Self-adaptive MSB planetary gear train fault diagnosis method based on meshing period characteristics

Abstract

The invention relates to the technical field of wind driven generator gearbox fault diagnosis and vibration signal processing, and particularly discloses a meshing period characteristic-based adaptive MSB planetary gear train fault diagnosis method, which comprises the following steps of: S1, acquiring an original vibration signal and a rotating speed pulse signal in an operation process of a planetary gearbox, resampling the original vibration signal, and acquiring a rotating speed pulse signal; a vibration signal after resampling is obtained; s2, calculating a reset period of the planetary gearbox, and segmenting the resampling signal according to the reset period to obtain a plurality of reset period signal segments; s3, calculating the sparse feature change trend of the separated fragment signals by using the GI index, and extracting the signal fragment with the maximum GI index as a fault signal fragment; and S4, applying the MSB to the fault signal fragment matrix to extract fault features. According to the method, the early fault information contained in the vibration signals can be identified, the MSB fragments are adaptively selected, weakening of weak fault characteristic signals to the signals in the averaging process is avoided, and the fault characteristic frequency is effectively extracted.

Description

Technical Field

[0001] The present invention relates to the technical fields of fault diagnosis of wind turbine gearboxes and vibration signal processing, and specifically provides an adaptive MSB planetary gear train fault diagnosis method based on meshing period characteristics. Background Art

[0002] As a high-efficiency, compact, low-noise, long-life and high-load-capacity transmission device, the planetary gearbox plays a crucial role in wind turbines with its unique structure and superior performance. The wind turbine gearbox needs to operate for a long time under high wind speeds, heavy loads and complex working conditions, and the compact design and high-efficiency transmission characteristics of the planetary gearbox make it an indispensable core component in the wind power generation system. However, during use, the planetary gearbox is prone to faults such as pitting and cracks, which will lead to a decline in the performance of the gearbox. In severe cases, it may cause equipment damage or even shutdown, resulting in huge economic losses. Therefore, timely detection of the fault conditions of the planetary gearbox is of great significance for improving the reliability, operation stability and economic benefits of wind turbines.

[0003] Traditional fault diagnosis methods have obvious limitations when facing planetary gearboxes. Different from fixed-axis gear trains, the continuous relative motion between the gears in the planetary gear train causes the meshing position and the vibration signal transmission path to change continuously. This time-variation brings great challenges to the extraction and diagnosis of fault characteristics. The vibration signals collected by sensors are often modulated by complex transmission paths, masking the true fault information and making accurate diagnosis difficult to achieve.

[0004] In order to improve the accuracy of planetary gearbox fault diagnosis, it is necessary to weaken the influence of the time-varying transmission path on fault vibrations. Therefore, it is necessary to design a planetary gear train fault diagnosis method to avoid the additional modulation influence of the time-varying transmission path on vibration signals, effectively extract the fault characteristic frequencies, and improve the accuracy in fault diagnosis. Summary of the Invention

[0005] To solve the above problems, the present invention provides an adaptive MSB planetary gear train fault diagnosis method based on meshing period characteristics, which can adaptively select MSB segments, avoid the weakening of signals by weak fault characteristic signals during averaging, effectively extract the fault characteristic frequencies, and improve the accuracy in fault diagnosis.

[0006] An adaptive MSB planetary gear train fault diagnosis method based on meshing period characteristics provided by the present invention includes the following steps: S1. Obtain the original vibration signal and the rotational speed pulse signal during the operation of the planetary gearbox, resample the original vibration signal through angle resampling, and obtain the resampled vibration signal; S2. Calculate the reset period of the planetary gearbox according to the meshing period characteristics of the planetary gear train, segment the resampled signal according to the reset period, and obtain several reset period signal segments; S3. Use the GI index to calculate the change trend of the sparse characteristics of the separated segment signals, and extract the signal segment with the largest GI index as the fault signal segment; S4. Apply the MSB to the fault signal segment matrix, suppress the participation noise and demodulation coupling frequency in the signal, and extract the fault characteristics.

[0007] Further, the calculation formula for the reset period of the planetary gear train in step S2 is as follows:

[0008] Among them, LCM represents the least common multiple, represents the number of teeth of the ring gear, represents the number of teeth of the faulty planetary gear or the sun gear.

[0009] Further, in step S3, the selection method of the fault signal segment is as follows: S31. Segment the separated different reset period signal segments through a time-varying sliding window; S32. Calculate the change trend of the sparse characteristics of different signal segments through the GI index; S33. Select the signal segment with the largest sparse characteristic as the fault signal segment for separation, and construct a fault signal segment matrix.

[0010] Further, the calculation formula for the GI index in step S32 is as follows:

[0011] Among them, is the ascending order of ; ; represents of norm; N represents the signal x total length of, n represents the index position of the current summation.

[0012] Further, step S4 is implemented through the following formula:

[0013] Among them, is the set average value of the MSB matrix obtained from multiple modulated signals; is complex conjugate of; is the carrier frequency; is the modulation frequency; is Upper and lower sideband frequencies; Is the bispectrum of the modulation signal of the signal.

[0014] Further, in step S1, the original vibration signal is obtained by an acceleration sensor installed on the planetary gearbox housing, and the rotational speed pulse signal is obtained by a photoelectric sensor on an encoder installed on the shaft.

[0015] Further, in step S1, when resampling, the resampling frequency is set to the number of points rotated by the planet carrier per revolution.

[0016] Further, in step S1, a simulation signal model is used to simulate the vibration signal when a planetary gearbox actually fails. Specifically, an amplitude-modulated and frequency-modulated signal is used to simulate the gear meshing modulation signal, a Hanning window is used to simulate the time-varying transmission path of the planetary gear train vibration, and a Gaussian white noise signal with a signal-to-noise ratio of 10 dB is added; the simulation signal model is:

[0017] Among them, Represents the original vibration signal, Represents the meshing amplitude-modulated and frequency-modulated signal, Represents the Hanning window, Represents the Gaussian white noise signal with a signal-to-noise ratio of 10 dB.

[0018] Further, the gear meshing amplitude-modulated and frequency-modulated signal is expressed as:

[0019] Among them, > 0, > 0 respectively represent the amplitude sizes of amplitude modulation and frequency modulation; Represents the fault characteristic frequency of the gear, t represents time, Represents the meshing frequency of the gear, , And Are respectively the initial phase of amplitude modulation, the initial phase of frequency modulation, and the phase offset of the carrier.

[0020] Further, the Hanning window signal is expressed as:

[0021] Among them, Represents the length of a single window function, Represents the time-domain position of the window function.

[0022] Compared with the prior art, the present invention can achieve the following beneficial effects: 1. The present invention aims to identify early fault information contained in vibration signals. According to the kinematic law of planetary gear train meshing, an adaptive MSB fault diagnosis method for planetary gear trains based on meshing period characteristics is specifically proposed. Compared with the prior art, the present invention combines the meshing period characteristics of the planetary gear train, separates the resampled signals according to the reset period, and makes the vibration signals within different reset periods correspond one by one to the meshing time sequence relationship of the planetary gear train, so as to separate the vibration signal segment closest to the distance sensor, avoiding the additional modulation effect of the time-varying transmission path on the vibration signal.

[0023] 2. The present invention uses the fault signal segment closest to the distance sensor for the MSB method, solves the problem that the MSB does not consider the influence of the time-varying transmission path on the vibration signal during signal processing, avoids the reduction of the feature enhancement effect during the averaging process of the MSB, and improves the accuracy of the MSB method in fault diagnosis. Description of the Drawings

[0024] Figure 1 is the overall flowchart of the planetary gear train fault diagnosis method provided by the embodiment of the present invention; Figure 2 is the time-domain diagram of the simulated vibration signal in the planetary gear train fault diagnosis method provided by the embodiment of the present invention; Figure 3 is the frequency spectrum diagram of the simulated vibration signal in the planetary gear train fault diagnosis method provided by the embodiment of the present invention; Figure 4 is the variation law diagram of the GI index feature of the vibration signal in different reset periods in the time domain in the planetary gear train fault diagnosis method provided by the embodiment of the present invention; Figure 5 is the MSB diagram of the same region selecting the maximum average GI index in different reset periods in the planetary gear train fault diagnosis method provided by the embodiment of the present invention; Figure 6 is the MSB diagram of the same region selecting the minimum average GI index in different reset periods in the planetary gear train fault diagnosis method provided by the embodiment of the present invention. Detailed Embodiment

[0025] In order to make the objectives, technical solutions and advantages of the present invention clearer, the following further elaborates on the present invention with reference to the Figure 1-6 accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and do not constitute a limitation to the present invention.

[0026] An adaptive MSB fault diagnosis method for planetary gear trains based on meshing period characteristics, as Figure 1 shown, includes the following steps: S1. Obtain the original vibration signal and rotational speed pulse signal during the operation of the planetary gearbox. The original vibration signal is obtained by the acceleration sensor installed on the housing of the planetary gearbox, and the rotational speed pulse signal is obtained by the photoelectric sensor on the encoder installed on the shaft. In this example, the number of teeth of the sun gear in the planetary gear train , the number of teeth of the planetary gear , and the number of teeth of the ring gear .

[0027] In step S1, a simulation signal model is used to simulate the vibration signal when the planetary gearbox actually fails. Specifically, the amplitude-modulated and frequency-modulated signal is used to simulate the gear meshing modulation signal, the Hanning window is used to simulate the time-varying transmission path of the vibration of the planetary gear train, and the Gaussian white noise signal with a signal-to-noise ratio of 10 dB is added. The simulation signal model is:

[0028] Among them, represents the original vibration signal, represents the meshing amplitude-modulated and frequency-modulated signal, represents the Hanning window, represents the Gaussian white noise signal with a signal-to-noise ratio of 10 dB.

[0029] The gear meshing amplitude-modulated and frequency-modulated signal is expressed as:

[0030] Among them, > 0, > 0 respectively represent the amplitude sizes of amplitude modulation and frequency modulation; represents the fault characteristic frequency of the gear, t represents time, represents the meshing frequency of the gear, , and are the initial phases of amplitude modulation, frequency modulation and the phase offset of the carrier respectively.

[0031] The Hanning window signal is expressed as:

[0032] Among them, represents the length of a single window function, represents the time-domain position of the window function.

[0033] The time-domain diagram of the simulated vibration signal in the planetary gear train fault diagnosis method is as shown in Figure 2 , and the frequency-spectrum diagram of the simulated vibration signal in the planetary gear train fault diagnosis method is as shown in Figure 3 .

[0034] In step S1, the original vibration signal needs to be resampled by angle resampling to obtain the resampled vibration signal. In this example, gear meshing refers to only considering the meshing process between a single planetary gear and the sun gear in the planetary gear train. When performing resampling, the resampling frequency is set to the ratio of the number of acquisition points per revolution of the planet carrier to the time used, so as to ensure that the amount of data collected by the sensor is the same for each revolution of the planet carrier.

[0035] S2. Calculate the reset period of the planetary gearbox according to the meshing period characteristics of the planetary gear train. The calculation formula for the reset period of the planetary gear train is as follows:

[0036] where LCM represents the least common multiple, represents the number of teeth of the ring gear, represents the number of teeth of the faulty planetary gear or the sun gear.

[0037] , that is, the reset period of this planetary gearbox is. where represents the rotation period of a planet carrier.

[0038] Then segment the resampled signal according to the reset period to obtain several reset period signal segments.

[0039] S3. Use the GI index to calculate the change trend of the sparse characteristics of the separated segment signals, and extract the signal segment with the largest GI index as the fault signal segment. The change law diagram of the GI index characteristics of different reset period signals in the time-domain vibration signal in the planetary gear train fault diagnosis method is as Figure 4 shown.

[0040] The selection method of the fault signal segment is as follows: S31. Segment the separated different reset period signal segments through a time-varying sliding window.

[0041] S32. Calculate the change trend of the sparse characteristics of different signal segments through the GI index; where the calculation formula for the GI index is as follows:

[0042] where, is sorted in ascending order, ; represents of norm; N represents the signal x total length of, n represents the index position of the current summation.

[0043] S33. Select the signal segment with the largest sparse feature as the fault signal segment for separation, and construct a fault signal segment matrix.

[0044] S4. Apply the MSB to the fault signal segment matrix to suppress the participation noise and demodulation coupling frequency in the signal, and extract the fault features.

[0045] Step S4 is implemented by the following formula:

[0046] where, is the set average value of the MSB matrix obtained from multiple segments of modulated signals; is 's complex conjugate; is the carrier frequency; is the modulation frequency; are the upper and lower sideband frequencies; is the modulation signal bispectrum of the signal.

[0047] The MSB diagrams of the same region with the maximum average GI index in different reset periods in the planetary gear train fault diagnosis method are as Figure 5 shown, and the MSB diagrams of the same region with the minimum average GI index in different reset periods in the planetary gear train fault diagnosis method are as Figure 6 shown.

[0048] Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. An adaptive MSB planetary gear train fault diagnosis method based on meshing period characteristics, characterized in that: The steps include: S1. Acquire the original vibration signal and the speed pulse signal during the operation of the planetary gearbox, resample the original vibration signal by angle resampling, and obtain the resampled vibration signal; S2, calculating the reset period of the planetary gearbox according to the meshing period characteristics of the planetary gear train, segmenting the resampled signal according to the reset period, and obtaining a plurality of reset period signal segments; S3, using the GI index to calculate the sparse feature change trend of the separated segment signal, and extracting the signal segment with the largest GI index as the fault signal segment; S4. Apply MSB to the fault signal fragment matrix to suppress the participating noise and demodulate the coupled frequency in the signal and extract the fault features.

2. The adaptive MSB planetary gear train fault diagnosis method based on meshing period characteristics according to claim 1, characterized in that: The calculation formula for the reset period of the planetary gear train in step S2 is as follows: ； Among them, LCM means the least common multiple, Indicates the number of teeth on the ring gear. Indicates the number of faulty planetary gear or sun gear teeth.

3. The adaptive MSB planetary gear train fault diagnosis method based on meshing period characteristics according to claim 1, characterized in that: In step S3, the fault signal segment is selected as follows: S31, segmenting the separated signal fragments with different reset periods through a time-varying sliding window; S32, calculating the sparse feature change trend of different signal segments through the GI index; S33, selecting the signal segment with the largest sparse feature as the fault signal segment for separation, and constructing a fault signal segment matrix.

4. The adaptive MSB planetary gear train fault diagnosis method based on meshing period characteristics according to claim 3 is characterized in that: The calculation formula of the GI index in step S32 is as follows: ； in, Yes In ascending order, ; express of norm; N Indicates signal x The total length of n Indicates the index position of the current summation.

5. The adaptive MSB planetary gear train fault diagnosis method based on meshing period characteristics according to claim 1, characterized in that: Step S4 is implemented by the following formula: ； in, To obtain the collective average of the MSB matrix from multiple modulation signals; for The complex conjugate of is the carrier frequency; is the modulation frequency; are the upper and lower sideband frequencies; is the bispectrum of the modulation signal.

6. The adaptive MSB planetary gear train fault diagnosis method based on meshing period characteristics according to claim 1, characterized in that: In step S1, the original vibration signal is obtained by the acceleration sensor installed on the planetary gearbox housing, and the speed pulse signal is obtained by the photoelectric sensor on the encoder installed on the shaft.

7. The adaptive MSB planetary gear train fault diagnosis method based on meshing period characteristics according to claim 1, characterized in that: In step S1, when resampling is performed, the resampling frequency is set to the number of points per rotation of the planet carrier.

8. The adaptive MSB planetary gear train fault diagnosis method based on meshing period characteristics according to claim 7, characterized in that: In step S1, a simulation signal model is used to simulate the vibration signal of the planetary gearbox when a fault actually occurs. Specifically, the gear meshing modulation signal is simulated by an amplitude-frequency modulation signal, the time-varying transmission path of the planetary gear train vibration is simulated by a Hanning window, and a Gaussian white noise signal with a signal-to-noise ratio of 10 dB is added together; the simulation signal model is: ； in, represents the original vibration signal, Indicates the meshing AM / FM signal, represents the Hanning window, Represents a Gaussian white noise signal with a signal-to-noise ratio of 10dB.

9. The adaptive MSB planetary gear train fault diagnosis method based on meshing period characteristics according to claim 8, characterized in that: The gear meshing amplitude and frequency modulation signal is expressed as: ； in, >0, >0 indicates the amplitude of amplitude modulation and frequency modulation respectively; represents the fault characteristic frequency of the gear, t represents the time, represents the meshing frequency of the gears, , and They are the initial phase of amplitude modulation, frequency modulation and the phase offset of the carrier respectively.

10. The adaptive MSB planetary gear train fault diagnosis method based on meshing period characteristics according to claim 9, characterized in that: The Hanning window signal is expressed as: ； in, Represents the length of a single window function, Represents the time domain position of the window function.

Images