] In accordance with aspects of the present disclosure, the exemplary method for monitoring the condition of the wind turbine may also be used to identify a faulty turbine blade. To that end, in one embodiment, phase information from the frequency spectrum of the derived parameter may be employed. However, in another embodiment, the identification of faulty turbine blade may entail use of positional information of the turbine blades. The existence of faults in different turbine blades may result in different harmonic patterns in the frequency spectrum of the derived parameter. Furthermore, the existence of different types of faults, such as, but not limited to, a crack in a blade, a blade icing condition, a blade delamination, a blade core breakdown in a turbine blade and the like, may result in different harmonic patterns in the frequency spectrum of the derived parameter. Hence, based on the different harmonic patterns in the frequency spectrum of the derived parameter the type of fault may be determined, in one example. Also, it may be noted that the frequency spectrum of the derived parameter may have relatively fewer harmonic components when one turbine blade is faulty than when multiple turbine blades of the wind turbine are in a faulty condition. Accordingly, when multiple turbine blades are in a faulty condition, the frequency spectrum of the derived parameter may be different from a frequency spectrum of the derived parameter when only a single turbine blade is faulty. In one non-limiting example, the position of the one or more turbine blades may change the frequency spectrum of the derived parameter. Consequently, based on the position of the one or more turbine blades, a particular turbine blade having a faulty condition may be identified.

[0051] Turning now to FIG. 4, a flow chart 400 representing a step for determining presence of a non-stationary sub-signal corresponding to a derived parameter is depicted, by way of example. In particular, step 306 of FIG. 3 is presented in FIG. 4. In one example, step 306 for determining the presence of a non-stationary sub-signal corresponding to the derived parameter may be coordinated by the processing unit 106 of FIG. 1 and in particular, by the signal processing unit 208 of FIG. 2. The method begins at 402 by obtaining the derived parameter. As previously noted, the derived parameter maybe generated subsequent to the processing of step 304.

[0052] Furthermore, at step 404, a recursive mean of the derived parameter obtained at different instants of time in a measurement stage may be determined. Also, a covariance of the mean of the derived parameter may be determined, as depicted by step 406. In addition, at step 408, a check may be carried out to compare the covariance of the mean of the derived parameter with a threshold covariance value (othieshow)- At step 408, if it is determined that the covariance of the mean of the derived parameter is greater than or equal to the threshold covariance value, presence of the non-stationary sub-signal
corresponding to the derived parameter may be determined as depicted by reference numeral 410. However, at step 408, if the covariance of the mean of the derived parameter is less than the threshold covariance value that may be indicative of the absence of the non-stationary sub-signal corresponding to the derived parameter, as depicted by reference numeral 412.

[0053] Referring to FIG. 5, a diagrammatical representation 500 of the identification of a faulty condition of a wind turbine is depicted. In particular, step 316 of FIG. 3 is depicted in FIG. 5. By way of example, in FIG. 5, a frequency spectrum 502 corresponding to a first condition of the derived parameter and a second condition of the derived parameter, at a tower pass frequency is represented. Reference numeral 504 generally represents a portion of the frequency spectrum 502. Also, reference numeral 514 is representative of a magnitude of the frequency spectrum while a frequency is represented by reference numeral 516. Furthermore, the portion 504 of the frequency spectrum may include a first frequency spectrum 506 corresponding to the first condition of the derived parameter and a second frequency spectrum 508 corresponding to the second condition of the derived parameter.

[0054] In one non-limiting example, at a frequency (ft) 510, the first frequency spectrum 506 of the derived parameter deviates from the second frequency spectrum 508 of the derived parameter. Reference numeral 512 is generally indicative of the deviation of the first frequency spectrum 506 from the second frequency spectrum 508. In one embodiment, the first condition may include a normal condition and the second condition may include an anomaly condition. Also, in one example, the anomaly condition may include a faulty condition while the normal condition may include a healthy condium. In one embodiment, the first frequency spectrum 506 of the derived parameter may be employed as a threshold value. As noted hereinabove, the threshold value may be determined based on a field trial, an experimental simulation, and the like. Furthermore, the deviation 512 may be indicative of a faulty condition. Also, the value of deviation may be indicative of the severity of the fault condition of the wind turbine. In one non-limiting example, when the deviation is a relatively smaller value, the deviation may be indicative of a fault of lesser intensity. Furthermore, in one example, based on the value of the deviation 512, the type of fault, such as, but not limited to, blade crack, blade icing condition, or blade delamination may be identified.

[0055] Furthermore, the foregoing examples, demonstrations, and process steps such as those that may be performed by the system may be implemented by suitable code on a processor-based system, such as a general-purpose or special-purpose computer. It should also be noted that different implementations of the present disclosure may perform some or all of the steps described herein in different orders or substantially concurrently, that is, in parallel. Furthermore, the functions may be implemented in a variety of programming languages, including but not limited to C++ or Java. Such code may be stored or adapted for storage on one or more tangible, machine readable media, such as on data repository chips, local or remote hard disks, optical disks (that is, CDs or DVDs), memory or other media, which may be accessed by a processor-based system to execute the stored code. Note that the tangible media may comprise paper or anouier suitable medium upon which the instructions are printed. For instance, the instructions may be electronically captured via optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in the data repository or memory.

[0056] The system and method for monitoring the condition of the wind turbine
described hereinabove aid in the early detection of a faulty condition of the wind turbine, thereby preventing catastrophic failure. Also, the system and method for monitoring the condition of the wind turbine employ sensing devices currently existing in the wind turbine, circumventing the need for any additional sensors, thereby reducing system complexity and cost. Furthermore, the use of the line parameters, die electromagnetic parameters, and the parameters estimated from the electromagnetic parameters allow efficient identification of faulty condition even during a transient operating condition.

[0057] While the invention has been described with reference to exempla'y embodiments, it will be understood by those skilled in die art mat various changes may be made and equivalents may be substituted for elements thereof without departing from the scope of the invention. In addition, many modifications may be made to adapt a particular situation or material to the teachings of the invention without departing from the essential scope thereof. Therefore, it is intended that die invention not be limited to the particular embodiment disclosed as me best mode contemplated for carrying out mis invention, but that me invention will include all embodiments falling within the scope of the appended claims.

CLAIMS:

1. A method for monitoring a wind turbine, comprising:

acquiring one or more line parameters at an output terminal of a generator, wherein the generator is operatively coupled to the wind turbine;

estimating a derived parameter based on the one or more acquired line parameters;

determining a frequency spectrum of the estimated derived parameter;

determining a magnitude of the determined frequency spectrum at a tower pass frequency; and

identifying a condition of the wind turbine based on a comparison of the magnitude of the determined frequency spectrum at the tower pass frequency with a threshold value.

2. The method of claim 1, further comprising identifying a recommendation for the wind turbine based on the identified condition of the wind turbine.

3. The method of claim 2, wherein the recommendation for the wind turbine comprises operating the wind turbine in a partial capacity mode, predicting life of the wind turbine, estimating a maintenance cycle of the wind turbine, correcting a pitch demand from a controller, or combinations thereof.

4. The method of claim 1, further comprising:

identifying a non-stationary sub-signal corresponding to the derived parameter; and

processing the derived parameter to minimize an error introduced by the non-stationary sub-signal.

5. The method of claim 4, wherein processing the derived parameter comprises
using a non-stationary signal processing technique.

6. The method of claim 5, wherein the non-stationary signal processing technique comprises a wavelet transform technique, an empirical mode decomposition technique, a windowed spectral analysis technique, or combinations thereof.

7. The method of claim 1, wherein determining the frequency spectrum of the derived parameter comprises processing the derived parameter using a fast Fourier transformation technique.

8. The method of claim 1, further comprising adaptively learning at least one of a trend of progress of faults and a correlation of one or more of the one or more acquired line parameters, the derived parameter, the magnitude of the determined frequency spectrum at the tower pass frequency, the identified condition of the wind turbine, an identified recommendation for the wind turbine, or combinations thereof.

9. The method of claim 8, further comprising storing the one or more the parameters, the derived parameter, the magnitude of the determined frequency spectrum at the tower pass frequency, the identified condition of the wind turbine, the identified recommendation for the wind turbine, or combinations thereof.

10. The method of claim 1, wherein the one or more line parameters comprise a current, a voltage, or a combination thereof.

11. The method of claim 1, wherein the derived parameter comprises an electromagnetic power, an electromagnetic torque, a mechanical power, a mechanical torque, or combinations thereof.

12. The method of claim 1, wherein the identified condition of the wind turbine comprises a fault in one or more blades of the wind turbine.

13. The method of claim 12, wherein the fault in the one or more blades of the wind turbine comprises a crack in a blade, a blade icing condition, a blade delamination, a blade core breakdown, or combinations thereof.

14. The method of claim 1, further comprising obtaining wind turbine rotor rotational speed data corresponding to the one or more acquired line parameters.

15. A non-transitory computer readable medium comprising one or more tangible media, wherein the one or more tangible media comprise routines for causing a computer
to perform steps of:

acquiring one or more line parameters at an output terminal of a generator, wherein the generator is operatively coupled to a wind turbine;

estimating a derived parameter based on the one or more acquired line parameters;

determining a frequency spectrum of the estimated derived parameter;

determining a magnitude of the determined frequency spectrum at a tower pass frequency; and

identifying a condition of the wind turbine based on a comparison of the magnitude of the determined frequency spectrum at the tower pass frequency with a threshold value.

16. The non-transitory computer readable medium of claim 15, wherein causing computer to perform step of identifying a recommendation for die wind turbine based on the identified condition of the wind turbine.

17. A device, comprising:

a processing unit configured to acquire one or more line parameters at an output terminal of a generator operatively coupled to a wind turbine, wherein the processing unit comprises:

a derived parameter estimating unit configured to estimate a derived parameter based on the one or more acquired line parameters;

a mathematical modeling unit configured to identify a non-stationary sub-signal corresponding to the derived parameter;

a signal processing unit configured to process the derived parameter to minimize an error introduced in the derived parameter due to the non-stationary sub-signal; and

an adaptive learning unit configured to adaptively learn at least one of a trend of progress of faults and a correlation of one or more of the one or more acquired line parameters, the derived parameter, a magnitude of a determined frequency spectrum at a tower pass frequency, an identified condition of the wind turbine, a recommendation for the wind turbine, or combinations thereof.

18. The device of claim 17, wherein the adaptive learning unit comprises a data repository configured to store the one or more acquired line parameters, the derived parameter, the magnitude of the determined frequency spectrum at the tower pass frequency, an identified condition of the wind turbine, the recommendation for the wind turbine, or combinations thereof.

19. The device of claim 17, wherein the derived parameter estimating unit comprises an electromagnetic parameter estimating unit, a mechanical parameter estimating unit, or both the electromagnetic parameter estimating unit and die mechanical parameter estimating unit.

20. The device of claim 17, wherein the signal processing unit is further configured for:

determining a frequency spectrum of the derived parameter;

obtaining a tower pass frequency from the determined frequency spectrum of the derived parameter; and

determining a magnitude of me frequency spectrum at the obtained tower pass frequency.