Wind turbine generator variable pitch gear box signal acquisition method, fault detection method and system

Abstract

The invention relates to a signal acquisition method of a transmission system, in particular to a signal acquisition method, a fault detection method and a fault detection system of a variable pitch gear box of a wind turbine generator. The method comprises the steps that a first angle signal of the input end of a variable pitch motor is collected, and a sample angle signal is generated according to the first angle signal; collecting a second angle signal generated when the output shaft of the variable pitch gear box is driven by the variable pitch motor; synchronously receiving the sample angle signal and the second angle signal; and according to the sample angle signal and the second angle signal, obtaining a torsional vibration transmission error signal of the variable pitch gear box. According to the invention, the original angle signal is collected at the input end of the variable-pitch motor as the sample angle signal, and the sample angle signal and the second angle signal collected at the output shaft end of the variable-pitch gear box are synchronously received and processed, so that the torsional vibration transmission error signal can be more accurately obtained; and signal attenuation and distortion caused by step-by-step transmission of a multi-stage mechanical interface are avoided to the greatest extent.

Description

Technical Field

[0001] This invention relates to a signal acquisition method for transmission systems, and in particular to a signal acquisition method, fault detection method, and system for wind turbine pitch gearboxes. Background Technology

[0002] In wind turbine units, the pitch gearbox is a key transmission component for blade angle control. Its performance directly affects the power regulation, overspeed protection, and emergency shutdown functions of the wind turbine. Once a fault occurs, such as output shaft breakage, planetary carrier cracking, or gear tooth breakage, the pitch control accuracy will decrease, the blades will be subjected to abnormal loads, and in severe cases, it may lead to blade breakage or failure of other major components.

[0003] Existing fault detection technologies generally include two types: operational parameter analysis and housing vibration signal analysis. Among them, operational parameter analysis often lags behind in responding to potentially progressive but ultimately severe failure modes, such as output shaft cracking in pitch gearboxes, which may lead to instantaneous brittle fracture. Housing vibration signal analysis suffers from insufficient fault sensitivity, resulting in a high false alarm rate, and strong interference environments lead to a high false alarm rate. Summary of the Invention

[0004] In a first aspect, embodiments of this application provide a method for acquiring signals from a wind turbine pitch gearbox, in order to solve the problem of not being able to reflect faults in the wind turbine pitch gearbox in a timely and accurate manner.

[0005] The signal acquisition method includes: Acquire the first angle signal from the input end of the pitch motor, and generate a sample angle signal based on the first angle signal; The second angle signal generated when the output shaft of the pitch gearbox is driven by the pitch motor is acquired; Simultaneously receive the sample angle signal and the second angle signal; Based on the sample angle signal and the second angle signal, the torsional vibration transmission error signal of the pitch gearbox is obtained.

[0006] By adopting the above scheme, this method collects the original signal at the input end of the pitch motor as a sample angle signal, and synchronously receives and processes it with the second angle signal collected at the output shaft end of the pitch gearbox. This allows for more accurate acquisition of the torsional vibration transmission error signal, minimizing signal attenuation and distortion caused by the step-by-step transmission through multiple mechanical interfaces.

[0007] In one possible implementation, the first angle signal is a raw differential analog signal or a raw pulse train signal acquired from the input of the pitch motor.

[0008] In one possible implementation, generating the sample angle signal based on the first angle signal includes: The original differential analog signal is allocated to generate an execution feedback signal and the sample angle signal; The execution feedback signal is sent to the pitch driver, and the pitch driver performs real-time closed-loop control of the pitch motor based on the execution feedback signal.

[0009] In one possible implementation, the acquisition of the second angle signal generated when the output shaft of the pitch gearbox is driven by the pitch motor includes: An output shaft encoder is mounted on the wind turbine hub using a mounting bracket, and the output shaft encoder is connected to the process hole machined on the pitch output gear via a connecting fixture. The second angle signal is obtained through the output shaft encoder.

[0010] In one possible implementation, obtaining the torsional vibration transmission error signal of the pitch gearbox based on the sample angle signal and the second angle signal further includes: The first instantaneous angular velocity signal is obtained by performing a first-order differential operation on the sample angle signal; The second instantaneous angular velocity signal is obtained by performing a first-order differential operation on the second angle signal; The torsional vibration transmission error signal is obtained based on the first instantaneous angular velocity signal and the second instantaneous angular velocity signal.

[0011] In one possible implementation, obtaining the torsional vibration transmission error signal of the pitch gearbox based on the sample angle signal and the second angle signal further includes: The first instantaneous angular acceleration signal is obtained by performing a second-order differential operation on the sample angle signal; The second instantaneous angular acceleration signal is obtained by performing a second-order differential operation on the second angle signal; The torsional vibration transmission error signal is obtained based on the first instantaneous angular acceleration signal and the second instantaneous angular acceleration signal.

[0012] In one possible implementation, the torsional vibration transmission error signal includes an angle transmission error signal, which is the difference between the angle signal calculated from the second angle signal according to the speed ratio and the sample angle signal.

[0013] In one possible implementation, the torsional vibration transmission error signal includes an angular velocity transmission error signal, which is the difference between the angular velocity signal calculated from the second instantaneous angular velocity signal according to the speed ratio and the first instantaneous angular velocity signal.

[0014] In one possible implementation, the torsional vibration transmission error signal includes an angular acceleration transmission error signal, which is the difference between the angular acceleration signal after the second instantaneous angular acceleration signal is converted according to the speed ratio and the first instantaneous angular acceleration signal.

[0015] Secondly, embodiments of this application provide a method for detecting faults in a wind turbine pitch gearbox, comprising: The torsional vibration transmission error signal is obtained based on the torsional vibration transmission error signal acquisition method of the wind turbine pitch gearbox described in the first aspect. Based on the torsional vibration transmission error signal, feature extraction is performed on the torsional vibration transmission error signal, and it is compared with a preset standard state threshold to determine whether the pitch gearbox has a fault.

[0016] In one possible implementation, The torsional vibration transmission error signal includes the angle transmission error signal, the angular velocity transmission error signal, and / or the angular acceleration transmission error signal; The process of determining whether the pitch gearbox has a fault includes: The angle transmission error signal is feature extracted and compared with a preset standard state threshold. If the error exceeds the preset standard state threshold, it is determined that the pitch gear box backlash is too large or the output shaft of the pitch gear box has a deep crack. Alternatively, feature extraction can be performed on the angular velocity transmission error signal, and it can be compared with a preset standard state threshold. If the preset standard state threshold is exceeded, it can be determined that there is tooth surface damage in the gear of the pitch gearbox. Alternatively, feature extraction can be performed on the angular acceleration transmission error signal, and it can be compared with a preset standard state threshold. If the preset standard state threshold is exceeded, it can be determined that there is a crack in the spline of the pitch gearbox output shaft and / or a deep crack in the output shaft of the pitch gearbox.

[0017] In one possible implementation, the step of extracting features from the angle transmission error signal and comparing it with a preset standard state threshold, and determining that the pitch gearbox backlash is too large when it exceeds the preset standard state threshold, includes: Control the pitch motor to rotate in the forward direction, calculate and record the average value of the forward angle transmission error signal under steady-state conditions; Control the pitch motor to rotate in the opposite direction, calculate and record the average value of the reverse angle transmission error signal under steady state; The absolute difference of the angle transmission error signal is obtained by subtracting the average value of the forward angle transmission error signal from the average value of the reverse angle transmission error signal. The absolute difference is compared with a preset standard state threshold, which is the pitch gearbox backlash threshold. When the absolute difference is greater than the preset standard state threshold, it is determined that the pitch gearbox backlash is too large.

[0018] In one possible implementation, the forward and reverse rotations of the pitch motor are small-angle, low-torque rotations.

[0019] In one possible implementation, the step of extracting features from the angle transmission error signal and comparing it with a preset standard state threshold, and determining that the output shaft of the pitch gearbox has a deep crack when the error exceeds the preset standard state threshold, includes: Obtain the spectrum of the angle transmission error signal based on the angle transmission error signal; Monitor the amplitude of the output shaft rotational frequency harmonics in the spectrum of the angle transmission error signal; The amplitude of the output shaft frequency harmonic in the spectrum of the angle transmission error signal is compared with a preset standard state threshold, wherein the preset standard state threshold is the harmonic amplitude threshold. When the preset standard state threshold is exceeded continuously, it is determined that the output shaft of the pitch gearbox has a deep crack.

[0020] In one possible implementation, the step of extracting features from the angular velocity transmission error signal and comparing it with a preset standard state threshold, and determining that the gears of the pitch gearbox have tooth surface damage when the error exceeds the preset standard state threshold, includes: Spectral analysis is performed on the angular velocity transmission error signal to obtain the angular velocity transmission error signal spectrum / order spectrum; Identify significant peak frequencies or orders in the spectrum / order spectrum of the angular velocity transmission error signal; The peak frequency or order is matched with the theoretical fault characteristic frequencies of each stage of the transmission gear in the pitch gearbox. When a peak value or order component matching the theoretical fault characteristic frequency of a certain gear is identified in the spectrum of the second angular velocity transmission error signal, it is determined that there is tooth surface damage in that gear of the pitch gearbox.

[0021] In one possible implementation, the step of extracting features from the angular acceleration transmission error signal and comparing it with a preset standard state threshold, and determining that the output shaft spline of the pitch gearbox has an early-middle crack when the threshold is exceeded, includes: The signal spectrum of the output shaft spline of the pitch gearbox is obtained based on the angular acceleration transmission error signal. The energy of the torsional resonance peak of the signal spectrum of the output shaft spline of the pitch gearbox is monitored; The energy of the torsional resonance peak of the signal spectrum of the output shaft spline of the pitch gearbox is compared with the preset standard state threshold, which is an energy threshold. When the energy of the torsional resonance peak of the signal spectrum of the output shaft spline of the pitch gearbox continuously exceeds the preset standard state threshold, it is determined that there is a crack in the output shaft spline of the pitch gearbox.

[0022] In one possible implementation, the step of extracting features from the angular acceleration transmission error signal and comparing it with a preset standard state threshold, and determining that the output shaft spline of the pitch gearbox has a crack when the threshold is exceeded, includes: Obtain the reference angular acceleration transmission error signal of the output shaft of the pitch gearbox under healthy conditions; A short-time Fourier transform is performed on the reference angular acceleration transmission error signal of the output shaft of the pitch gearbox under healthy conditions to obtain a first time-frequency diagram; Monitor the current angular acceleration transmission error signal of the pitch gearbox under operating conditions; A short-time Fourier transform is performed on the current angular acceleration transmission error signal of the pitch gearbox in its working state to obtain a second time-frequency diagram; Monitor the modulation amplitude of the output shaft's natural frequency in the second time-frequency diagram; The modulation amplitude of the output shaft's natural frequency in the second time-frequency diagram is compared with the modulation amplitude of the output shaft's natural frequency in the first time-frequency diagram; When the modulation amplitude of the output shaft natural frequency in the second time-frequency diagram exceeds the preset standard state threshold compared with the modulation amplitude of the output shaft natural frequency in the first time-frequency diagram, it is determined that there is a crack in the spline of the output shaft of the pitch gearbox. The preset standard state threshold is the modulation amplitude threshold.

[0023] In one possible implementation, the step of extracting features from the angular acceleration transmission error signal and comparing it with a preset standard state threshold, and determining that the output shaft spline of the pitch gearbox has a crack when the threshold is exceeded, includes: Obtain the reference angular acceleration transmission error signal of the output shaft of the pitch gearbox under healthy conditions; Perform spectral coherence analysis on the reference angular acceleration transmission error signal to obtain a first spectral coherence diagram; Monitor the current angular acceleration transmission error signal of the pitch gearbox under operating conditions; Perform spectral coherence analysis on the current angular acceleration transmission error signal to obtain a second spectral coherence map; Calculate the number of output shaft rotational frequency harmonics near the output shaft torsional natural frequency in the second spectral coherence diagram; The number of output shaft frequency harmonics is compared with the harmonic number threshold. When the number of output shaft frequency harmonics monitored continuously exceeds the health baseline data, it is determined that there are early to mid-stage cracks in the spline of the output shaft of the pitch gearbox.

[0024] In one possible implementation, the step of extracting features from the angular acceleration transmission error signal and comparing it with a preset standard state threshold, and determining that the output shaft of the pitch gearbox has a deep crack when the threshold is exceeded, includes: The signal spectrum is obtained based on the angular acceleration transmission error signal; Identify and track the first-order torsional natural frequency in the signal spectrum; The offset of the first-order torsional natural frequency in the signal spectrum towards the lower frequency direction is compared with the preset standard state threshold, which is the offset threshold. When the first-order torsional natural frequency is detected to have shifted to a lower frequency direction for an extended period exceeding a preset standard state threshold, it is determined that there is a deep crack in the output shaft of the pitch gearbox.

[0025] Thirdly, embodiments of this application provide a wind turbine pitch gearbox fault detection system, comprising: The first sensing module acquires the first angle signal from the input end of the pitch motor; The second sensing module acquires the second angle signal generated when the output shaft of the pitch gearbox is driven by the pitch motor. The signal distribution module receives the first angle signal output by the first sensing module and generates an execution feedback signal and a sample angle signal based on the first angle signal. The torsional vibration signal acquisition module synchronously receives the sample angle signal and the second angle signal; and obtains the torsional vibration transmission error of the pitch gearbox based on the sample angle signal and the second angle signal. The fault analysis module extracts features from the torsional vibration transmission error signal, compares it with a preset standard state threshold, obtains the analysis results of the torsional vibration transmission error signal, and determines whether the pitch gearbox has a fault based on the analysis results of the torsional vibration transmission error signal. Attached Figure Description

[0026] Figure 1 This is a flowchart of the first embodiment; Figure 2 This is a schematic diagram of the structure upon which the second embodiment is based; Figure 3 for Figure 2 Internal structure diagram; Figure 4 This is a flowchart of the second embodiment; Figure 5 This is a schematic diagram of the encoder mounting structure in the second embodiment; Figure 6 Flowchart of the fourth embodiment; Figure 7 This is a flowchart of the fifth embodiment; Figure 8 This is a flowchart of the sixth embodiment; Figure 9 This is a flowchart of the seventh embodiment; Figure 10 This is a time-domain waveform diagram of the angular acceleration torsional vibration signal of the output shaft under healthy conditions in the seventh embodiment; Figure 11 The time-domain waveform of the angular acceleration torsional vibration signal of the output shaft spline crack state of the pitch gearbox in the seventh embodiment is shown. Figure 12 This is a diagram showing the spline crack state of the output shaft of the pitch gearbox in the seventh embodiment. Figure 13 The torsional vibration spectrum of the output shaft under healthy conditions in the seventh embodiment; Figure 14 The torsional vibration spectrum of the output shaft spline crack state of the pitch gearbox in the seventh embodiment; Figure 15 This is a flowchart of the eighth embodiment; Figure 16 This is a time-frequency diagram of the torsional vibration signal under healthy conditions of the output shaft in the eighth embodiment; Figure 17 The time-frequency diagram of torsional vibration signal under spline crack condition of output shaft of pitch gearbox in the eighth embodiment; Figure 18 This is a flowchart of the ninth embodiment; Figure 19 This is a coherence diagram of the torsional vibration signal spectrum under healthy conditions of the output shaft in the ninth embodiment; Figure 20 The image shows the torsional vibration signal spectrum coherence diagram under the spline crack condition of the output shaft of the pitch gearbox in the ninth embodiment. Figure 21 This is a flowchart of the tenth embodiment; Figure 22 This is a schematic diagram of the module structure of the eleventh embodiment. Detailed Implementation

[0027] The following detailed description, in conjunction with specific embodiments and accompanying drawings, clarifies that the described embodiments are only a portion, not all, of the embodiments. All other embodiments obtained by those skilled in the art based on the following embodiments without inventive effort are also within the scope of protection of this invention.

[0028] It should be understood that if the controllers or control circuits involved in the embodiments are conventional control technologies or units for those skilled in the art, such as the control circuits of the controllers, they can be implemented by those skilled in the art using existing technologies.

[0029] The disclosure of the embodiments provides many different implementations or examples for different ways of implementing the invention. To simplify the disclosure of the invention, specific examples of components and arrangements are described in the embodiments. Of course, these are merely examples and are not intended to limit the invention. Furthermore, reference numerals and / or reference letters may be repeated in different examples in the embodiments; such repetition is for simplification and clarity and does not in itself indicate a relationship between the various implementations and / or arrangements discussed.

[0030] To facilitate understanding of the problems to be solved and the background and principles of the technical solutions implemented in the embodiments of this application, the existing basic situation will be explained first.

[0031] To detect faults in the transmission components of wind turbine generators, existing wind turbine monitoring and data acquisition systems (SCADA) can perform health status analysis of the pitch system by monitoring operating parameters. Commonly used parameters include the pitch motor's current, torque, and temperature, as well as the pitch angle setpoint / feedback value and the blade's wind energy utilization coefficient. The operating status of the pitch system is inferred by analyzing the time series and statistical characteristics of these parameters or building models (such as threshold-based, trend analysis, or machine learning models). The core drawback of this monitoring scheme is its insufficient early warning capability. Although SCADA data is readily available and inexpensive, this monitoring method is inherently indirect. It reflects the overall effect of the pitch system's execution of commands or its macroscopic operating status, rather than directly capturing key mechanical components. For example, identifying and detecting early damage signals of the output shaft cannot be accomplished through the analysis of wind turbine monitoring and data acquisition systems.

[0032] Therefore, for fault types such as output shaft cracking of the pitch gearbox, which may be a progressive development but could eventually lead to serious failures such as instantaneous brittle fracture, monitoring through SCADA data is often delayed, and the early warning window may not be sufficient to support effective preventive maintenance measures before a catastrophic failure occurs. Once the output shaft fractures, it will not only cause the pitch system to fail completely, but may also trigger a chain reaction, causing blade cracking or even more serious unit damage accidents.

[0033] The proposed vibration signal monitoring solution involves installing accelerometers, such as piezoelectric or impact pulse sensors, on key components of the pitch gearbox housing to continuously collect high-frequency vibration signals. Subsequently, signal processing techniques, such as wavelet transform and variational mode decomposition, are used to analyze the signals, aiming to extract sensitive feature components related to faults. Based on these features, either statistical index thresholds are constructed, or machine learning models are trained, to achieve monitoring of the pitch system's operating status and provide early warning of anomalies.

[0034] However, due to the lengthy transmission path of vibration signals, which involves multiple stages from the crack fault source → shaft → bearing → bearing housing → gearbox housing → sensor, vibration signals caused by early, weak cracks attenuate significantly during this multi-stage transmission. Furthermore, the output shaft speed of the pitch gearbox is typically extremely low, falling into the ultra-low-speed operating range. The inherent characteristics of the accelerometer in the low-frequency range cause a sharp drop in its signal-to-noise ratio, making the weak vibration energy generated by early cracks in the output shaft easily masked by mechanical background noise. Moreover, in actual operating environments, vibrations from the wind turbine nacelle drivetrain and the energy generated by the rotor rotation itself often account for over 90% of the total vibration energy of the pitch gearbox housing. Interference signals such as rotational modulation during blade rotation, impacts from the tower shadow effect, and abnormal vibrations caused by nacelle drivetrain faults are easily misinterpreted by the vibration monitoring system as faults in the pitch gearbox itself, leading to frequent invalid maintenance alarms and unplanned downtime.

[0035] Therefore, embodiments of this application provide a method for analyzing fault detection of wind turbine pitch gearbox by collecting and transmitting torsional vibration transmission error signals. Combined with the transmission characteristics of the pitch gearbox, the method can accurately detect changes in torsional vibration transmission error signals caused by early cracks and determine the fault location.

[0036] Torsional vibration refers to the periodic or random angular deflection and speed fluctuation of a drive shaft caused by periodic or random torque disturbances in a transmission system. Continuous torsional vibration subjects components such as drive shafts, gears, and connecting fixtures to alternating stresses, accelerating wear and fatigue, and potentially leading to cracking or even breakage. Therefore, torsional vibration signals can directly reflect faults in transmission components. Combining the above characteristics and the transmission structure and component composition of a pitch gearbox, the first embodiment of this application provides a signal acquisition method for a wind turbine pitch gearbox, such as... Figure 1 As shown, the method includes: S01. Acquire the first angle signal at the input end of the pitch motor, and generate a sample angle signal based on the first angle signal; S02. Acquire the second angle signal generated when the output shaft of the pitch gearbox is driven by the pitch motor; S03. Simultaneously receive the sample angle signal and the second angle signal; S04. Obtain the torsional vibration transmission error signal of the pitch gearbox based on the sample angle signal and the second angle signal.

[0037] The steps S01-S04 above represent the composition of the execution content of the method, and do not represent the execution order of the method. When the subsequent steps do not depend on the execution results of other steps, their order of expression can be adjusted. The following descriptions of method steps are explained with reference to this rule.

[0038] Specifically, the first embodiment proposes a method for acquiring torsional vibration transmission error signals based on the characteristics of torsional vibration signals, and combines this method with the working characteristics of the pitch gearbox. Figure 2 and Figure 3 As shown, the pitch motor 1 and pitch gearbox 2 are installed in the wind turbine hub 3. The output end of the pitch motor 1 is connected to and drives the input end of the pitch gearbox 2. The pitch gearbox 2 has transmission components such as a sun gear, planet gears, a gear ring, and a planet carrier (not shown in the figure). The planet carrier is connected to and drives the output shaft of the pitch gearbox. The inner end of the output shaft of the pitch gearbox is connected to the planet carrier through an output shaft spline. The output end of the output shaft of the pitch gearbox is provided with a pitch output tooth 4. The pitch output tooth 4 meshes with the inner ring of the lateral pitch bearing 5. The inner ring of the pitch bearing 5 is connected to the blades. The outer ring of the pitch bearing 5 is installed on the wind turbine hub 3. During operation, the pitch motor 1 rotates, and the input end of the pitch gearbox 2 is driven by the pitch motor 1. After being decelerated by the transmission components inside the pitch gearbox 2, the output shaft of the pitch gearbox 2 drives the pitch output tooth 4 of the pitch gearbox to rotate, thereby driving the inner ring of the pitch bearing 5 to rotate, and finally driving the blades to rotate to achieve pitch control.

[0039] In this embodiment, the second angle signal of the output shaft of the pitch gearbox needs to be collected as a reference quantity for the torsional vibration transmission error signal. However, when the torsional vibration of the pitch motor is transmitted to the pitch gearbox, the output speed of the pitch gearbox is extremely low, usually less than 5 rpm. It is usually difficult to detect the subtle changes in the torsional vibration signal by directly collecting and testing the second angle signal of the pitch gearbox. Although the abnormal torsional vibration signal of the output shaft of the pitch gearbox can detect the fault in the transmission-related structural components when a fault occurs, it is difficult to determine the degree of the fault source.

[0040] Therefore, in this embodiment, a torsional vibration transmission error signal will be generated between the torsional vibration of the pitch motor and the torsional vibration of the pitch gearbox. Analysis of the torsional vibration transmission error signal from different dimensions can accurately analyze the fault source and fault severity. To obtain the torsional vibration transmission error signal, a rotary transformer or encoder inherent at the tail end of the pitch motor is used to collect the first angle signal, which is the original differential analog signal or original pulse train signal collected from the input end of the pitch motor. Specifically, when using a rotary transformer, the collected signal is the original differential analog signal; when using an encoder, the collected signal is the original pulse signal. Taking a rotary transformer as an example, the original differential analog signal collected by the rotary transformer is typically used for angle position feedback in pitch control. The first angle signal of the motor input shaft is collected. In this embodiment, based on the first angle signal The sample angle signal is generated. A high-precision absolute encoder is additionally installed at the end of the output shaft of the pitch gearbox to directly acquire the second angle signal of the output shaft. .

[0041] Since the output shaft of the wind turbine pitch gearbox is an ultra-low speed reciprocating rotary device, the high-precision absolute encoder is a 15-bit or higher absolute encoder, that is, a 32,768-line or higher absolute encoder.

[0042] During signal acquisition, to ensure strict clock alignment between the two signals and avoid phase errors affecting subsequent analysis, the system employs a precise time protocol for network clock synchronization, combined with a hardware triggering mechanism, to achieve high-precision clock synchronization. In traditional solutions, the analog signal from the input resolver must first be sent to the pitch driver, processed by the internal resolver decoding chip, and then the digital angle information is provided externally. This process is affected by the driver's internal processing circuitry, decoding delay, and communication transmission, introducing millisecond-level lag. This delay degrades the quality of the torsional vibration transmission error signal, significantly reducing the accuracy and reliability of the gearbox condition monitoring algorithm.

[0043] To address this issue and ensure precise signal synchronization, this embodiment first inputs the acquired first angle signal into a signal distributor. The signal distributor divides the input first angle signal into two signals. The signal distributor employs a high-input-impedance buffer amplifier to ensure that the current drawn from the signal source is minimal, not affecting the amplitude and phase of the original signal. Simultaneously, the signal distributor integrates an isolation circuit to ensure that the two distributed signals are electrically isolated from each other, cutting off loop interference. One signal output from the signal distributor is normally sent to the pitch driver. The resolver decoding chip inside the pitch driver converts it into digital angle and speed signals for real-time closed-loop control of the pitch control system; this path function remains unaffected. The other identical real-time signal output from the signal distributor, as a sample angle signal, is sent to a separate external decoding module dedicated to condition monitoring. This external decoding module also includes a high-performance resolver decoding chip, whose decoding algorithm is optimized for condition monitoring (e.g., higher resolution, faster update rate). The decoded digital angle signal is then sent as a sample angle signal to an external torsional vibration signal acquisition unit.

[0044] The torsional vibration signal acquisition unit simultaneously receives the sample angle signal and the digital signal directly from the output absolute encoder. Since the two signals originate from the synchronous sampling clock of the same acquisition unit, and the input signal avoids driver delay, microsecond-level synchronization between the input and output signals is achieved. This provides a precise data foundation for the subsequent generation of torsional vibration transmission error signals and the reliable operation of the state monitoring algorithm.

[0045] To accurately acquire the second angle signal at the output shaft end of the pitch gearbox, in this embodiment, due to the structural limitations of the hub, the traditional method of mounting an encoder on the output shaft of the pitch gearbox cannot be used. Furthermore, because the output shaft of the pitch gearbox rotates at extremely low speeds, the encoder on the output shaft needs to accurately and directly acquire the second angle signal as a reference quantity for calculating the torsional vibration transmission error signal. Ordinary methods such as setting the encoder laterally via a code disk or code tape, or other indirect acquisition methods, cannot meet the accuracy requirements of this embodiment. Therefore, a second embodiment is further proposed. The second embodiment, based on the first embodiment, provides an encoder installation method and a method for accurately acquiring the second angle signal, such as... Figure 4 As shown, it specifically includes: S021. An output shaft encoder is mounted on the wind turbine hub by means of a mounting bracket, and the input end of the output shaft encoder is connected to the process hole machined on the pitch output tooth by means of a connecting fixture; S022. Obtain the second angle signal through the encoder.

[0046] Among them, combined Figure 2, Figure 3 and Figure 5 As shown, the output shaft encoder 6 is mounted on the mounting bracket 7, which is mounted on the hub and provides a mounting surface for the pitch output tooth 4 located on the output shaft of the pitch gearbox. The output shaft encoder 6 is connected to the center process hole 41 of the pitch output tooth 4 of the pitch gearbox via a connecting fixture 10, ensuring precise alignment between the encoder shaft centerline and the output shaft centerline, minimizing measurement errors caused by installation misalignment. The center process hole of the pitch output tooth 4 of the pitch gearbox serves as both the mounting hole for the pitch output tooth 4 and the mounting hole for the connecting fixture 10 to which the input end of the output shaft encoder 6 is connected. After installation, the encoder shaft centerline remains coaxial with the pitch gearbox shaft centerline. Under these conditions, the rotary transformer or encoder at the pitch motor end and the output shaft encoder of the pitch gearbox constitute two encoders. The angle signals generated by both encoders are then connected to a dedicated PLC torsional vibration data acquisition system.

[0047] To ensure precise alignment between the encoder shaft centerline and the output shaft centerline, the mounting bracket 7 can be installed using the combination of the first bolt 8 and the second bolt 9. The first bolt 8 connects the encoder 6 to the mounting bracket 7, and the second bolt 9 connects the mounting bracket 7 to the hub 3. This combination adjustment achieves fine-tuning. The core of the entire adjustment process is to first perform coarse adjustment, then fine adjustment, and finally tighten in stages. The detailed operating steps are as follows: First, clean all contact surfaces, ensuring that the process holes 41 on the pitch output teeth of the pitch gearbox, the mating surfaces of the connecting fixture 10, and the mounting surface of the encoder 6 are free of oil, dust, or burrs. Screw the threaded end of the connecting fixture 10 into the process hole 41 of the pitch output teeth of the pitch gearbox, ensuring it is in place. The encoder body is connected to the mounting bracket 7 via the first bolt 8. Do not tighten the first bolt 8 immediately, allowing for slight movement of the encoder. Pry the mounting bracket to adjust the approximate position of the entire assembly, aligning the encoder output shaft approximately with the center line of the right end of the connecting fixture. Connect the encoder output shaft to the right end of the connecting fixture.

[0048] Next, using the first bolt 8 as a fine-tuning mechanism, and taking advantage of the small gap between the first bolt 8 and the mounting hole of the mounting bracket, the position of the encoder 6 in the X and Y directions (radial) and its angular runout are finely adjusted by alternately tightening and loosening these bolts. The goal is to adjust the center of the encoder output shaft to coincide with the center height of the process hole of the pitch output tooth. The mounting bracket is then fixed to the hub using the second bolt 9, but not tightened yet, so that the entire assembly still has a certain adjustment margin.

[0049] Finally, the radial runout of the encoder shaft was measured using a dial indicator, with a target runout value set at 0.05 mm. After confirming that the alignment accuracy met the requirements, the first bolt 8 was tightened to secure the encoder to the mounting bracket. Subsequently, the second bolt 9 was tightened to secure the mounting bracket to the hub 3. After all bolts were tightened, the radial runout of the encoder output shaft was checked again to ensure that the alignment accuracy had not shifted during the tightening process.

[0050] The connecting fixture 10 can be a coupling, and the encoder 6 is connected to the process hole 41 through the connecting fixture 10.

[0051] In this embodiment, the torsional vibration transmission error signal includes one or more of the following dimensions: angle transmission error signal, angular velocity transmission error signal, and angular acceleration transmission error signal. Analysis of the torsional vibration transmission error signals in each dimension allows for further analysis of the fault source and fault severity. The calculation methods for the torsional vibration transmission error signals in each of the above dimensions are explained below: The raw angle signal acquired by the encoder at the motor end Perform first-order differentiation to obtain the angular velocity signal of the input shaft. ;right Perform second-order differentiation to obtain the instantaneous angular acceleration signal of the input axis. Similarly, the original angle signal acquired by the encoder at the output end... Perform first-order differentiation to obtain the angular velocity signal of the output shaft. ;right Perform second-order differentiation to obtain the angular acceleration signal of the output shaft. .

[0052] Based on the acquired and processed motion state signals at both ends, the torsional vibration transmission error signal of the pitch gearbox can be calculated. The torsional vibration transmission error signal is key dynamic data characterizing the gear meshing state and potential faults (such as output shaft cracking, tooth surface damage, etc.) inside the pitch gearbox. Assuming the theoretical reduction ratio of the pitch gearbox is r, the torsional vibration transmission error signal can be calculated along the following dimensions: angle transmission error signal. Defined as the deviation between the actual angle of the motor input shaft and the angle of the output shaft after conversion according to the theoretical reduction ratio, the calculation formula is: Angular velocity transmission error signal Defined as the deviation between the actual angular velocity of the motor input shaft and the angular velocity of the output shaft after being converted according to the theoretical reduction ratio, the calculation formula is: Angular acceleration transmits error signals Defined as the deviation between the actual angular acceleration of the motor input shaft and the angular acceleration of the output shaft after being converted according to the theoretical reduction ratio, the calculation formula is: .

[0053] The above angle transmits error signals Angular velocity transmission error signal and angular acceleration transmission error signal As different dimensions of the torsional vibration transmission error signal, it contains rich information on the dynamic characteristics of the gearbox and is the core input data for subsequent condition monitoring and fault early warning.

[0054] It should also be noted that in the above signal acquisition process, the encoder at the tail end of the pitch motor is inherent, that is, in order to control the pitch motor and provide motion angle feedback, an existing encoder is preferred, and a rotary transformer is preferred.

[0055] Because of the above method, the wind turbine pitch gearbox signal acquisition method in the first embodiment generates a sample angle signal through the first angle signal at the pitch motor end, and processes it synchronously with the second angle signal at the output shaft end to generate a torsional vibration transmission error signal. This torsional vibration transmission error signal has a shorter transmission path, which avoids signal attenuation and distortion caused by the step-by-step transmission of the vibration signal from the crack → shaft → bearing → bearing housing → gearbox housing → sensor during the monitoring of pitch gearbox vibration signal. It can more accurately and sensitively detect pitch gearbox faults.

[0056] Furthermore, by calculating the transmission error signal between the first angle signal at the input shaft end of the pitch motor and the second angle signal at the output shaft of the pitch gearbox, the common interference signals caused by factors such as the nacelle drivetrain, hub, and rotor rotation can be canceled out. Compared with traditional methods for measuring gearbox vibration, the wind turbine pitch gearbox signal acquisition method of the first embodiment significantly reduces the possibility of false alarms caused by external interference sources and improves the reliability of monitoring results. At the same time, compared with traditional methods for measuring gearbox vibration, the torsional vibration transmission error signal facilitates order tracking analysis, giving it excellent variable speed adaptability and enabling effective extraction of pitch gearbox fault characteristics under actual operating conditions of pitch motor speed variations.

[0057] The first embodiment described above obtains the torsional vibration transmission error signal to determine whether there is a fault in the pitch gearbox. In fact, by processing the torsional vibration transmission error signal in different forms, it is possible to further determine the fault source and fault type. The following embodiments further propose multiple embodiments based on the torsional vibration transmission error signal collected in the first embodiment to detect different dimensions of the torsional vibration transmission signal to obtain the fault source and fault type of the wind turbine pitch gearbox fault detection method.

[0058] In general, based on the first embodiment described above, a third embodiment is further provided. The wind turbine pitch gearbox fault detection method of the third embodiment includes: The torsional vibration transmission error signal is obtained according to the torsional vibration transmission error signal acquisition method of the wind turbine pitch gearbox described in the first embodiment. Based on the torsional vibration transmission error signal, it is compared with a preset standard state threshold to obtain the fault source and / or fault type of the pitch gearbox.

[0059] In the third embodiment, the torsional vibration transmission error signal is analyzed in different dimensions and compared with the corresponding preset standard state thresholds in different dimensions. This can accurately obtain the fault source and fault type of the wind turbine pitch gearbox, which is beneficial for accurate fault source finding and early fault detection.

[0060] In this embodiment, the torsional vibration transmission error signal includes an angle transmission error signal, an angular velocity transmission error signal, and / or an angular acceleration transmission error signal; The fault types of the pitch gearbox are as follows: The angle transmission error signal is feature extracted and compared with a preset standard state threshold. If the error exceeds the preset standard state threshold, it is determined that the pitch gearbox backlash is too large or the output shaft has a deep crack. Alternatively, feature extraction can be performed on the angular velocity transmission error signal, and it can be compared with a preset standard state threshold. If the preset standard state threshold is exceeded, it can be determined that there is tooth surface damage in the gear of the pitch gearbox. Alternatively, feature extraction can be performed on the angular acceleration transmission error signal, and it can be compared with a preset standard state threshold. If the preset standard state threshold is exceeded, it can be determined that there is a crack in the spline of the pitch gearbox output shaft and / or a deep crack in the output shaft of the pitch gearbox.

[0061] In general, this embodiment classifies the torsional vibration error signal into three types: angle transmission error signal, angular velocity transmission error signal, and / or angular acceleration transmission error signal. By processing, analyzing, and comparing the three types of signals, the different fault types at different positions of the gearbox can be determined.

[0062] The analysis of the torsional vibration error signal in different dimensions needs to be compared with different types of standard state thresholds to analyze their differences. The standard state thresholds are the corresponding thresholds obtained by analyzing or calculating the signals of various dimensions acquired by the pitch gearbox in a healthy state. In order to further explain in detail how to obtain the specific and accurate detection results of fault source and fault type and / or degree by analyzing the torsional vibration transmission error signal in different dimensions and setting the corresponding preset standard state thresholds, the following will describe in detail through the fourth to tenth embodiments.

[0063] A fourth embodiment is further proposed based on the third embodiment described above, such as... Figure 6As shown, in the fourth embodiment, feature extraction is performed on the angle transmission error signal, and it is compared with a preset standard state threshold. If the error exceeds the preset standard state threshold, it is determined that the pitch gearbox backlash is too large or the output shaft has a deep crack, including: S11. Control the pitch motor to rotate in the forward direction, calculate and record the average value of the forward second angle transmission error signal in the steady state; S12. Control the pitch motor to rotate in the opposite direction, calculate and record the average value of the reverse second angle transmission error signal in the steady state; S13. Calculate the absolute difference of the angle transmission error signal by subtracting the average value of the positive second angle transmission error signal from the average value of the negative second angle transmission error signal; S14. Compare the absolute difference with a preset standard state threshold, wherein the preset standard state threshold is the pitch gearbox backlash threshold. S15. When the absolute difference is greater than the preset standard state threshold, it is determined that the pitch gearbox backlash is too large.

[0064] The pitch gearbox backlash refers to the clearance between the teeth in the gear meshing pair. Excessive backlash is usually caused by gear wear, bearing wear, or improper assembly, which reduces transmission accuracy, leads to inaccurate pitch positioning, generates impact noise, and accelerates the wear of transmission components. Processing and analyzing the torsional vibration transmission error signal can detect and diagnose faults caused by excessive pitch gearbox backlash. This requires controlling the pitch motor to rotate forward and backward to obtain the difference in angle transmission error signals and identify the problem of excessive gearbox backlash. The pitch motor is controlled to rotate forward, and the average value of the angle transmission error signal under stable conditions is calculated and recorded, denoted as . At this point, the clearance in the transmission chain is on the side where the gears are meshing; similarly, control the pitch motor to rotate in the opposite direction, calculate and record the average value of the angle transmission error signal under steady-state conditions, denoted as . At this point, the clearance in the transmission chain is on the other side of the gear mesh. Calculate the absolute difference between the angular transmission error signals obtained from the forward and reverse tests. This difference It directly reflects the total backlash accumulated throughout the entire drivetrain of the pitch gearbox from input to output. When the monitored total backlash value... When the pitch gearbox backlash continuously exceeds the preset threshold, an excessive pitch gearbox backlash warning is triggered.

[0065] Preferably, based on the actual operating conditions of the pitch motor and the pitch gearbox, the forward and reverse rotation of the pitch motor are small-angle, low-torque rotations. Here, "small angle" refers to the normal pitch angle range, and "low torque" refers to a torque range of 1-5 kN·m. A fifth embodiment is further proposed based on the third embodiment described above, such as... Figure 7 As shown, in the fifth embodiment, the step of extracting features from the angle transmission error signal and comparing it with a preset standard state threshold, and determining whether the pitch gearbox backlash is too large or the output shaft has a deep crack when the backlash exceeds the preset standard state threshold, includes: When the torsional vibration transmission error signal is an angle transmission error signal S21. Obtain the spectrum of the angle transmission error signal based on the angle transmission error signal; S22. Monitor the amplitude of the output shaft rotational frequency harmonics in the spectrum of the angle transmission error signal; S23. Compare the output shaft frequency harmonic amplitude in the spectrum of the angle transmission error signal with a preset standard state threshold, wherein the preset standard state threshold is the harmonic amplitude threshold. S24. When the preset standard state threshold is continuously exceeded, it is determined that there is a deep crack in the output shaft.

[0066] In this embodiment, the degree of output shaft cracking is mainly analyzed based on the angle transmission error signal. Output shaft cracks often originate from the spline teeth and then gradually extend to the output shaft, resulting in uneven torsional stiffness. As the crack depth increases, the torsional stiffness asymmetry intensifies, causing a significant increase in the amplitude of the output shaft's rotational frequency (1x) and its harmonic frequencies (2x, 3x) in the angle transmission error signal. The spectrum of the angle transmission error signal is calculated. When the amplitude of the output shaft rotational frequency harmonics (1x or 2x) in the spectrum continuously exceeds a preset threshold, an output shaft crack warning is triggered, indicating the presence of a deep crack in the output shaft. A deep crack is defined as a crack area that accounts for 15% or more of the cross-sectional area of ​​the output shaft.

[0067] A sixth embodiment is further proposed based on the third embodiment described above, such as... Figure 8 As shown, in the sixth embodiment, the step of extracting features from the angular velocity transmission error signal and comparing it with a preset standard state threshold, and determining that the gears of the pitch gearbox have tooth surface damage when the error exceeds the preset standard state threshold, includes: S31. Perform spectral analysis on the angular velocity transmission error signal to obtain the angular velocity transmission error signal spectrum / order spectrum; S32. Identify significant peak frequencies or orders in the spectrum / order spectrum of the angular velocity transmission error signal; S33. Perform precise mode matching between the peak frequency or order and the theoretical fault characteristic frequencies of each stage of the transmission gears in the pitch gearbox. S34. When a spectral peak or order component that matches the theoretical fault characteristic frequency of a certain gear is identified in the spectrum of the second angular velocity transmission error signal, it is determined that the gear has tooth surface damage.

[0068] In this embodiment, the tooth surface damage and its degree of the pitch gearbox are mainly analyzed based on the angular velocity transmission error signal. The tooth surface damage of the pitch gearbox includes pitting, spalling, and tooth breakage, which will generate periodic impacts during gear meshing and will be represented by the characteristic peaks of the corresponding gears in the transmission error signal spectrum.

[0069] Transmitting error signals for angular velocity When performing spectrum analysis, under varying pitch motor speeds, the error signal is first transmitted to the angular velocity. Equal-angle resampling is performed to convert the time-domain non-stationary signal into an angular-domain stationary signal. Then, spectral analysis is conducted to obtain the angular velocity transmission error signal spectrum / order spectrum, identifying significant peak frequencies or orders. These peak frequencies or orders are then precisely pattern-matched with the theoretical fault characteristic frequencies (such as meshing frequency, sideband frequency, planetary gear passing frequency, etc.) of each stage of the pitch gearbox transmission gears (including the sun gear, planet gears, and ring gear). When a spectral peak (or order component) clearly matches the theoretical fault characteristic frequency of a certain stage of gear is clearly identified in the angular velocity transmission error signal spectrum, it can be determined that tooth surface damage exists in that stage of gear, triggering a corresponding tooth surface damage warning. The theoretical fault characteristic frequencies of the transmission gears of the pitch gearbox can be obtained from the gear parameters provided by the gear supplier.

[0070] A seventh embodiment is further proposed based on the third embodiment described above, such as... Figure 9 As shown in the seventh embodiment, the step of extracting features from the angular acceleration transmission error signal and comparing it with a preset standard state threshold, and determining that the output shaft spline of the pitch gearbox has a crack and / or the output shaft of the pitch gearbox has a deep crack when the threshold is exceeded, includes: S41. Obtain the signal spectrum of the output shaft spline of the pitch gearbox based on the angular acceleration transmission error signal; S42. Monitor the energy of the torsional resonance peak of the signal spectrum of the output shaft spline of the pitch gearbox; S43. Compare the energy of the torsional resonance peak of the signal spectrum of the output shaft spline of the pitch gearbox with the preset standard state threshold, wherein the preset standard state threshold is an energy threshold. S44. When the energy of the torsional resonance peak of the signal spectrum of the output shaft spline of the pitch gearbox continuously exceeds the preset standard state threshold, it is determined that the output shaft spline of the pitch gearbox is cracked.

[0071] The seventh embodiment is used to determine the fault type and severity of the output shaft spline of the pitch gearbox. The function of the output shaft spline of the pitch gearbox is to transmit torque. When cracks appear at the teeth of the output shaft spline of the pitch gearbox, the stability of torque transmission is weakened, resulting in a significant increase in torsional vibration energy at the torsional resonance frequency of the output shaft spline of the pitch gearbox. When the energy of the torsional resonance peak of the output shaft spline of the pitch gearbox continuously exceeds a preset threshold in the transmission error signal spectrum, a crack warning for the output shaft spline of the pitch gearbox is triggered.

[0072] like Figure 10 As shown, with the output shaft spline of the pitch gearbox intact, the pitch gearbox angular acceleration transmission error signal is measured by the torsional vibration data acquisition system of this invention. The time-domain waveform. For example... Figure 11 As shown, after a crack was confirmed in the spline of the output shaft of the same pitch gearbox, the angular acceleration transmission error signal was measured under the same operating conditions. The time-domain waveform. Figure 12 These are magnetic particle inspection photos of a spline crack on the output shaft of the pitch gearbox, clearly showing the crack located at the root of the spline tooth. (Regarding health status signals...) Spectral analysis was performed to obtain the spectrum of its angular acceleration transmission error signal, such as... Figure 14 As shown. This spectrum represents the reference torsional vibration characteristics of the output shaft spline of the pitch gearbox in a crack-free state. For fault condition signals... Perform the same spectral analysis to obtain the spectrum of its angular acceleration propagation error signal, such as Figure 14 As shown.

[0073] Will Figure 14 medium fault spectrum and Figure 13A direct comparison with the health baseline spectrum clearly reveals a significant increase in vibration energy near the torsional natural frequency of the output shaft spline of the pitch gearbox (approximately 150Hz). This energy increase at a specific torsional resonance frequency is the core feature targeted by the torsional resonance peak energy increase detection in this embodiment. Its physical root lies in the additional nonlinear friction and damping effects caused by the spline crack. In the crack region, with shaft rotation and torque alternation, the crack surface undergoes periodic opening and closing and friction, leading to a decrease in torque transmission stability. The combined effect manifests as a significant amplification of vibration energy at the torsional resonance frequency of the output shaft spline of the pitch gearbox. In practical applications, continuous online monitoring of the transmission error signal spectrum, with particular attention to the energy level of key torsional resonance peaks, such as the spline torsional resonance peak near 150Hz in this embodiment, is crucial. When the energy value of the resonance peak consistently exceeds a preset threshold based on health baseline data, the fault source can be identified. At this point, the system will automatically trigger an output shaft crack warning for the pitch gearbox, prompting maintenance personnel to conduct an inspection, thereby enabling early or developmental detection of spline cracks in the output shaft of the pitch gearbox. An early-stage spline crack in the output shaft indicates that the crack area is within 5% of the cross-sectional area of ​​the output shaft or within 10% of the spline load-bearing area.

[0074] like Figure 15 As shown, based on the third embodiment, an eighth embodiment is further proposed. In the eighth embodiment, the step of extracting features from the angular acceleration transmission error signal and comparing it with a preset standard state threshold, and determining that the output shaft spline of the pitch gearbox has a crack and / or the output shaft of the pitch gearbox has a deep crack when the preset standard state threshold is exceeded, includes: When the torsional vibration transmission error signal is an angular acceleration transmission error signal S51. Obtain the reference angular acceleration transmission error signal of the output shaft of the pitch gearbox under healthy conditions; S52. Perform a short-time Fourier transform on the reference angular acceleration transmission error signal of the output shaft of the pitch gearbox under healthy conditions to obtain a first time-frequency diagram; S53. Monitor the current angular acceleration transmission error signal under the working state of the pitch gearbox; S54. Perform a short-time Fourier transform on the current angular acceleration transmission error signal of the pitch gearbox in its working state to obtain a second time-frequency diagram; S55. Monitor the modulation amplitude of the output shaft's natural frequency in the second time-frequency diagram; S56. Compare the modulation amplitude of the output shaft's natural frequency in the second time-frequency diagram with the modulation amplitude of the output shaft's natural frequency in the first time-frequency diagram; S57. When the modulation amplitude of the output shaft natural frequency in the second time-frequency diagram exceeds the preset standard state threshold compared with the modulation amplitude of the output shaft natural frequency in the first time-frequency diagram, it is determined that the spline of the output shaft of the pitch gearbox is cracked, and the preset standard state threshold is the modulation amplitude threshold.

[0075] Reference signal for the output shaft health status of the pitch gearbox Performing a short-time Fourier transform yields the following: Figure 16 The first time-frequency diagram is shown. Observe. Figure 16 It was found that, due to the inherent nonlinear characteristics of the spline connection, even in a healthy state, the torsional natural frequency (around 175Hz) of the output shaft exhibits a certain degree of periodic frequency modulation in the time-frequency diagram. This manifests as slight, periodic fluctuations (frequency oscillations) in the spectral line of the torsional natural frequency (around 175Hz) on the time axis. However, when the output shaft is healthy, the amplitude (frequency deviation) of this frequency modulation is small, and the degree of modulation is relatively gentle. This relates to the angular velocity transmission error signal after a crack appears in the spline of the output shaft of the pitch gearbox. Performing the same short-time Fourier analysis yields the following results: Figure 17 The second time-frequency diagram is shown. Observation reveals that, compared to... Figure 16 Compared to the health status in the middle, Figure 17 Under fault conditions, the modulation of the output shaft's torsional natural frequency (around 175Hz) is significantly aggravated. This manifests as a marked increase in the amplitude of fluctuations (frequency deviation) in the characteristic frequency spectrum, along with significant amplitude modulation. The root cause of this aggravated torsional natural frequency modulation is that the spline crack disrupts the circumferential stiffness uniformity of the spline meshing region. When the output shaft rotates, the spline tooth region without cracks (or where the cracks are not engaged) provides higher constraint stiffness; the spline tooth region where the cracks are located (or where the cracks are engaged) provides significantly reduced constraint stiffness. This periodically varying constraint stiffness directly acts on the torsional vibration of the output shaft, causing its natural frequency to shift significantly periodically during rotation (i.e., strong frequency modulation). The more severe the crack, the greater the stiffness difference, and the greater the frequency modulation amplitude.

[0076] A ninth embodiment is further proposed based on the eighth embodiment described above, such as... Figure 18 As shown, in the ninth embodiment, the step of extracting features from the angular acceleration transmission error signal and comparing it with a preset standard state threshold, and determining that the output shaft spline of the pitch gearbox has a crack and / or the output shaft of the pitch gearbox has a deep crack when the threshold is exceeded, includes: When the torsional vibration transmission error signal is an angular acceleration transmission error signal S61. Obtain the reference angular acceleration transmission error signal of the output shaft of the pitch gearbox under healthy conditions. S62. Perform spectral coherence analysis on the reference angular acceleration transmission error signal to obtain a first spectral coherence diagram; S63. Monitor the current angular acceleration transmission error signal under the working state of the pitch gearbox; S64. Perform spectral coherence analysis on the current angular acceleration transmission error signal to obtain a second spectral coherence map; S65. Calculate the number of output shaft rotational frequency harmonics near the torsional natural frequency of the output shaft in the second spectral coherence diagram.

[0077] S66. Compare the number of output shaft frequency harmonics with the harmonic number threshold; S67. When the number of output shaft frequency harmonics monitored continuously exceeds the health baseline data, it is judged as an early spline crack.

[0078] The ninth embodiment is used to quantize the intensity of the frequency modulation in the eighth embodiment and to transmit an error signal to the reference. Transmitting error signals under fault conditions Spectral coherence analysis was performed separately. Figure 19 This indicates the spline in a healthy state. The spectral coherence analysis results show that only weak spectral coherence is observed near the torsional natural frequency of the output shaft. Near the natural frequency, the number and amplitude of harmonics in the output shaft's rotational frequency are low, indicating that the torsional natural frequency modulation is weak under healthy conditions. Figure 20 This indicates the state of spline cracking. The spectral coherence analysis results show a significantly enhanced spectral coherence near the torsional natural frequency of the output shaft. Abundant high-order harmonic components (such as 3x, 4x, 5x, 6x...) of the output shaft rotational frequency appear near the natural frequency, reflecting a strong periodic modulation phenomenon of the torsional natural frequency under spline crack conditions in the output shaft of the pitch gearbox. Based on the above analysis, the number of harmonics related to the output shaft rotational frequency (1x) appearing near the torsional natural frequency in the spectral coherence analysis diagram, or their total energy, can serve as an effective indicator for quantifying the degree of natural frequency modulation. A greater number of sidebands, higher order, and stronger energy indicate more severe torsional natural frequency modulation. In practical applications, this is crucial for real-time acquisition of speed transmission error signals. Online spectral coherence analysis is performed to calculate the number of output shaft rotational frequency harmonics near the output shaft's torsional natural frequency. When the number of monitored output shaft rotational frequency harmonics continuously exceeds a threshold set based on health baseline data, the trigger condition for intensified torsional resonance peak modulation is met. At this point, the system will automatically trigger an early spline crack warning, specifically indicating the presence of damage leading to circumferential unevenness in meshing stiffness. An early spline crack can be defined as within 5% of the output shaft's cross-sectional area.

[0079] A tenth embodiment is further proposed based on the third embodiment, such as... Figure 21 As shown, in the tenth embodiment, the step of extracting features from the angular acceleration transmission error signal and comparing it with a preset standard state threshold, and determining that the output shaft spline of the pitch gearbox has a crack and / or the output shaft of the pitch gearbox has a deep crack when the preset standard state threshold is exceeded, includes: When the torsional vibration transmission error signal is an angular acceleration transmission error signal S71. Obtain the signal spectrum of the output shaft of the pitch gearbox based on the angular acceleration transmission error signal; S72. Identify and track the first-order torsional natural frequency of the signal spectrum of the output shaft.

[0080] S73. The offset of the first-order torsional natural frequency towards the lower frequency direction is compared with the preset standard state threshold, wherein the preset standard state threshold is the offset threshold. S74. When the first-order torsional natural frequency is detected to have shifted to a lower frequency direction for an extended period exceeding a preset standard state threshold, it is determined that there is a deep crack in the output shaft.

[0081] Output shaft cracks can cause a decrease in the overall torsional stiffness of the output shaft. According to the principles of vibration dynamics, the torsional natural frequency of the output shaft will drift to a lower frequency. The spectrum of the angular acceleration transmission error signal is calculated to identify and track the first-order torsional natural frequency of the pitch output shaft. When the shift of the natural frequency to a lower frequency direction is continuously detected to exceed the preset threshold, a deep crack in the output shaft can be identified, triggering a crack warning and indicating that a deep crack exists in the output shaft.

[0082] The eleventh embodiment provides a wind turbine pitch gearbox fault detection system, such as... Figure 22 As shown, the system includes: The first sensing module 1 acquires the first angle signal at the input end of the pitch motor; The second sensing module 2 acquires the second angle signal generated when the output shaft of the pitch gearbox is driven by the pitch motor. Signal distribution module 3 receives the first angle signal output by the first sensing module and generates an execution feedback signal and a sample angle signal based on the first angle signal; Torsional vibration signal acquisition module 4 synchronously receives the sample angle signal and the second angle signal; and obtains the torsional vibration transmission error signal of the pitch gearbox based on the sample angle signal and the second angle signal. The fault analysis module 5 extracts features from the torsional vibration transmission error signal, compares it with a preset standard state threshold, obtains the torsional vibration transmission error signal analysis result, and determines whether the pitch gearbox has a fault based on the torsional vibration transmission error signal analysis result.

[0083] The signal distribution module 3 generates an execution feedback signal based on the first angle signal and sends it to the pitch controller for closed-loop control of the pitch motor.

[0084] The wind turbine pitch gearbox fault detection system of this embodiment is used to execute the signal acquisition method and fault detection method in the above embodiments, which will not be repeated here. The above description is only a preferred embodiment of the embodiments of this application, and does not limit the scope of disclosure of the embodiments of this application. Any equivalent structural or procedural transformations made based on the description and drawings of the embodiments of this application, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection supported by the embodiments of this application.

Claims

1. A method for collecting signals of a pitch gear box of a wind turbine generator unit, characterized in that, The method comprises: collecting a first angle signal at the input end of the pitch motor, and generating a sample angle signal according to the first angle signal; collecting a second angle signal generated when the output shaft of the pitch gear box is driven by the pitch motor; synchronously receiving the sample angle signal and the second angle signal; obtaining a torsional vibration transmission error signal of the pitch gear box according to the sample angle signal and the second angle signal.

2. The wind turbine variable pitch gearbox signal acquisition method of claim 1, wherein, The first angle signal is a raw differential analog signal or a raw pulse train signal collected from the input end of the pitch motor.

3. The wind turbine variable pitch gearbox signal acquisition method of claim 2, wherein: The generation of the sample angle signal according to the first angle signal comprises: distributing the raw differential analog signal to generate an execution feedback signal and the sample angle signal; The execution feedback signal is sent to the pitch driver, and the pitch motor is controlled in real time according to the execution feedback signal.

4. The wind turbine variable pitch gearbox signal acquisition method of claim 1, wherein, The collection of the second angle signal generated when the output shaft of the pitch gear box is driven by the pitch motor comprises: installing an output shaft encoder on the fan hub through a mounting bracket, and connecting the output shaft encoder to the process hole of the pitch output gear through a connecting tool; The second angle signal is obtained through the output shaft encoder.

5. The wind turbine variable pitch gearbox signal acquisition method of any of claims 1-4, wherein, The obtaining of the torsional vibration transmission error signal of the pitch gear box according to the sample angle signal and the second angle signal further comprises: performing first-order differential operation on the sample angle signal to obtain a first instantaneous angular velocity signal; performing first-order differential operation on the second angle signal to obtain a second instantaneous angular velocity signal; obtaining the torsional vibration transmission error signal according to the first instantaneous angular velocity signal and the second instantaneous angular velocity signal.

6. The wind turbine variable pitch gearbox signal acquisition method of any of claims 1-4, wherein, The obtaining of the torsional vibration transmission error signal of the pitch gear box according to the sample angle signal and the second angle signal further comprises: performing second-order differential operation on the sample angle signal to obtain a first instantaneous angular acceleration signal; performing second-order differential operation on the second angle signal to obtain a second instantaneous angular acceleration signal; obtaining the torsional vibration transmission error signal according to the first instantaneous angular acceleration signal and the second instantaneous angular acceleration signal.

7. The wind turbine variable pitch gearbox signal acquisition method of any of claims 1-4, wherein, The torsional vibration transmission error signal comprises an angle transmission error signal, which is the difference between the angle signal obtained by converting the second angle signal according to the speed ratio and the sample angle signal.

8. The wind turbine variable pitch gearbox torsional vibration transmission error signal acquisition method of claim 5, wherein, The torsional vibration transmission error signal comprises an angular velocity transmission error signal, which is the difference between the angular velocity signal obtained by converting the second instantaneous angular velocity signal according to the speed ratio and the first instantaneous angular velocity signal.

9. The wind turbine variable pitch gearbox signal acquisition method of claim 6, wherein, The torsional vibration transmission error signal comprises an angular acceleration transmission error signal, which is the difference between the angular acceleration signal obtained by converting the second instantaneous angular acceleration signal according to the speed ratio and the first instantaneous angular acceleration signal.

10. A method of wind turbine generator unit pitch gear box fault detection, characterized by, The method comprises: obtaining a torsional vibration transmission error signal based on the wind turbine pitch gear box torsional vibration transmission error signal acquisition method according to any one of claims 1-9; According to the torsional vibration transmission error signal, the torsional vibration transmission error signal is extracted, compared with the preset standard state threshold, and whether the pitch gear box has a fault is obtained.

11. The wind turbine variable pitch gearbox fault detection method of claim 10, wherein: the torsional vibration transmission error signal comprises an angle transmission error signal, an angular velocity transmission error signal, and / or an angular acceleration transmission error signal; the obtaining of the variable pitch gearbox fault comprises: feature extraction of the angle transmission error signal, and comparison with a preset standard state threshold value, and when the preset standard state threshold value is exceeded, it is determined that the variable pitch gearbox side clearance is too large or that the output shaft of the variable pitch gearbox has a deep crack; or, feature extraction of the angular velocity transmission error signal, and comparison with a preset standard state threshold value, and when the preset standard state threshold value is exceeded, it is determined that the gear of the variable pitch gearbox has a gear surface damage; or, feature extraction of the angular acceleration transmission error signal, and comparison with a preset standard state threshold value, and when the preset standard state threshold value is exceeded, it is determined that the output shaft spline of the variable pitch gearbox has a crack and / or that the output shaft of the variable pitch gearbox has a deep crack.

12. The wind turbine generator variable pitch gearbox fault detection method of claim 11, wherein, the feature extraction of the angle transmission error signal and comparison with a preset standard state threshold value, and when the preset standard state threshold value is exceeded, it is determined that the variable pitch gearbox side clearance is too large, comprises: controlling the variable pitch motor to rotate in a forward direction, calculating and recording the average value of the forward angle transmission error signal in a stable state; controlling the variable pitch motor to rotate in a reverse direction, calculating and recording the average value of the reverse angle transmission error signal in a stable state; calculating the absolute difference value of the angle transmission error signal by subtracting the average value of the reverse angle transmission error signal from the average value of the forward angle transmission error signal; comparing the absolute difference value with a preset standard state threshold value, and the preset standard state threshold value is a variable pitch gearbox side clearance threshold value; when the absolute difference value is greater than the preset standard state threshold value, it is determined that the variable pitch gearbox side clearance is too large.

13. The wind turbine generator variable pitch gearbox fault detection method of claim 12, wherein, the forward rotation and reverse rotation of the variable pitch motor are small angle and low torque rotation.

14. The wind turbine generator variable pitch gearbox fault detection method of claim 11, wherein, the feature extraction of the angle transmission error signal and comparison with a preset standard state threshold value, and when the preset standard state threshold value is exceeded, it is determined that the output shaft of the variable pitch gearbox has a deep crack, comprises: obtaining the angle transmission error signal frequency spectrum according to the angle transmission error signal; monitoring the output shaft rotation frequency harmonic amplitude in the angle transmission error signal frequency spectrum; comparing the output shaft rotation frequency harmonic amplitude in the angle transmission error signal frequency spectrum with a preset standard state threshold value, and the preset standard state threshold value is a harmonic amplitude threshold value; when the preset standard state threshold value is continuously exceeded, it is determined that the output shaft crack of the variable pitch gearbox has a deep crack.

15. The wind turbine generator variable pitch gearbox fault detection method of claim 11, wherein, the feature extraction of the angular velocity transmission error signal and comparison with a preset standard state threshold value, and when the preset standard state threshold value is exceeded, it is determined that the gear of the variable pitch gearbox has a gear surface damage, comprises: spectrum analysis of the angular velocity transmission error signal to obtain an angular velocity transmission error signal frequency spectrum / order spectrum; identifying a significant peak frequency or order in the angular velocity transmission error signal frequency spectrum / order spectrum; pattern matching of the peak frequency or order with the theoretical fault characteristic frequency of each level of the variable pitch gearbox transmission gear; When a spectral peak or order component matching a theoretical fault characteristic frequency of a gear of a certain stage is identified in the second angular velocity transmission error signal spectrum, it is determined that the gear of the certain stage of the variable pitch gear box has gear surface damage.

16. The wind turbine generator variable pitch gearbox fault detection method of claim 11, wherein, The feature extraction of the angular acceleration transmission error signal and comparison with the preset standard state threshold value, when exceeding the preset standard state threshold value, determine that the output shaft spline of the variable pitch gear box has early cracks include: According to the angular acceleration transmission error signal, a signal spectrum of the output shaft spline of the variable pitch gear box is obtained; The energy of the torsional resonance peak of the signal spectrum of the output shaft spline of the variable pitch gear box is monitored; The energy of the torsional resonance peak of the signal spectrum of the output shaft spline of the variable pitch gear box is compared with the preset standard state threshold value, and the preset standard state threshold value is an energy threshold value; When the energy of the torsional resonance peak of the signal spectrum of the output shaft spline of the variable pitch gear box continuously exceeds the preset standard state threshold value, it is determined that the output shaft spline of the variable pitch gear box has cracks.

17. The wind turbine generator variable pitch gearbox fault detection method of claim 11, wherein, The feature extraction of the angular acceleration transmission error signal and comparison with the preset standard state threshold value, when exceeding the preset standard state threshold value, determine that the output shaft spline of the variable pitch gear box has cracks include: A reference angular acceleration transmission error signal under the health state of the output shaft of the variable pitch gear box is obtained; A first time-frequency graph is obtained by performing short-time Fourier transform on the reference angular acceleration transmission error signal under the health state of the output shaft of the variable pitch gear box; A current angular acceleration transmission error signal under the working state of the variable pitch gear box is monitored; A second time-frequency graph is obtained by performing short-time Fourier transform on the current angular acceleration transmission error signal under the working state of the variable pitch gear box; The modulation amplitude of the inherent frequency of the output shaft in the second time-frequency graph is monitored; The modulation amplitude of the inherent frequency of the output shaft in the second time-frequency graph is compared with the modulation amplitude of the inherent frequency of the output shaft in the first time-frequency graph; When the modulation amplitude of the inherent frequency of the output shaft in the second time-frequency graph exceeds the modulation amplitude of the inherent frequency of the output shaft in the first time-frequency graph by more than the preset standard state threshold value, it is determined that the output shaft spline of the variable pitch gear box has cracks, and the preset standard state threshold value is a modulation amplitude threshold value.

18. The wind turbine generator variable pitch gearbox fault detection method of claim 11, wherein, The feature extraction of the angular acceleration transmission error signal and comparison with the preset standard state threshold value, when exceeding the preset standard state threshold value, determine that the output shaft spline of the variable pitch gear box has cracks include: A reference angular acceleration transmission error signal under the health state of the output shaft of the variable pitch gear box is obtained; A first spectral coherence graph is obtained by performing spectral coherence analysis on the reference angular acceleration transmission error signal; A current angular acceleration transmission error signal under the working state of the variable pitch gear box is monitored; A second spectral coherence graph is obtained by performing spectral coherence analysis on the current angular acceleration transmission error signal; The number of output shaft rotation frequency harmonics near the inherent frequency of the torsional vibration of the output shaft in the second spectral coherence graph is calculated; The number of output shaft rotation frequency harmonics is compared with the harmonic number threshold value; When the number of monitored output shaft rotation frequency harmonics continuously exceeds the health benchmark data, it is determined that the output shaft spline of the pitch gear box has a medium-early stage crack.

19. The wind turbine generator variable pitch gearbox fault detection method of claim 11, wherein, The feature extraction of the angular acceleration transmission error signal is compared with the preset standard state threshold, and when the preset standard state threshold is exceeded, it is determined that the output shaft of the pitch gear box has a deep crack, which includes: Obtaining a signal frequency spectrum according to the angular acceleration transmission error signal; Identifying and tracking the first-order torsional natural frequency in the signal frequency spectrum; Comparing the shift amount of the first-order torsional natural frequency in the signal frequency spectrum to the low frequency direction with the preset standard state threshold, which is a shift amount threshold; When the shift amount of the first-order torsional natural frequency in the signal frequency spectrum to the low frequency direction is continuously monitored to exceed the preset standard state threshold, it is determined that the output shaft of the pitch gear box has a deep crack.

20. A wind turbine generator variable pitch gearbox fault detection system characterized by, It includes: The first sensing module obtains the first angle signal of the input end of the pitch motor; The second sensing module obtains the second angle signal generated by the output shaft of the pitch gear box when driven by the pitch motor; The signal distribution module receives the first angle signal output by the first sensing module, and generates an execution feedback signal and a sample angle signal according to the first angle signal; The torsional vibration signal acquisition module synchronously receives the sample angle signal and the second angle signal; according to the sample angle signal and the second angle signal, the torsional vibration transmission error of the pitch gear box is obtained; the fault analysis module extracts the features of the torsional vibration transmission error signal, compares it with the preset standard state threshold, obtains the torsional vibration transmission error signal analysis result, and determines whether the pitch gear box has a fault according to the torsional vibration transmission error signal analysis result.

Images