Early fault diagnosis method and system for wind turbine gearbox

Abstract

The application discloses a wind turbine gearbox early fault diagnosis method and system, and relates to the technical field of wind power equipment state monitoring. The method comprises the following steps: collecting multi-channel vibration signals of a gearbox and unit operation state data; obtaining a demodulation spectrum by performing angular domain resampling and envelope demodulation analysis on the vibration signals; positioning and calculating fault energy in the demodulation spectrum according to a preset fault characteristic frequency order ratio, and then constructing a fault feature vector; calculating a comprehensive evaluation index of the gearbox operation health degree in combination with the unit operation state data; and generating a fault diagnosis result when the index exceeds a preset threshold. In the above manner, the method effectively eliminates the influence of variable speed conditions, realizes accurate extraction and quantitative evaluation of early weak faults of the gearbox, improves the reliability and timeliness of fault diagnosis, and provides support for predictive maintenance of the wind turbine.

Description

Technical Field

[0001] This application relates to the field of wind power equipment condition monitoring technology, and in particular to a method and system for early fault diagnosis of wind turbine gearboxes. Background Technology

[0002] As a critical transmission component, the operating status of the wind turbine gearbox directly affects the overall reliability and power generation efficiency of the unit. Due to the randomness of wind speed, the gearbox operates under non-stationary conditions of varying speed and load for extended periods, posing significant challenges to condition monitoring and early fault diagnosis. Traditional vibration analysis methods are mostly based on steady-state assumptions. When processing variable speed signals, the fault characteristic frequencies become blurred with speed fluctuations, causing early, weak fault characteristics to be drowned out by background noise and difficult to identify effectively. Existing technologies struggle to provide accurate early warnings in the nascent stage of a fault, often only detecting it when it has progressed to a severe stage, potentially resulting in serious component damage and substantial economic losses.

[0003] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main purpose of this application is to provide a method and system for early fault diagnosis of wind turbine gearboxes, which aims to solve the technical problem that it is difficult to extract early fault characteristics under variable speed conditions in the prior art.

[0005] To achieve the above objectives, this application provides a method for early fault diagnosis of wind turbine gearboxes, the method comprising: Acquire multi-channel vibration signals and unit operating status data of the wind turbine gearbox within a preset detection time window; The multi-channel vibration signal is resampled in the angular domain to obtain angular domain vibration signals with equal angular intervals. The modulation component is extracted from the angular domain vibration signal by envelope demodulation analysis to obtain the demodulation spectrum. Based on the preset fault characteristic frequency order ratio, the fault energy is located from the demodulation spectrum and calculated. Based on the fault energy, a fault characteristic vector characterizing the state of the internal components of the gearbox is obtained. Based on the fault feature vector and the unit operating status data, a comprehensive evaluation index of gearbox operating health is calculated. When the comprehensive evaluation index is greater than the preset fault threshold, a fault diagnosis result is generated based on the comprehensive evaluation index.

[0006] In one embodiment, the step of performing angular domain resampling on the multi-channel vibration signal to obtain angular domain vibration signals with equal angular intervals, and extracting the modulation component from the angular domain vibration signal through envelope demodulation analysis to obtain the demodulation spectrum includes: The instantaneous angle sequence is calculated based on the speed pulse signal in the unit's operating status data; The multi-channel vibration signal within the preset detection time window is resampled in the angular domain according to the instantaneous angle sequence, and the non-stationary time domain signal is converted into a stationary angular domain signal to obtain angular domain vibration signals with equal angular intervals. The envelope is obtained by performing Hilbert transform on the angular domain vibration signal. The modulation component is extracted based on the envelope, and the demodulation spectrum is obtained by performing order spectrum analysis on the modulation component.

[0007] In one embodiment, the step of calculating the instantaneous angle sequence based on the speed pulse signal in the unit operating status data includes: Identify the rising edge time of each key phase pulse in the speed pulse signal of the unit operating status data; Based on the time interval between adjacent pulse moments, linear interpolation is used to calculate the instantaneous angle corresponding to each sampling time point within the preset detection time window; Arrange the instantaneous angles corresponding to all sampling time points in chronological order to obtain the instantaneous angle sequence.

[0008] In one embodiment, the step of locating and calculating the fault energy from the demodulation spectrum based on a preset fault characteristic frequency order, and obtaining a fault feature vector characterizing the state of internal components of the gearbox based on the fault energy includes: The expected number of failures of key components within one revolution is calculated based on the design parameters of the wind turbine gearbox, and the expected number of failures is used as the preset fault characteristic frequency order ratio. On the fault characteristic frequency order ratio axis in the demodulation spectrum, a target frequency range is located with a preset width centered on each preset fault characteristic frequency order ratio. Calculate the root mean square value of the amplitude within each target frequency interval, and use the root mean square value as the fault energy; The fault energies are combined in a preset order to obtain a fault feature vector that characterizes the state of the internal components of the gearbox.

[0009] In one embodiment, the step of calculating the root mean square value of the amplitude within each of the target frequency intervals and using the root mean square value as the fault energy includes: From the spectral data of the demodulated spectrum, extract the amplitude of all discrete spectral lines that fall within the target frequency range; The extracted amplitude values ​​are squared, averaged, and then squared to obtain the root mean square value of the vibration energy within the positioning interval. The root mean square value is used as the fault energy.

[0010] In one embodiment, the step of acquiring multi-channel vibration signals and unit operating status data of the wind turbine gearbox within a preset detection time window includes: Multi-channel vibration signals are obtained by synchronously measuring the signals of multiple vibration sensors installed at different bearing positions in the wind turbine gearbox within a preset detection time window; The unit's operating status data is obtained by collecting real-time speed pulse signals, power parameters, and operating condition parameters from the wind turbine monitoring system.

[0011] In one embodiment, the step of calculating the comprehensive evaluation index of gearbox operating health based on the fault feature vector and the unit operating status data includes: Extract the current torque value of the unit from the unit's operating status data; Based on the torque value, a preset load-energy benchmark table is queried to obtain the benchmark fault feature vector corresponding to the gearbox health status under the current torque value; Determine the relative degradation degree of each corresponding element in the fault feature vector and the baseline fault feature vector; The relative deterioration is weighted and fused to obtain a comprehensive evaluation index of the gearbox's operational health.

[0012] In one embodiment, the step of generating a fault diagnosis result based on the comprehensive evaluation index when the comprehensive evaluation index is greater than a preset fault threshold includes: When the comprehensive evaluation index exceeds the preset fault threshold, the excess value is determined based on the comprehensive evaluation index and the preset fault threshold. The severity level of the fault is determined based on the threshold range in which the exceeded value is located and the short-term trend of the comprehensive evaluation index. A fault diagnosis result is generated based on the fault severity level and the fault feature vector.

[0013] In one embodiment, after the step of generating a fault diagnosis result based on the comprehensive evaluation index when the comprehensive evaluation index is greater than a preset fault threshold, the method further includes: The fault diagnosis results, the corresponding fault feature vectors, the unit operating status and timestamps are stored in the historical database to form historical diagnosis data. Based on the historical diagnostic data, a trend graph of the energy of key fault characteristics and comprehensive evaluation indicators over time was plotted. When the trend graph shows abnormal changes, a gearbox status early warning report is automatically generated and sent to the wind farm central control system via a communication interface.

[0014] Furthermore, to achieve the above objectives, this application also proposes an early fault diagnosis system for wind turbine gearboxes, which includes: The data acquisition module is used to acquire multi-channel vibration signals and unit operating status data of the wind turbine gearbox within a preset detection time window; The signal processing module is used to perform angular domain resampling on the multi-channel vibration signal to obtain angular domain vibration signals with equal angular intervals, and to extract the modulation component from the angular domain vibration signal through envelope demodulation analysis to obtain the demodulation spectrum. The positioning calculation module is used to locate and calculate the fault energy from the demodulation spectrum according to the preset fault characteristic frequency order ratio, and obtain the fault feature vector characterizing the state of the internal components of the gearbox based on the fault energy. The comprehensive evaluation module is used to calculate a comprehensive evaluation index of the gearbox's operational health based on the fault feature vector and the unit's operating status data. The fault diagnosis module is used to generate a fault diagnosis result based on the comprehensive evaluation index when the comprehensive evaluation index is greater than the preset fault threshold.

[0015] In addition, to achieve the above objectives, this application also proposes an early fault diagnosis device for wind turbine gearboxes, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. The computer program is configured to implement the steps of the early fault diagnosis method for wind turbine gearboxes as described above.

[0016] In addition, to achieve the above objectives, the present invention also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the early fault diagnosis method for wind turbine gearbox as described above.

[0017] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the wind turbine gearbox early fault diagnosis method described above.

[0018] This application provides a method for early fault diagnosis of wind turbine gearboxes. The method involves collecting multi-channel vibration signals from the gearbox and turbine operating status data; performing angular domain resampling and envelope demodulation analysis on the vibration signals to obtain a demodulation spectrum; locating and calculating the fault energy in the demodulation spectrum based on a preset fault characteristic frequency order, thereby constructing a fault feature vector; and calculating a comprehensive evaluation index of the gearbox's operational health based on the turbine operating status data. When this index exceeds a preset threshold, a fault diagnosis result is generated. Through this approach, the method effectively eliminates the influence of variable speed operating conditions, achieves accurate extraction and quantitative evaluation of early, minor gearbox faults, improves the reliability and timeliness of fault diagnosis, and provides support for predictive maintenance of wind turbines. Attached Figure Description

[0019] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0020] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a flowchart illustrating an embodiment of the early fault diagnosis method for wind turbine gearboxes in this application. Figure 2 This is a schematic representation of the load-energy reference of an embodiment of the wind turbine gearbox early fault diagnosis method of this application; Figure 3 This is a schematic diagram of the module structure of the early fault diagnosis system for wind turbine gearboxes according to an embodiment of this application; Figure 4 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the early fault diagnosis method of wind turbine gearbox in the embodiments of this application.

[0022] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0023] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0024] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0025] The main solution of this application embodiment is to acquire multi-channel vibration signals and unit operating status data of the wind turbine gearbox within a preset detection time window; The multi-channel vibration signal is resampled in the angular domain to obtain angular domain vibration signals with equal angular intervals. The modulation component is extracted from the angular domain vibration signal by envelope demodulation analysis to obtain the demodulation spectrum. Based on the preset fault characteristic frequency order ratio, the fault energy is located from the demodulation spectrum and calculated. Based on the fault energy, a fault characteristic vector characterizing the state of the internal components of the gearbox is obtained. Based on the fault feature vector and the unit operating status data, a comprehensive evaluation index of gearbox operating health is calculated. When the comprehensive evaluation index is greater than the preset fault threshold, a fault diagnosis result is generated based on the comprehensive evaluation index.

[0026] Currently, the gearbox of a wind turbine is a critical transmission component, and its operating status directly affects the reliability and power generation efficiency of the entire unit. Due to the randomness of wind speed, the gearbox operates under non-stationary conditions of varying speed and load for extended periods, posing significant challenges to condition monitoring and early fault diagnosis. Traditional vibration analysis methods are mostly based on steady-state assumptions. When processing variable speed signals, the fault characteristic frequencies become blurred with speed fluctuations, causing early, weak fault characteristics to be drowned out by background noise and difficult to identify effectively. Existing technologies struggle to provide accurate early warnings in the nascent stage of a fault, often only detecting it when it has progressed to a severe stage, potentially resulting in serious component damage and substantial economic losses.

[0027] This application provides a solution that collects multi-channel vibration signals from the gearbox and unit operating status data; obtains the demodulation spectrum by performing angular domain resampling and envelope demodulation analysis on the vibration signals; locates and calculates the fault energy in the demodulation spectrum based on a preset fault characteristic frequency order, and then constructs a fault feature vector; calculates a comprehensive evaluation index of the gearbox's operational health by combining the unit operating status data; and generates a fault diagnosis result when the index exceeds a preset threshold. Through this method, the influence of variable speed operating conditions is effectively eliminated, enabling accurate extraction and quantitative evaluation of early-stage minor gearbox faults, improving the reliability and timeliness of fault diagnosis, and providing support for predictive maintenance of wind turbine units.

[0028] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device capable of performing the above functions, such as a wind turbine gearbox early fault diagnosis device. This embodiment does not specifically limit it in this way. The following uses a wind turbine gearbox early fault diagnosis device as an example to describe this embodiment and the following embodiments.

[0029] All actions involving the acquisition of signals, information, or data in this application are carried out in accordance with the relevant data protection laws and policies of the country where the application is located, and with the authorization of the owner of the relevant device.

[0030] This application provides a method for diagnosing early faults in wind turbine gearboxes, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the wind turbine gearbox early fault diagnosis method of this application.

[0031] In this embodiment, the method for diagnosing early faults in wind turbine gearboxes includes steps S10 to S50: Step S10: Obtain multi-channel vibration signals and unit operating status data of the wind turbine gearbox within a preset detection time window.

[0032] It should be noted that the preset detection time window refers to a continuous time period set for a one-time fault diagnosis analysis. The length of this window needs to be carefully considered: if it is too long, although it can include a more complete operating cycle to improve analysis accuracy, it will reduce the system response speed and may introduce interference due to changes in operating conditions; if it is too short, it may not be able to collect enough vibration cycle signals, resulting in insufficient accuracy of spectrum analysis. Typically, this window will cover data from the low-speed shaft of the gearbox at several tens of revolutions to ensure that recurring fault impact signals can be captured. The specific setting should be based on the actual situation.

[0033] Understandably, the process of acquiring multi-channel vibration signals and turbine operating status data of the wind turbine gearbox within a preset detection time window is accomplished collaboratively by hardware sensors and a data acquisition system. Accelerometer vibration sensors installed at different bearing housings or key locations within the gearbox, such as near the high-speed shaft, intermediate shaft, and low-speed shaft, synchronously acquire vibration signals. Simultaneously, the wind turbine's main control system acquires real-time turbine operating status data, including speed pulse signals (key phase signals), output power, and gear oil temperature, through communication interfaces (such as CAN bus or Modbus). All this data is timestamped with a clock synchronization mechanism and cached within a unified preset time window, providing the raw data foundation for subsequent synchronous analysis.

[0034] In one feasible implementation, the step of acquiring multi-channel vibration signals and unit operating status data of the wind turbine gearbox within a preset detection time window includes: Multi-channel vibration signals are obtained by synchronously measuring the signals of multiple vibration sensors installed at different bearing positions in the wind turbine gearbox within a preset detection time window; The unit's operating status data is obtained by collecting real-time speed pulse signals, power parameters, and operating condition parameters from the wind turbine monitoring system.

[0035] It should be noted that synchronous measurement signals refer to data that are strictly aligned in time and collected from multiple vibration sensors distributed on various bearing housings of the gearbox (such as high-speed shaft, medium-speed shaft, and low-speed shaft).

[0036] In practical implementation, vibration signals are typically acquired by IEPE (Integrated Circuit Piezoelectric) accelerometers. The output signal, after being filtered to prevent aliasing, is then converted to digital signal at high speed by a synchronous data acquisition card. Crucial synchronization is achieved through a shared sampling clock across the acquisition cards, ensuring that sampling points from all channels are acquired at the same microsecond-level timestamp. For vibration signal analysis, its time-domain waveform can be represented as follows: ,

[0037] in, Indicates the first The time-domain voltage signal collected by the vibration sensor is proportional to the acceleration of the mechanical vibration. : Time variable. The start time of data collection. : The preset detection time window length.

[0038] The acquisition of unit operating status data relies on the wind turbine monitoring and control system (SCADA). The speed pulse signal, also known as the key phase signal, represents one or more pulses generated per revolution. Acquired through a magnetic probe or encoder, it serves as the reference for subsequent angular domain resampling. Power parameters (such as instantaneous active power)... ) and operating parameters (such as gearbox oil temperature) The vibration signal is recorded at a lower sampling rate. To achieve accurate fusion of vibration signals and operational data, the system adds a high-precision absolute timestamp (usually from GPS or the system clock) to each speed pulse and operational data sample. Speed It can be estimated by calculating the reciprocal of the time interval between consecutive pulses.

[0039] in It is the pulse interval time. It is the pulse interval. It is the number of pulses per revolution.

[0040] Step S20: The multi-channel vibration signal is resampled in the angular domain to obtain angular domain vibration signals with equal angular intervals. The modulation component is extracted from the angular domain vibration signal by envelope demodulation analysis to obtain the demodulation spectrum. It should be noted that angular domain resampling is a signal processing technique that converts vibration signals originally sampled at equal time intervals in the time domain into signals sampled at equal angular intervals on the rotating shaft. This operation can eliminate the influence of speed fluctuations on signal analysis, allowing the fault characteristic frequencies of rotating components such as gears and bearings to be stably presented in the frequency spectrum.

[0041] Understandably, the process begins by using the previously acquired rotational speed pulse signal as an angular reference. A linear interpolation algorithm is then used to calculate the precise instantaneous shaft rotation angle for each time-interval vibration sample. Next, based on the calculated angle sequence, the vibration signal is resampled in the angular domain at preset equal angular intervals, such as 1024 points per revolution, generating an angular domain signal. Subsequently, envelope demodulation analysis is performed on the angular domain signal. Typically, a Hilbert transform is used to calculate the signal envelope to highlight the periodic impact components caused by local faults. Finally, a Fourier transform is performed on the angular domain envelope signal to transform the signal from the angular domain to the angular frequency domain. The resulting spectrum is the demodulation spectrum. The demodulation spectrum clearly reveals the modulation sidebands centered on the meshing frequency and its harmonics, or on the bearing fault characteristic order, thus revealing potential fault information.

[0042] In one feasible implementation, the steps of performing angular domain resampling on the multi-channel vibration signal to obtain angular domain vibration signals with equal angular intervals, and extracting the modulation component from the angular domain vibration signal through envelope demodulation analysis to obtain the demodulation spectrum include: The instantaneous angle sequence is calculated based on the speed pulse signal in the unit's operating status data; The multi-channel vibration signal within the preset detection time window is resampled in the angular domain according to the instantaneous angle sequence, and the non-stationary time domain signal is converted into a stationary angular domain signal to obtain angular domain vibration signals with equal angular intervals. The envelope is obtained by performing Hilbert transform on the angular domain vibration signal. The modulation component is extracted based on the envelope, and the demodulation spectrum is obtained by performing order spectrum analysis on the modulation component.

[0043] It should be noted that order ratio spectrum analysis is a spectral analysis technique performed in the angular or order domain. The horizontal axis, order ratio, is defined as the ratio of frequency to the fundamental frequency of rotation. For example, the meshing order ratio of a gear is equal to its number of teeth, while the order ratio of bearing fault characteristics is determined by its geometric dimensions. Since the order ratio is an inherent property of rotating machinery and does not change with rotational speed, the order ratio spectrum can stably map vibration characteristics under variable speed conditions to a fixed order ratio position.

[0044] Understandably, angular domain transformation is performed first, utilizing a high-precision rotational speed pulse signal. Assuming within a time window... Internal collection pulse arrival time Then, the instantaneous angle corresponding to each pulse moment can be calculated by linear interpolation (usually one revolution is set as...). (radians) and instantaneous rotational speed. Furthermore, for any sampling time point... instantaneous angle This can be obtained through pulse time interpolation. The goal of angular domain resampling is to achieve equal angular intervals. (like ) sequence The corresponding vibration signal amplitude is found through an interpolation algorithm. The core formula in this process is to establish a mapping relationship between angle and time. and its inverse function .

[0045]

[0046] in, For any point in time The instantaneous axis rotation angle (in radians). For the first The precise timestamp of each rotation speed pulse. t represents the number of pulses generated per revolution of the speed sensor, m is the pulse index, and t is the right boundary pulse number of the time interval.

[0047] Obtain angular domain signal Then, envelope demodulation is performed to extract the amplitude modulation information caused by the fault impulse. The Hilbert transform is a commonly used method; it constructs an analytic signal by convolving the signal with the envelope, producing a 90-degree phase-shifted copy. The magnitude of the analytic signal is the signal envelope. It contains the modulation component. The Hilbert transform and envelope calculation formulas are as follows: Analyzing the signal:

[0048] Envelope signal:

[0049] in, This is the vibration signal after angular domain resampling. For Hilbert transform operators, Constructed analytic signal, The calculated signal envelope is the modulation component.

[0050] Finally, regarding the envelope signal Perform order ratio spectral analysis. Because... Since it is an angular variable, the horizontal axis obtained by directly performing a Fourier Transform (FFT) on it is the "order ratio". This spectrum is the demodulation spectrum, which can clearly show the family of modulation sidebands centered on the gear meshing order ratio or its harmonics. The order ratio intervals between these sidebands correspond to the characteristic order ratio of the faulty component (such as the rotational frequency order ratio of a certain shaft), thereby accurately locating the fault source. The formula for calculating the order ratio spectrum is:

[0051] in, For the angular domain envelope signal, For Fast Fourier Transform, To demodulate the spectrum, It is the order ratio.

[0052] In one feasible implementation, the step of calculating the instantaneous angle sequence based on the speed pulse signal in the unit operating status data includes: Identify the rising edge time of each key phase pulse in the speed pulse signal of the unit operating status data; Based on the time interval between adjacent pulse moments, linear interpolation is used to calculate the instantaneous angle corresponding to each sampling time point within the preset detection time window; Arrange the instantaneous angles corresponding to all sampling time points in chronological order to obtain the instantaneous angle sequence.

[0053] As can be understood, a key phase pulse refers to a standard pulse signal generated by a key phase sensor (such as a photoelectric encoder or electromagnetic proximity switch) every revolution or fixed angle on a rotating shaft. It is typically triggered by a key phase marker (such as a keyway, protrusion, or reflective patch) on the shaft, providing a precise angular reference origin (or zero-phase point) for the entire rotating machinery system. The rising edge of each pulse precisely corresponds to the instant the shaft rotates to the position of this fixed marker, serving as the absolute time reference for calculating instantaneous rotational speed and performing angular domain resampling.

[0054] It should be understood that the data acquisition system records the voltage signal of the pulse channel at a sampling rate much higher than the pulse frequency (e.g., 100 kHz). First, a digital comparator identifies the moment when the voltage exceeds a preset threshold, and the rising edge of each pulse is precisely captured. These moments constitute the basic timestamps for angle calculation. Pulse period The calculation formula is:

[0055] in, For the first The precise timestamp of the rising edge of each bond phase pulse. For the first The precise timestamp of the rising edge of the (i.e., the previous) bond phase pulse.

[0056] After obtaining the pulse time sequence, the system needs to sample the vibration signal at each equally spaced time point. Calculate the corresponding instantaneous angle Assuming one key phase pulse is generated per revolution (i.e., the pulse is synchronized with the shaft rotation), linear interpolation is a computationally efficient method that meets accuracy requirements. Its core idea is: between two adjacent key phase pulses... and Assuming the shaft rotation speed changes uniformly (uniform acceleration model), then for any sampling time within this interval... ( (), which is relative to the previous reference point The percentage of the angle rotated is directly proportional to the percentage of time taken. The formula for linear interpolation of instantaneous angle is:

[0057] in, The vibration signal is the first Each sampling time point Sampling points The timing of the two bond phase pulses before and after the given time interval. This represents the total number of complete rotations counted from the start of the data window. This represents the proportion of the current sampling time within the current rotation cycle. Multiplying the radian corresponding to one revolution by the total ratio yields the instantaneous angle accumulated from the starting moment. (Unit: radians)

[0058] Finally, the system iterates through all vibration signal sampling time points within the preset detection time window. Calculate the corresponding values ​​for each point in sequence. These angle values ​​are then arranged in chronological order to obtain the desired instantaneous angle sequence. .

[0059] Step S30: Locate and calculate the fault energy from the demodulation spectrum according to the preset fault characteristic frequency order ratio, and obtain the fault characteristic vector characterizing the state of the internal components of the gearbox based on the fault energy. It should be noted that the fault characteristic frequency order refers to the frequency components of vibration signals exhibited in the order spectrum (angular frequency spectrum) of rotating machinery (such as gears and bearings in a gearbox) when a fault occurs. The ratio of the frequencies of these components to the rotational speed of the equipment is a fixed value, i.e., the order. The order is determined by the physical geometry of the component and is an inherent property independent of the rotational speed. Therefore, even under variable speed operating conditions, fault characteristics will stably appear at these fixed order positions, thus providing a reliable basis for fault diagnosis.

[0060] Understandably, the system pre-stores a fault characteristic order ratio library, which defines the characteristic order ratios and their sideband intervals corresponding to typical faults (such as pitting and spalling) of various components (such as bearings and gears at each stage). During analysis, the algorithm retrieves the corresponding preset fault characteristic frequency order ratios from the library based on the basic parameters of the current unit (such as the number of teeth and bearing model). On the demodulated spectrum, it locates the energy peaks within a certain bandwidth range (such as ±5% of the target order ratio) centered on these preset order ratios. Then, it integrates or calculates the root mean square value of this peak region as the fault energy of the fault mode. Finally, it combines the fault energy values ​​corresponding to all monitored components and fault types into a multi-dimensional array in a predetermined order, which constitutes the fault characteristic vector characterizing the overall health status of the gearbox.

[0061] In one feasible implementation, the step of locating and calculating the fault energy from the demodulation spectrum according to a preset fault characteristic frequency order, and obtaining a fault feature vector characterizing the state of internal components of the gearbox based on the fault energy includes: The expected number of failures of key components within one revolution is calculated based on the design parameters of the wind turbine gearbox, and the expected number of failures is used as the preset fault characteristic frequency order ratio. On the fault characteristic frequency order ratio axis in the demodulation spectrum, a target frequency range is located with a preset width centered on each preset fault characteristic frequency order ratio. Calculate the root mean square value of the amplitude within each target frequency interval, and use the root mean square value as the fault energy; The fault energies are combined in a preset order to obtain a fault feature vector that characterizes the state of the internal components of the gearbox.

[0062] It should be noted that the fault feature vector is a mathematical object that numerically represents the health status of the gearbox. Each element corresponds to a preset energy intensity index of a possible fault mode. It is formed by arranging the "fault energy" extracted from the vicinity of different fault feature frequency orders in the demodulation spectrum in a predetermined order.

[0063] In practical implementation, the characteristic order ratios of key components (such as bearings and gears) first need to be calculated based on the gearbox design parameters. For bearings, the fault characteristic order ratios (BPFO, BPFI, BSF, FTF) are determined by geometric parameters such as the number of balls and contact angle. For gears, the fault characteristic order ratio is the number of teeth (meshing frequency order ratio) or its multiples. These calculated order ratio values ​​are the preset fault characteristic frequency order ratios. These represent the theoretical number of cyclic impacts between the fault point and other components during one rotation of the component. Example of calculating the characteristic order ratio of a bearing outer ring fault:

[0064] in, This represents the number of balls. The diameter of the ball bearing is [value]. This refers to the bearing pitch diameter. It represents the contact angle. This is the fault characteristic order of the bearing outer ring, which is the number of impacts between the fault point and the balls per revolution of the shaft.

[0065] Subsequently, on the order ratio coordinate axis of the demodulated spectrum, for each preset fault characteristic order... Set a target frequency range centered on it, for example Interval width The choice of [specific parameter] requires a trade-off between interference resistance and robustness to speed fluctuations. Next, the root mean square (RMS) value of all spectral line amplitudes within this interval is calculated as the fault energy for this fault mode. The RMS value effectively reflects the overall vibration energy level within the frequency band, is insensitive to random noise and transient fluctuations, and is more stable. The formula for calculating fault energy is:

[0066] in, Located in the target frequency range The number of spectral lines within. For the demodulation spectrum The amplitude of each spectral line. The calculated fault energy is characterized by the intensity of vibrational energy near the fault's characteristic frequency.

[0067] Finally, the energy values ​​of all target failure modes are calculated. They are combined into a multidimensional array according to a preset, fixed order (e.g., bearings first, then gears, arranged according to the transmission chain). This array That is the final fault feature vector.

[0068] In one feasible implementation, the step of calculating the root mean square value of the amplitude within each of the target frequency intervals and using the root mean square value as the fault energy includes: From the spectral data of the demodulated spectrum, extract the amplitude of all discrete spectral lines that fall within the target frequency range; The extracted amplitude values ​​are squared, averaged, and then squared to obtain the root mean square value of the vibration energy within the positioning interval. The root mean square value is used as the fault energy.

[0069] It should be noted that the root mean square (RMS) value is calculated by first squaring each instantaneous value of the signal, then averaging these squares, and finally taking the square root of the average to restore the result to the same dimensions as the original signal. When processing vibration signals, the RMS value can stably reflect the effective amplitude or energy level of the signal over a period of time. It is insensitive to instantaneous peak values ​​and therefore represents the sustained intensity of the signal better than peak values ​​or average values. It is often used for fault energy assessment of rotating machinery.

[0070] In its implementation, the algorithm first determines the order of fault characteristics based on a preset ratio. and bandwidth A continuous target frequency range is determined on the order axis of the demodulated spectrum. The demodulated spectrum itself is composed of discrete spectral lines, each corresponding to a specific order and amplitude. The algorithm scans all spectral lines, filters out all spectral lines whose order coordinates fall within the target interval, and records their amplitudes. Extract the target frequency band amplitude sequence:

[0071] in, For the first The order corresponding to each discrete spectral line. For the first The amplitude of each discrete spectral line. This represents the total number of spectral lines falling within the target frequency range.

[0072] Subsequently, these extracted amplitude sequences Perform root mean square (RMS) calculation. This calculation first squares each amplitude value to obtain the instantaneous power; then sums all the squared values ​​and divides by the number of spectral lines. The average power is obtained; finally, the square root of the average power is the root mean square value of the vibration energy within the target frequency range. This calculation method comprehensively reflects the overall level of all resonant sidebands and modulation energy within the frequency band, providing a more robust characterization of fault severity than simply searching for the maximum amplitude. The formula for calculating fault energy (root mean square value of the frequency band) is:

[0073] in, For the target frequency range, the first The amplitude of each spectral line. This represents the total number of spectral lines within the target frequency range. The calculated fault energy is physically represented as the average value of the vibration energy near the target fault characteristic frequency.

[0074] Step S40: Based on the fault feature vector and the unit operating status data, calculate the comprehensive evaluation index of gearbox operating health.

[0075] It should be noted that the comprehensive evaluation index of gearbox operational health is used to comprehensively reflect the overall technical status of the gearbox under current operating conditions. It is obtained by information fusion and weighted calculation of the fault characteristic vectors reflecting each component and the unit operating status data (such as load, speed, etc.) that affect its status.

[0076] Understandably, models based on weighted scoring or machine learning regression are typically used. For weighted scoring, the system assigns a weight to each fault energy element in the fault feature vector according to its importance, while dynamically adjusting the weights or scores based on operational status data (such as appropriately relaxing the vibration threshold under high load conditions). Finally, a health score between 0 and 100 is calculated through weighted summation or other comprehensive functions. For more intelligent methods, the fault feature vector and key operational status data are directly used as input to a pre-trained regression model (such as gradient boosting trees or neural networks). The model automatically learns the complex nonlinear relationship between the fault feature vector and health and directly outputs a comprehensive evaluation index.

[0077] In one feasible implementation, the step of calculating the comprehensive evaluation index of gearbox operating health based on the fault feature vector and the unit operating status data includes: Extract the current torque value of the unit from the unit's operating status data; Based on the torque value, a preset load-energy benchmark table is queried to obtain the benchmark fault feature vector corresponding to the gearbox health status under the current torque value; Determine the relative degradation degree of each corresponding element in the fault feature vector and the baseline fault feature vector; The relative deterioration is weighted and fused to obtain a comprehensive evaluation index of the gearbox's operational health.

[0078] It should be noted that the load-energy reference table establishes the correspondence between the unit's output torque and the theoretical reference values ​​of each fault characteristic vector of the gearbox under healthy conditions.

[0079] In practical implementation, the first step is to acquire parameters that characterize the gearbox load level in real time, typically the torque value. This is because it is directly related to the internal stress and vibration energy levels of the gearbox. The system will look up this information in a predefined load-energy reference table. Figure 2 This table uses torque values. As input, output a baseline fault feature vector with the same dimensions as the current fault feature vector. Each element in the vector Represents torque Next, the The theoretical fault energy baseline value that a fault mode (such as a specific bearing or gear fault) should exhibit under a fully healthy state. Then, the relative degradation degree of each corresponding element in the fault feature vector and the baseline fault feature vector is determined, using the following formula:

[0080] in, The fault feature vector obtained in real time is the first... The fault energy value of each element. To be based on the current torque The first one found from the benchmark table The baseline energy value for each failure mode. For the first The relative degradation of each failure mode quantitatively reflects the degree of deterioration of that mode relative to the healthy baseline.

[0081] Next, the real-time fault feature vector is calculated. Each element Relative to the corresponding element in the reference vector Relative degradation It is usually defined using ratios or relative deviations. Then, considering the varying importance of different components and failure modes within the gearbox and their differing impacts on overall health, it is necessary to assess each degree of degradation. Perform weighted fusion. For each Assign a weight coefficient (satisfy Finally, a comprehensive evaluation index is obtained through weighted calculation. The weighted calculation formula is:

[0082] in, For the first The relative degradation of each failure mode. For the first The weight of each failure mode represents its importance. It is an optional function used to scale or transform the degradation (such as taking the logarithm to smooth large values). It is a final mapping function used to normalize the weighted sum to a specified range (such as 0-100). It is a calculated comprehensive health index; the lower the value, the worse the health status.

[0083] Step S50: When the comprehensive evaluation index is greater than the preset fault threshold, generate a fault diagnosis result based on the comprehensive evaluation index.

[0084] It should be noted that the preset fault threshold is a numerical limit used to divide continuous comprehensive evaluation indicators into different health status levels, thereby determining whether the gearbox is faulty or the severity of the fault. It is set by the gearbox design requirements, industry standards or expert experience.

[0085] Understandably, the system will compare the calculated comprehensive evaluation index HI with the preset multi-level fault thresholds in the database in real time. If HI is below the minimum threshold (e.g., HI<70), a fault or critical alarm diagnosis result will be generated, and an emergency shutdown recommendation may be triggered. If HI is in the middle range (e.g., 70 ≤ HI < 85), a performance degradation or early warning diagnosis result will be generated, indicating that close monitoring or planned maintenance is required. If the HI is higher than the highest threshold (e.g., HI ≥ 85), it is considered to be in a healthy state.

[0086] This judgment logic is implemented through conditional judgment statements (if-else or switch-case). The diagnostic results are stored together with the timestamp and indicator values ​​and sent to the monitoring interface, providing clear decision-making basis for operation and maintenance personnel.

[0087] In one feasible implementation, the step of generating a fault diagnosis result based on the comprehensive evaluation index when the comprehensive evaluation index is greater than a preset fault threshold includes: When the comprehensive evaluation index exceeds the preset fault threshold, the excess value is determined based on the comprehensive evaluation index and the preset fault threshold. The severity level of the fault is determined based on the threshold range in which the exceeded value is located and the short-term trend of the comprehensive evaluation index. A fault diagnosis result is generated based on the fault severity level and the fault feature vector.

[0088] It should be noted that the excess value refers to the portion of the comprehensive evaluation index that exceeds the preset fault threshold. It is used to quantify the degree to which the current health status deviates from the normal benchmark and reflects the specific extent of abnormal gearbox operation.

[0089] In practical implementation, the first step is to calculate the out-of-limit value. The system presets multiple levels of fault thresholds (such as early warning thresholds). Alarm threshold (etc.). When comprehensive evaluation indicators When the value falls below a certain threshold, an alarm of that level is considered triggered. (Exceeding the limit) The calculation is usually for the current The difference between the value and the nearest next more severe threshold can be expressed as:

[0090] in, It is the current The fault threshold violated by the value (e.g., if Below the warning threshold But it is higher than the alarm threshold ,but ), It is a comprehensive health index obtained through calculation. The larger the value, the greater the deviation from a healthy state.

[0091] After obtaining the out-of-limit value, the system maps it to a preset threshold range (such as "slightly out of limit", "moderately out of limit", "severely out of limit"), and analyzes it simultaneously. The short-term trend of the indicator. Trend analysis can be performed by calculating the trend within a recent time window. The first derivative of the value or the slope of the linear regression To achieve:

[0092] in, It's a point in time. It is a health indicator for the corresponding time period. and These are the average values ​​of time and health within the window, respectively. This is the number of data points within the time window. Slope Characterized the rate of deterioration of health ( This indicates that the indicator is declining, meaning the situation is deteriorating.

[0093] Finally, the system combines the range in which the out-of-limit value falls with the deterioration trend of the health index (e.g., "severely out of limit" and "rapidly deteriorating"), and uses a predefined rule base or fuzzy inference system to determine the final fault severity level (e.g., "Level 1 Warning", "Level 2 Alert", "Level 3 Emergency"). This level is then combined with specific frequency components or energy distributions prominent in the fault feature vector to generate a refined fault diagnosis result that includes the severity of the fault and the potential location of the fault (e.g., "severe bearing wear and rapid deterioration").

[0094] In one feasible implementation, after the step of generating a fault diagnosis result based on the comprehensive evaluation index when the comprehensive evaluation index is greater than a preset fault threshold, the method further includes: The fault diagnosis results, the corresponding fault feature vectors, the unit operating status and timestamps are stored in the historical database to form historical diagnosis data. Based on the historical diagnostic data, a trend graph of the energy of key fault characteristics and comprehensive evaluation indicators over time was plotted. When the trend graph shows abnormal changes, a gearbox status early warning report is automatically generated and sent to the wind farm central control system via a communication interface.

[0095] It should be noted that the gearbox status early warning report is a comprehensive document generated based on trend analysis of historical diagnostic data. By capturing abnormal change patterns in key fault characteristic energy and comprehensive evaluation indicators, it provides early warnings of potential future gearbox failures. This report not only includes a current status assessment but also provides predictive maintenance recommendations based on historical evolution trends. Furthermore, it pushes structured early warning information to the wind farm's central control system via standardized communication protocols, providing forward-looking support for operational and maintenance decisions.

[0096] In its implementation, the system first stores multidimensional data in time-series format. Assume each record in the historical diagnostic database contains a timestamp. Comprehensive evaluation indicators Fault feature vector (in Representing the Energy values ​​for each frequency band or fault mode), and unit operating status parameters (such as power). Rotation speed (etc.). Creating trend charts essentially involves visualizing this time-series data and using statistical process control (SPC) or machine learning methods to detect anomalies. For key fault characteristic energy... It can be calculated within a certain time window. (e.g., moving average over 30 days) and moving standard deviation :

[0097] in, It is a time window The number of data points within the range. When the latest data point... Exceeding the preset control limit (e.g.) ) or indicators When the rate of decline continuously exceeds the threshold, it is judged as an "abnormal change".

[0098] After an anomaly triggers the generation of an early warning report, the system automatically extracts relevant data within the abnormal time period and generates a structured document using a pre-set report template. The report typically includes: a description of the anomaly characteristics (e.g., "bearing outer ring fault energy increased by 30% within 24 hours"), a curve showing the change in the comprehensive health index, related operating condition analysis, and maintenance recommendations (e.g., "It is recommended to inspect the gearbox bearing during the next maintenance window"). Finally, the report is encapsulated in JSON / XML format and sent to a designated message queue or database in the wind farm control system via communication interfaces such as OPC UA, IEC 61400-25, or RESTful API, completing the closed-loop transmission of early warning information.

[0099] This embodiment provides a method for early fault diagnosis of wind turbine gearboxes. It involves collecting multi-channel vibration signals from the gearbox and turbine operating status data; obtaining a demodulation spectrum by performing angular domain resampling and envelope demodulation analysis on the vibration signals; locating and calculating the fault energy in the demodulation spectrum based on a preset fault characteristic frequency order, and then constructing a fault feature vector; calculating a comprehensive evaluation index of the gearbox's operational health based on the turbine operating status data; and generating a fault diagnosis result when this index exceeds a preset threshold. Through this method, the influence of variable speed operating conditions is effectively eliminated, enabling accurate extraction and quantitative evaluation of early, minor gearbox faults, improving the reliability and timeliness of fault diagnosis, and providing support for predictive maintenance of wind turbines.

[0100] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the early fault diagnosis method of wind turbine gearbox in this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0101] This application also provides an early fault diagnosis system for wind turbine gearboxes; please refer to [reference needed]. Figure 3 The early fault diagnosis system for wind turbine gearboxes includes: Data acquisition module 10 is used to acquire multi-channel vibration signals and unit operating status data of wind turbine gearbox within a preset detection time window; Signal processing module 20 is used to perform angular domain resampling on the multi-channel vibration signal to obtain angular domain vibration signals with equal angular intervals, and to extract the modulation component from the angular domain vibration signal through envelope demodulation analysis to obtain the demodulation spectrum; The positioning calculation module 30 is used to locate and calculate the fault energy from the demodulation spectrum according to the preset fault characteristic frequency order ratio, and obtain the fault feature vector characterizing the state of the internal components of the gearbox based on the fault energy. The comprehensive evaluation module 40 is used to calculate a comprehensive evaluation index of the gearbox operating health based on the fault feature vector and the unit operating status data. The fault diagnosis module 50 is used to generate a fault diagnosis result based on the comprehensive evaluation index when the comprehensive evaluation index is greater than the preset fault threshold.

[0102] In one feasible implementation, the signal processing module 20 is further configured to calculate an instantaneous angle sequence based on the speed pulse signal in the unit operating status data; The multi-channel vibration signal within the preset detection time window is resampled in the angular domain according to the instantaneous angle sequence, and the non-stationary time domain signal is converted into a stationary angular domain signal to obtain angular domain vibration signals with equal angular intervals. The envelope is obtained by performing Hilbert transform on the angular domain vibration signal. The modulation component is extracted based on the envelope, and the demodulation spectrum is obtained by performing order spectrum analysis on the modulation component.

[0103] In one feasible implementation, the signal processing module 20 is further configured to identify the rising edge time of each key phase pulse in the speed pulse signal of the unit operating status data; Based on the time interval between adjacent pulse moments, linear interpolation is used to calculate the instantaneous angle corresponding to each sampling time point within the preset detection time window; Arrange the instantaneous angles corresponding to all sampling time points in chronological order to obtain the instantaneous angle sequence.

[0104] In one feasible implementation, the positioning calculation module 30 is further configured to calculate the expected number of times a key component will fail within one revolution of rotation based on the design parameters of the wind turbine gearbox, and use the expected number of times as a preset fault characteristic frequency order ratio; On the fault characteristic frequency order ratio axis in the demodulation spectrum, a target frequency range is located with a preset width centered on each preset fault characteristic frequency order ratio. Calculate the root mean square value of the amplitude within each target frequency interval, and use the root mean square value as the fault energy; The fault energies are combined in a preset order to obtain a fault feature vector that characterizes the state of the internal components of the gearbox.

[0105] In one feasible implementation, the positioning calculation module 30 is further configured to extract the amplitude of all discrete spectral lines falling within the target frequency range from the spectral data of the demodulated spectrum. The extracted amplitude values ​​are squared, averaged, and then squared to obtain the root mean square value of the vibration energy within the positioning interval. The root mean square value is used as the fault energy.

[0106] In one feasible implementation, the data acquisition module 10 is further used to obtain multi-channel vibration signals based on the synchronous measurement signals of multiple vibration sensors located at different bearing seats of the wind turbine gearbox within a preset detection time window. The unit's operating status data is obtained by collecting real-time speed pulse signals, power parameters, and operating condition parameters from the wind turbine monitoring system.

[0107] In one feasible implementation, the comprehensive evaluation module 40 is further configured to extract the current torque value of the unit from the unit operating status data; Based on the torque value, a preset load-energy benchmark table is queried to obtain the benchmark fault feature vector corresponding to the gearbox health status under the current torque value; Determine the relative degradation degree of each corresponding element in the fault feature vector and the baseline fault feature vector; The relative deterioration is weighted and fused to obtain a comprehensive evaluation index of the gearbox's operational health.

[0108] In one feasible implementation, the fault diagnosis module 50 is further configured to determine an over-limit value based on the comprehensive evaluation index and the preset fault threshold when the comprehensive evaluation index is greater than the preset fault threshold. The severity level of the fault is determined based on the threshold range in which the exceeded value is located and the short-term trend of the comprehensive evaluation index. A fault diagnosis result is generated based on the fault severity level and the fault feature vector.

[0109] In one feasible implementation, the fault diagnosis module 50 is further configured to store the fault diagnosis result, the corresponding fault feature vector, the unit operating status and the timestamp into a historical database to form historical diagnosis data. Based on the historical diagnostic data, a trend graph of the energy of key fault characteristics and comprehensive evaluation indicators over time was plotted. When the trend graph shows abnormal changes, a gearbox status early warning report is automatically generated and sent to the wind farm central control system via a communication interface.

[0110] The early fault diagnosis system for wind turbine gearboxes provided in this application employs the early fault diagnosis method for wind turbine gearboxes described in the above embodiments, which can solve the technical problem of difficulty in extracting early fault characteristics under variable speed operating conditions. Compared with the prior art, the beneficial effects of the early fault diagnosis system for wind turbine gearboxes provided in this application are the same as those of the early fault diagnosis method for wind turbine gearboxes provided in the above embodiments, and other technical features of the early fault diagnosis system for wind turbine gearboxes are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0111] This application provides a wind turbine gearbox early fault diagnosis device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the wind turbine gearbox early fault diagnosis method in the above embodiment 1.

[0112] The following is for reference. Figure 4 This document illustrates a structural schematic diagram of a wind turbine gearbox early fault diagnosis device suitable for implementing embodiments of this application. The wind turbine gearbox early fault diagnosis device in this application embodiment may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and vehicle terminals (e.g., vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Figure 4 The wind turbine gearbox early fault diagnosis device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0113] like Figure 4As shown, the early fault diagnosis device for wind turbine gearboxes may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in ROM (Read Only Memory) 1002 or a program loaded from storage device 1003 into RAM (Random Access Memory) 1004. RAM 1004 also stores various programs and data required for the operation of the early fault diagnosis device for wind turbine gearboxes. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via bus 1005. Input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, LCDs (Liquid Crystal Displays), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the wind turbine gearbox early fault diagnosis equipment to communicate wirelessly or wiredly with other devices to exchange data. Although the figure shows a wind turbine gearbox early fault diagnosis equipment with various systems, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems can be implemented alternatively.

[0114] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0115] The early fault diagnosis device for wind turbine gearboxes provided in this application, employing the early fault diagnosis method for wind turbine gearboxes described in the above embodiments, can solve the technical problem of early fault diagnosis for wind turbine gearboxes. Compared with the prior art, the beneficial effects of the early fault diagnosis device for wind turbine gearboxes provided in this application are the same as those of the early fault diagnosis method for wind turbine gearboxes provided in the above embodiments, and other technical features of this early fault diagnosis device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0116] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0117] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0118] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the wind turbine gearbox early fault diagnosis method in the above embodiments.

[0119] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, RAM (Random Access Memory), ROM (Read Only Memory), EPROM (Erasable Programmable Read Only Memory or Flash Memory), optical fibers, CD-ROM (CD-Read Only Memory), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0120] The aforementioned computer-readable storage medium may be included in the early fault diagnosis equipment for wind turbine gearboxes; or it may exist independently and not be assembled into the early fault diagnosis equipment for wind turbine gearboxes.

[0121] The aforementioned computer-readable storage medium carries one or more programs. When the aforementioned one or more programs are executed by the wind turbine gearbox early fault diagnosis device, the wind turbine gearbox early fault diagnosis device acquires multi-channel vibration signals and unit operating status data of the wind turbine gearbox within a preset detection time window. The multi-channel vibration signal is resampled in the angular domain to obtain angular domain vibration signals with equal angular intervals. The modulation component is extracted from the angular domain vibration signal by envelope demodulation analysis to obtain the demodulation spectrum. Based on the preset fault characteristic frequency order ratio, the fault energy is located from the demodulation spectrum and calculated. Based on the fault energy, a fault characteristic vector characterizing the state of the internal components of the gearbox is obtained. Based on the fault feature vector and the unit operating status data, a comprehensive evaluation index of gearbox operating health is calculated. When the comprehensive evaluation index is greater than the preset fault threshold, a fault diagnosis result is generated based on the comprehensive evaluation index.

[0122] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including LAN (Local Area Network) or WAN (Wide Area Network)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0123] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0124] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0125] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described method for diagnosing early faults in wind turbine gearboxes, thereby solving the technical problem of diagnosing early faults in wind turbine gearboxes. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the wind turbine gearbox early fault diagnosis method provided in the above embodiments, and will not be repeated here.

[0126] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the wind turbine gearbox early fault diagnosis method described above.

[0127] The computer program product provided in this application can solve the technical problem of difficulty in extracting early fault characteristics under variable speed operating conditions. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the early fault diagnosis method for wind turbine gearboxes provided in the above embodiments, and will not be repeated here.

[0128] The above are only some embodiments of this application and do not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A method for early fault diagnosis of wind turbine gearboxes, characterized in that, The method for early fault diagnosis of wind turbine gearboxes includes: Acquire multi-channel vibration signals and unit operating status data of the wind turbine gearbox within a preset detection time window; The multi-channel vibration signal is resampled in the angular domain to obtain angular domain vibration signals with equal angular intervals. The modulation component is extracted from the angular domain vibration signal by envelope demodulation analysis to obtain the demodulation spectrum. Based on the preset fault characteristic frequency order ratio, the fault energy is located from the demodulation spectrum and calculated. Based on the fault energy, a fault characteristic vector characterizing the state of the internal components of the gearbox is obtained. Based on the fault feature vector and the unit operating status data, a comprehensive evaluation index of gearbox operating health is calculated. When the comprehensive evaluation index is greater than the preset fault threshold, a fault diagnosis result is generated based on the comprehensive evaluation index.

2. The method as described in claim 1, characterized in that, The steps of performing angular domain resampling on the multi-channel vibration signal to obtain angular domain vibration signals with equal angular intervals, and extracting the modulation component from the angular domain vibration signal through envelope demodulation analysis to obtain the demodulation spectrum include: The instantaneous angle sequence is calculated based on the speed pulse signal in the unit's operating status data; The multi-channel vibration signal within the preset detection time window is resampled in the angular domain according to the instantaneous angle sequence, and the non-stationary time domain signal is converted into a stationary angular domain signal to obtain angular domain vibration signals with equal angular intervals. The envelope is obtained by performing Hilbert transform on the angular domain vibration signal. The modulation component is extracted based on the envelope, and the demodulation spectrum is obtained by performing order spectrum analysis on the modulation component.

3. The method as described in claim 2, characterized in that, The step of calculating the instantaneous angle sequence based on the speed pulse signal in the unit operating status data includes: Identify the rising edge time of each key phase pulse in the speed pulse signal of the unit operating status data; Based on the time interval between adjacent pulse moments, linear interpolation is used to calculate the instantaneous angle corresponding to each sampling time point within the preset detection time window; Arrange the instantaneous angles corresponding to all sampling time points in chronological order to obtain the instantaneous angle sequence.

4. The method as described in claim 1, characterized in that, The step of locating and calculating the fault energy from the demodulation spectrum based on a preset fault characteristic frequency order, and obtaining a fault feature vector characterizing the state of internal components of the gearbox based on the fault energy includes: The expected number of failures of key components within one revolution is calculated based on the design parameters of the wind turbine gearbox, and the expected number of failures is used as the preset fault characteristic frequency order ratio. On the fault characteristic frequency order ratio axis in the demodulation spectrum, a target frequency range is located with a preset width centered on each preset fault characteristic frequency order ratio. Calculate the root mean square value of the amplitude within each target frequency interval, and use the root mean square value as the fault energy; The fault energies are combined in a preset order to obtain a fault feature vector that characterizes the state of the internal components of the gearbox.

5. The method as described in claim 4, characterized in that, The step of calculating the root mean square value of the amplitude within each of the target frequency intervals and using the root mean square value as the fault energy includes: From the spectral data of the demodulated spectrum, extract the amplitude of all discrete spectral lines that fall within the target frequency range; The extracted amplitude values ​​are squared, averaged, and then squared to obtain the root mean square value of the vibration energy within the positioning interval. The root mean square value is used as the fault energy.

6. The method as described in claim 1, characterized in that, The steps for acquiring multi-channel vibration signals and unit operating status data of the wind turbine gearbox within a preset detection time window include: Multi-channel vibration signals are obtained by synchronously measuring the signals of multiple vibration sensors installed at different bearing positions in the wind turbine gearbox within a preset detection time window; The unit's operating status data is obtained by collecting real-time speed pulse signals, power parameters, and operating condition parameters from the wind turbine monitoring system.

7. The method as described in claim 1, characterized in that, The step of calculating the comprehensive evaluation index of gearbox operating health based on the fault feature vector and the unit operating status data includes: Extract the current torque value of the unit from the unit's operating status data; Based on the torque value, a preset load-energy benchmark table is queried to obtain the benchmark fault feature vector corresponding to the gearbox health status under the current torque value; Determine the relative degradation degree of each corresponding element in the fault feature vector and the baseline fault feature vector; The relative deterioration is weighted and fused to obtain a comprehensive evaluation index of the gearbox's operational health.

8. The method as described in claim 1, characterized in that, The step of generating a fault diagnosis result based on the comprehensive evaluation index when the comprehensive evaluation index is greater than the preset fault threshold includes: When the comprehensive evaluation index exceeds the preset fault threshold, the excess value is determined based on the comprehensive evaluation index and the preset fault threshold. The severity level of the fault is determined based on the threshold range in which the exceeded value is located and the short-term trend of the comprehensive evaluation index. A fault diagnosis result is generated based on the fault severity level and the fault feature vector.

9. The method as described in claim 1, characterized in that, After the step of generating a fault diagnosis result based on the comprehensive evaluation index when the comprehensive evaluation index is greater than the preset fault threshold, the method further includes: The fault diagnosis results, the corresponding fault feature vectors, the unit operating status and timestamps are stored in the historical database to form historical diagnosis data. Based on the historical diagnostic data, a trend graph of the energy of key fault characteristics and comprehensive evaluation indicators over time was plotted. When the trend graph shows abnormal changes, a gearbox status early warning report is automatically generated and sent to the wind farm central control system via a communication interface.

10. A wind turbine gearbox early fault diagnosis system, characterized in that, The wind turbine gearbox early fault diagnosis system includes: The data acquisition module is used to acquire multi-channel vibration signals and unit operating status data of the wind turbine gearbox within a preset detection time window; The signal processing module is used to perform angular domain resampling on the multi-channel vibration signal to obtain angular domain vibration signals with equal angular intervals, and to extract the modulation component from the angular domain vibration signal through envelope demodulation analysis to obtain the demodulation spectrum. The positioning calculation module is used to locate and calculate the fault energy from the demodulation spectrum according to the preset fault characteristic frequency order ratio, and obtain the fault feature vector characterizing the state of the internal components of the gearbox based on the fault energy. The comprehensive evaluation module is used to calculate a comprehensive evaluation index of the gearbox's operational health based on the fault feature vector and the unit's operating status data. The fault diagnosis module is used to generate a fault diagnosis result based on the comprehensive evaluation index when the comprehensive evaluation index is greater than the preset fault threshold.

Images