Planetary gearbox fault diagnosis method based on time-varying modulation feature extraction

Abstract

The invention discloses a planetary gearbox fault diagnosis method based on time-varying modulation feature extraction. The method comprises the following steps: separating amplitude-modulated and frequency-modulated signals from original vibration signals through Hilbert transform; respectively generating time-frequency diagrams of amplitude-modulated and frequency-modulated signals by adopting multi-order synchronous compression transformation; extracting a fault related ridge line by using an adaptive multi-ridge line extraction algorithm; reconstructing the signal and carrying out non-uniform resampling in a time domain by using time warping; and analyzing the frequency spectrum of the amplitude-modulated and frequency-modulated signal after resampling, and confirming the specific fault type of the planetary gearbox under the working condition of the time-varying rotating speed. According to the method, the time-varying modulation characteristics of amplitude modulation and frequency modulation components in original signals can be extracted respectively, non-uniform resampling is carried out on time-varying signals in a time domain to eliminate interference of rotating speed fluctuation, and extraction of fault characteristics and fault diagnosis of the planetary gearbox under the time-varying rotating speed working condition are achieved.

Description

Technical Field

[0001] The invention relates to a rotating machinery fault diagnosis technology method under time-varying speed conditions, and in particular to a planetary gearbox fault diagnosis method based on time-varying modulation feature extraction. Background Art

[0002] Planetary gearboxes are the core transmission equipment in various complex mechanical systems. Due to their compact structure, high transmission ratio and good stability, planetary gearboxes are widely used in automobiles, ships, wind power generation, aerospace and other fields. However, planetary gearboxes are usually subject to severe operating conditions such as severe impact, heavy load and time-varying speed during operation, so the internal mechanical components (i.e. sun gear, planetary gear and ring gear) will inevitably suffer from wear and broken teeth. These local damages of the planetary gearbox may lead to the performance degradation of the entire mechanical system or even serious accidents.

[0003] Therefore, studying effective feature extraction and fault diagnosis methods is crucial to detect gear faults in planetary gearboxes and ensure their reliable operation.

[0004] Vibration signal processing is one of the most widely used technologies in planetary gearbox fault diagnosis. When a gear is partially damaged, pulse excitation will be excited into the vibration system, and the fault characteristic frequency will appear in the Fourier spectrum of the vibration signal. However, due to the complex internal structure of the planetary gearbox and the variable vibration transmission path, the time-varying vibration signal contains complex frequency sidebands and exhibits obvious amplitude modulation and frequency modulation characteristics. It is difficult to separate the amplitude modulation and frequency modulation components and extract the complete fault characteristics. In addition, the interference of speed fluctuations will cause spectrum tailing in the frequency domain, thereby interfering with the extraction of fault characteristic frequencies. Summary of the invention

[0005] Purpose of the invention: The purpose of the present invention is to provide a planetary gearbox fault diagnosis method based on time-varying modulation feature extraction.

[0006] Technical solution: The planetary gearbox fault diagnosis method based on time-varying modulation feature extraction of the present invention comprises the following steps:

[0007] Step S1, using Hilbert transform to separate the amplitude modulation and frequency modulation signals from the original vibration signal;

[0008] Step S2, using multi-order synchronous compression transformation to generate time-frequency diagrams of amplitude modulation and frequency modulation signals respectively;

[0009] Step S3, extracting fault-related ridges using an adaptive multi-ridge extraction algorithm;

[0010] Step S4, reconstructing the signal and performing non-uniform resampling in the time domain using time warping;

[0011] Step S5, analyzing the resampled frequency spectrum of the AM / FM signal to confirm the specific fault type of the planetary gearbox under the time-varying speed condition.

[0012] Furthermore, in step S1, based on the vibration signal model of the planetary gearbox gear fault, the original vibration signal x(t) is subjected to Hilbert transform to separate the amplitude-frequency modulation signal to eliminate the complex sidebands in the spectrum.

[0013] The model of the vibration signal x(t) during solar failure is as follows:

[0014]

[0015] In the formula, A s Indicates the amplitude of the amplitude modulation, B s Indicates the amplitude of the amplitude modulation, f sr is the rotation frequency of the sun gear, f m is the meshing frequency, f s is the characteristic frequency of the sun gear fault, and the initial phases of the amplitude modulation and amplitude modulation are denoted as θ, and α k .

[0016] AM signal x AM (t) and the frequency modulated signal x FM (t) can be calculated as follows:

[0017]

[0018] Wherein, H(.) represents the Hilbert transform.

[0019] Furthermore, in step S2, a high-resolution time-frequency diagram is generated by using a multi-order synchronous compression transformation, and the formula of the multi-order synchronous compression transformation is as follows:

[0020]

[0021] In the formula, represents short-time Fourier transform, τ represents time variable, ω represents frequency unit, i represents imaginary unit, Ts [N] (t, η) represents the polysynchronous compression transform, δ(.) represents the Dirac function, η represents the frequency, ω(t, ω) represents the instantaneous frequency, and [N] represents the order of the polysynchronous compression transform.

[0022] Furthermore, in step S3, an adaptive multi-ridge detection method is proposed by replacing the fixed bandwidth function with a frequency modulation operator. The frequency modulation operator formula is as follows:

[0023]

[0024] In the formula, and Represents the short-time Fourier transform calculated with different window functions.

[0025] Furthermore, in step S4, a time warping method based on the estimated phase inverse is used to implement non-uniform resampling to eliminate the interference of speed fluctuations and extract stable fault features. The formula for time warping can be expressed as:

[0026]

[0027] In the formula, represents the inverse of the estimated phase, E i (t) represents the error of the estimated phase.

[0028] Furthermore, in step S5, the resampled AM / FM signal spectrum is analyzed to confirm the specific fault type of the planetary gearbox under the time-varying speed condition. By comparing the theoretical fault frequency of the planetary gearbox with the fault characteristic frequency in the AM / FM signal spectrum, the gear fault type of the planetary gearbox when the vibration signal is collected can be determined.

[0029] Compared with the prior art, the beneficial effects of the present invention are:

[0030] The method of the present invention utilizes the vibration signal model of the planetary gearbox to theoretically reveal the fault mechanism and explore the inherent modulation mode. At the same time, based on the vibration signal model, the amplitude modulation and frequency modulation signals can be accurately captured to avoid complex sidebands and extract time-varying modulation features. In addition, the traditional fixed bandwidth function is replaced by a frequency modulation operator to adjust the search range and adaptively extract time-varying ridges. Through non-uniform resampling, time warping is applied to eliminate the interference of speed fluctuations, thereby extracting stable fault features. According to the proposed diagnostic procedure, a preliminary diagnostic conclusion can be drawn from the amplitude modulation feature extraction, and this conclusion can be further verified by the frequency modulation feature extraction. The above method is applied for the first time in the field of planetary gearbox fault diagnosis under time-varying speed conditions, and is further combined with the time-frequency analysis method to form a complete set of planetary gearbox fault diagnosis methods under variable speed conditions, which is superior in extracting the modulation characteristics of variable speed vibration signals of planetary gearboxes. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a flow chart of the planetary gearbox fault diagnosis method based on time-varying modulation feature extraction in the present invention;

[0032] Figure 2 It is the time domain diagram and frequency domain diagram of the original vibration signal of the planetary gearbox in the present invention;

[0033] Figure 3 It is a time domain diagram of the amplitude modulation signal and the frequency modulation signal in the present invention;

[0034] Figure 4 It is a time-frequency diagram of multi-order synchronous compression transformation of the amplitude modulation signal and the frequency modulation signal in the present invention;

[0035] Figure 5 A ridgeline diagram extracted from an amplitude modulated signal and a frequency modulated signal in the present invention;

[0036] Figure 6 It is a spectrum characteristic diagram of the amplitude modulation signal and the frequency modulation signal after time warping processing in the present invention. DETAILED DESCRIPTION

[0037] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be further described below.

[0038] The planetary gearbox fault diagnosis method based on time-varying modulation feature extraction of this embodiment includes the following specific steps:

[0039] Figure 1 It is a flow chart of the planetary gearbox fault diagnosis method based on time-varying modulation feature extraction in the present invention.

[0040] like Figure 1 As shown, the planetary gearbox fault diagnosis method based on time-varying modulation feature extraction involved in this embodiment includes the following steps:

[0041] Step S1, using Hilbert transform to separate the amplitude-frequency modulated signal from the original vibration signal.

[0042] Figure 2 It is the time domain diagram and frequency domain diagram of the original vibration signal of the planetary gearbox in the present invention.

[0043] Figure 3 It is a time domain diagram of the amplitude modulation signal and the frequency modulation signal in the present invention.

[0044] In this embodiment, an acceleration sensor is arranged on the planetary gearbox test bench to collect the vibration signal of the planetary gearbox under variable speed conditions in real time, with a sampling frequency of 20000Hz. Based on the vibration signal model of the planetary gearbox gear fault, the original vibration signal x(t) is separated from the amplitude-frequency modulation signal by Hilbert transform to eliminate the complex sidebands in the spectrum.

[0045] The model of the vibration signal x(t) during solar failure is as follows:

[0046]

[0047] In the formula, A s Indicates the amplitude of the amplitude modulation, B s Indicates the amplitude of the amplitude modulation, f sr is the rotation frequency of the sun gear, f m is the meshing frequency, f sis the characteristic frequency of the sun gear fault, and the initial phases of the amplitude modulation and amplitude modulation are denoted as θ, and α k .

[0048] AM signal x AM (t) and the frequency modulated signal x FM (t) can be calculated as follows:

[0049]

[0050] Wherein, H(.) represents the Hilbert transform.

[0051] like Figure 2 As shown, the sampling time of the original vibration signal of the planetary gearbox is 8 seconds. Figure 3 Shown are time domain plots of AM and FM signals.

[0052] Step S2, using multi-order synchronous compression transformation to generate time-frequency diagrams of amplitude modulation and frequency modulation signals respectively.

[0053] Figure 4 It is a time-frequency diagram of multi-order synchronous compression transformation of the amplitude modulation signal and the frequency modulation signal in the present invention;

[0054] In this embodiment, the high-resolution time-frequency diagram of the amplitude modulated and amplitude modulated signals is generated by a multi-order synchronous compression transformation, and the formula of the multi-order synchronous compression transformation is as follows:

[0055]

[0056] In the formula, represents short-time Fourier transform, τ represents time variable, ω represents frequency unit, i represents imaginary unit, Ts [N] (t, η) represents the polysynchronous compression transform, δ(.) represents the Dirac function, η represents the frequency, ω(t, ω) represents the instantaneous frequency, and [N] represents the order of the polysynchronous compression transform.

[0057] Step S3, extracting fault-related ridges using an adaptive multi-ridge extraction algorithm.

[0058] Figure 5 is a ridgeline diagram extracted from an AM signal and a FM signal in the present invention;

[0059] In this embodiment, an adaptive multi-ridge detection method is proposed. By replacing the fixed bandwidth function with a frequency modulation operator, the time-varying ridge features in the time-frequency representation are extracted. The frequency modulation operator formula is as follows:

[0060]

[0061] In the formula, and Represents the short-time Fourier transform calculated with different window functions.

[0062] Figure 5 The figure shows the extracted ridge lines of the AM signal and the FM signal. The extracted ridge lines are consistent with the variation trend of the actual instantaneous frequency ridge lines.

[0063] Step S4, reconstruct the signal and perform non-uniform resampling in the time domain using time warping.

[0064] Figure 6 It is a spectrum characteristic diagram of the amplitude modulation signal and the frequency modulation signal after time warping processing in the present invention.

[0065] In this embodiment, time warping can be used to achieve non-uniform resampling and adjust the scale of the time axis according to the inverse of the estimated phase. The spectrum tailing phenomenon is eliminated in the frequency domain of the warped monomer, and the fault characteristic frequency can be effectively extracted. The time warping process can be expressed as follows:

[0066]

[0067] In the formula, represents the inverse of the estimated phase, E i (t) represents the error of the estimated phase.

[0068] Figure 6 The figure shows the spectrum characteristics of the AM signal and FM signal after time warping. The extracted characteristic frequencies are not interfered by complex sidebands, nor are they drowned by background noise and irrelevant pulses. The spectrum tailing phenomenon caused by speed fluctuations is also effectively eliminated.

[0069] Step S5, analyzing the resampled AM / FM signal spectrum to confirm the specific fault type of the planetary gearbox under the time-varying speed condition. By comparing the theoretical fault frequency of the planetary gearbox with the fault characteristic frequency in the AM / FM signal spectrum, the gear fault type of the planetary gearbox when the vibration signal is collected can be determined.

[0070] According to the planetary gearbox fault diagnosis method based on time-varying modulation feature extraction involved in the present invention, based on the vibration signal model of the planetary gearbox with different faulty gears, the amplitude modulation and frequency modulation signals can be separated to avoid complex sidebands in the frequency domain, and the time-varying modulation characteristics can be comprehensively extracted. Secondly, an adaptive multi-ridge detection method is proposed to retrieve time-varying ridges by replacing the fixed bandwidth function with a frequency modulation operator. Finally, time warping is proposed to adjust the scale of the time axis, and non-uniform resampling is performed based on the inverse of the phase to eliminate speed fluctuation interference and extract the fault characteristic frequency.

[0071] The above is only a preferred embodiment of the present invention and does not limit the present invention in any way. Any technician in the relevant technical field, without departing from the scope of the technical solution of the present invention, makes any form of equivalent replacement or modification to the technical solution and technical content disclosed in the present invention, which does not depart from the content of the technical solution of the present invention and still falls within the protection scope of the present invention.

Claims

1. A planetary gearbox fault diagnosis method based on time-varying modulation feature extraction, characterized in that: The following steps are involved: Step S1, using Hilbert transform to separate the amplitude modulation and frequency modulation signals from the original vibration signal; Step S2, using multi-order synchronous compression transformation to generate time-frequency diagrams of amplitude modulation and frequency modulation signals respectively; Step S3, extracting fault-related ridges using an adaptive multi-ridge extraction algorithm; Step S4, reconstructing the signal and performing non-uniform resampling in the time domain using time warping; Step S5, analyzing the resampled frequency spectrum of the AM / FM signal to confirm the specific fault type of the planetary gearbox under the time-varying speed condition.

2. The planetary gearbox fault diagnosis method based on time-varying modulation feature extraction according to claim 1 is characterized in that: In step S1, based on the vibration signal model of the planetary gearbox and Hilbert transform, the amplitude-frequency modulation signal is separated to eliminate the complex sidebands in the spectrum. The model of the vibration signal x(t) during a solar failure is expressed as follows: In the formula, A s Indicates the amplitude of the amplitude modulation, B s Indicates the amplitude of the amplitude modulation, f sr is the rotation frequency of the sun gear, f m is the meshing frequency, f s is the characteristic frequency of the sun gear fault, and the initial phases of the amplitude modulation and amplitude modulation are denoted as θ, and α k , AM signal x AM (t) and the frequency modulated signal x FM (t) can be calculated as follows: In the formula, H(.) represents the Hilbert transform, Based on the vibration signal model of the planetary gearbox and Hilbert transform, the AM and FM signals are separated to eliminate the complex sidebands in the spectrum.

3. The planetary gearbox fault diagnosis method based on time-varying modulation feature extraction according to claim 2 is characterized in that: In step S2, a high-resolution time-frequency diagram is generated by using a multi-order synchronous compression transformation, and the formula of the multi-order synchronous compression transformation is as follows: In the formula, represents short-time Fourier transform, τ represents time variable, ω represents frequency unit, i represents imaginary unit, Ts [N] (t, η) represents the polysynchronous compression transform, δ(.) represents the Dirac function, η represents the frequency, ω(t, ω) represents the instantaneous frequency, and [N] represents the order of the polysynchronous compression transform.

4. The planetary gearbox fault diagnosis method based on time-varying modulation feature extraction according to claim 1 is characterized in that: In step S3, the extraction method is to replace the fixed bandwidth function with a frequency modulation operator, and the frequency modulation operator formula is as follows: In the formula, and Represents the short-time Fourier transform calculated with different window functions.

5. The planetary gearbox fault diagnosis method based on time-varying modulation feature extraction according to claim 1 is characterized in that: In step S4, a time warping method based on the estimated phase inverse is used to implement non-uniform resampling, so as to eliminate the interference of speed fluctuation and extract stable fault features. The time warping formula can be expressed as: In the formula, represents the inverse of the estimated phase, E i (t) represents the error of the estimated phase.

6. The planetary gearbox fault diagnosis method based on time-varying modulation feature extraction according to claim 1 is characterized in that: In the step S5, the gear fault type of the planetary gearbox when the vibration signal is collected is determined by comparing the theoretical fault frequency of the planetary gearbox with the fault characteristic frequency in the spectrum of the AM / FM signal.

Images