Modulation signal separation and demodulation method based on multi-vibration-source coupling

Abstract

The invention discloses a modulation signal separation and demodulation method based on multi-vibration-source coupling, and belongs to the technical field of mechanical equipment fault diagnosis. Comprising the steps that firstly, collected mechanical equipment measurement signals are decomposed into a series of sub-band signals, the generated sub-band signals are analyzed through modulation signal bispectrum, and synchronous separation of multi-vibration-source coupling modulation signals is achieved; subsequently, sub-band signals generated by modulation signal bispectrum demodulation are subjected to synchronous averaging to generate local modulation signal bispectrums, and random noise and interference frequency in the sub-band signals are eliminated. And finally, performing amplitude and amplitude-frequency analysis on the signals generated after the random noise and the interference frequency component are eliminated, and demodulating fault characteristics of different components of the mechanical equipment. According to the method, fault features of different components can be effectively separated and demodulated from the multi-vibration-source coupling modulation signals, and a brand new view angle is provided for fault diagnosis of mechanical equipment.

Description

Technical Field

[0001] The invention belongs to the technical field of mechanical equipment fault diagnosis, and in particular relates to a method for separating and demodulating modulation signals based on multi-vibration source coupling. Background Art

[0002] As a unique aircraft, helicopters play an important role in ensuring national security and people's lives due to their advantages such as hovering in the air, vertical take-off and landing, and low-speed maneuvering. In a helicopter, the main reducer is a core component connected to the input end of the engine and the output end of the rotor and tail rotor. Whether it works normally is the key to ensuring the safe flight of the helicopter. However, due to long-term service in harsh environments such as high speed, heavy load, and variable working conditions, the main reducer is prone to sudden failures such as fatigue failure and damage, which directly affect the safe operation of the helicopter and even cause catastrophic casualties. Therefore, it is of great significance to carry out research on main reducer fault diagnosis technology to avoid catastrophic accidents and improve the safe service performance of helicopters.

[0003] As a crucial core component of the helicopter main reducer, the planetary gear train has many vibration excitation sources caused by complex multi-tooth meshing motion and multi-axis transmission, and is affected by the time-varying meshing force direction and transmission path. Its vibration response signal presents the characteristics of multi-order meshing frequency bands and a large number of modulation sidebands, and each meshing frequency band and modulation sideband are coupled with each other. In the existing signal demodulation analysis method, the fault characteristics of the planetary gear train are accurately extracted by effectively separating the modulation component and the carrier component in the measurement signal of the single component signal, but it is difficult to demodulate and analyze multi-carrier and multi-modulation wave signals at the same time, and its analysis results often cannot meet the requirements of actual fault diagnosis. Modulation signal bispectrum analysis is one of the most widely used resonance demodulation methods in mechanical equipment fault diagnosis technology. It can demodulate the fault characteristic components from the resonance frequency band by suppressing random noise and interference components. However, modulation signal bispectrum analysis has not yet realized the demodulation of fault components of different components from multi-vibration source signals, which may bring huge challenges. Since the vibration signals collected by the planetary gear train are not simply composed of several single signals, but are coupled with different component signals through complex transmission paths, and the signals of each component will modulate and interfere with each other, how to study the separation and demodulation of multi-vibration source coupled modulation signals is an urgent research topic in the field of mechanical equipment condition monitoring and fault diagnosis. Summary of the invention

[0004] The purpose of the present invention is to provide a method for separating and demodulating modulation signals based on multi-vibration source coupling to solve the technical problems mentioned in the above background technology.

[0005] To achieve the above purpose, a specific technical solution of the present invention based on a multi-source coupling modulation signal separation and demodulation method is as follows:

[0006] The present invention provides a method for separating and demodulating modulation signals based on multi-vibration source coupling, which is characterized in that the method comprises the following steps:

[0007] Step 1: Signal separation: Decompose the collected mechanical equipment multi-vibration source measurement signals into a series of sub-band signals, and analyze the generated sub-band signals by using the bi-spectrum of the modulation signal to achieve synchronous separation of the mechanical equipment multi-vibration source measurement signals;

[0008] Step 2, interference component elimination: synchronously average the sub-band signals generated by the dual spectrum demodulation of the modulation signal to generate a local modulation signal dual spectrum, and eliminate random noise and interference frequency in the sub-band signals;

[0009] Step 3: Demodulation of coupled signals: by performing amplitude and amplitude-frequency analysis on the signals generated after eliminating random noise and interference frequency components, the fault characteristics of different components of the mechanical equipment are demodulated;

[0010] The specific steps of step one are as follows:

[0011] Step 1: Measure the collected mechanical equipment multi-vibration source signal x q (t) is transformed by Fourier, which is expressed as:

[0012]

[0013] In formula (1), X q (f) represents the spectral components of the signal.

[0014] Step 2: Convert the spectrum X q (f) Divide into K sub-band signals, each sub-band signal contains M values, and adjacent sub-band segments are allowed to overlap. When the number of sampling points in a sub-band segment is less than zero, automatic filling can be performed, and K is defined as:

[0015] K=floor(N-nfft) / (nfft-nlap)+1(2)

[0016] In formula (2), N represents the original signal length, nfft represents the window function length, nlap represents the overlapped sample length, and floor(·) represents the rounded-down function.

[0017] The specific steps of step 2 are as follows:

[0018] Step 1: Calculate the bispectrum of the modulation signal of each sub-band signal, which is expressed as:

[0019]

[0020] In formula (3), M1=max(0,-f2,f2), M2=min(M-1,M-1-f2,M-1+f2), and i represents the number of sub-band signals.

[0021] Step 2: Synchronously average each sub-band signal to generate a local modulation signal bispectrum, which is defined as:

[0022]

[0023] The specific steps of step three are as follows:

[0024] Step 1: Considering that there are multiple harmonic components of meshing frequency in each vibration source signal, its signal model y(t) is expressed as:

[0025]

[0026] In formula (5), f z represents the meshing frequency, A n and θ n represents the amplitude and initial phase of the nth meshing frequency component, a n (t) and b n (t) represents the amplitude modulation and frequency modulation components of the nth meshing frequency component, which can be expressed as:

[0027]

[0028] In formulas (6) and (7), A n，k , A n，k and Respectively represent a n (t) and b n The amplitude and phase of the corresponding order in (t), f m Indicates the fault characteristic frequency. To ensure that the phase information does not affect the amplitude modulation, b m (t) is set to zero and its signal model y(t) is redefined as:

[0029]

[0030] Step 2: According to the equivalence of trigonometric functions, perform Fourier transform on y(t), which is defined as:

[0031]

[0032] In formula (9), δ(·) represents the impulse function.

[0033] Step 3: Substitute equation (9) into equation (4), and the amplitude modulation result of the local modulation signal bispectrum at (f, f2) is expressed as:

[0034]

[0035] In formula (10), there are sidebands with fault characteristic frequencies as intervals near the meshing frequency. When a local fault occurs in the mechanical equipment, the sidebands in the spectrum will appear at the position (m, n and k are integers).

[0036] Step 4: Without loss of generality, set the amplitude a n (t) = 0, initial phase θ n =β n，l =0, then the analog signal model in equation (8) is simplified to:

[0037]

[0038] Step 5: For when and When certain conditions are met, its y(t) is expressed as:

[0039]

[0040] Step 6: Use Bessel's formula y(t) is redefined as:

[0041]

[0042] In formula (13), J λ (β) represents the first order λ and parameter β of the Bessel function, and the Fourier transform of y(t) is expressed as:

[0043]

[0044] Step 7: Substitute equation (14) into equation (4), and the frequency modulation result of the local modulation signal bispectrum at (f, f2) is expressed as:

[0045]

[0046] In formula (15), the fault characteristic frequency is at the meshing frequency f m Near nf, the sidebands in the spectrum appear z +λlf m location.

[0047] The present invention provides a method for separating and demodulating multi-source coupled modulation signals, which has the following advantages:

[0048] (1) The present invention proposes a local modulation signal bispectrum method to eliminate random noise and interference frequency in the sub-band signal generated by bispectral demodulation of the modulation signal, thereby eliminating the interference component in the vibration signal.

[0049] (2) The present invention constructs a method for separating and demodulating multi-source coupled modulation signals, thereby realizing the separation and demodulation of multi-source coupled modulation signals and overcoming the problem that traditional demodulation methods cannot simultaneously demodulate multiple carriers and multiple modulation components. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a local modulation signal bispectrum diagram of embodiment 1 of the method for separation and demodulation of multi-vibration source coupling modulation signals of the present invention;

[0051] Figure 2 The analysis results of the local modulation signal bispectrum of Example 1 of the method for separating and demodulating multi-vibration source coupling modulation signals of the present invention are as follows ((a) meshing frequency is 300 Hz; (b) meshing frequency is 1000 Hz; (c) meshing frequency is 1600 Hz); DETAILED DESCRIPTION

[0052] In order to better understand the purpose, structure and function of the present invention, the following is a further detailed description of the method for separating and demodulating multi-source coupled modulation signals of the present invention in conjunction with the accompanying drawings.

[0053] Embodiment 1:

[0054] The present invention provides a method for separating and demodulating modulation signals based on multi-vibration source coupling, which is characterized in that the method comprises the following steps:

[0055] Step 1: Decompose the collected multi-vibration source measurement signals of mechanical equipment into a series of sub-band signals, and analyze the generated sub-band signals by using the bi-spectrum of the modulation signal to achieve synchronous separation of the multi-vibration source measurement signals of the mechanical equipment;

[0056] The specific steps of step one are as follows:

[0057] Step 1: Measure the collected mechanical equipment multi-vibration source signal x q (t) is transformed by Fourier, which is expressed as:

[0058]

[0059] In formula (1), X q (f) represents the spectral components of the signal.

[0060] Step 2: Convert the spectrum X q (f) Divide into K sub-band signals, each sub-band signal contains M values, and adjacent sub-band segments are allowed to overlap. When the number of sampling points in a sub-band segment is less than zero, automatic filling can be performed, and K is defined as:

[0061] K=floor(N-nfft) / (nfft-nlap)+1(2)

[0062] In formula (2), N represents the original signal length, nfft represents the window function length, nlap represents the overlapped sample length, and floor(·) represents the rounded-down function.

[0063] Step 2: Synchronously average the sub-band signals generated by the demodulation of the modulation signal bispectrum to generate the local modulation signal bispectrum. The result is as follows: Figure 1 shown by Figure 1 It can be seen that by eliminating random noise and interference frequency in the sub-band signal, the interference component in the signal can be eliminated.

[0064] The specific steps of step 2 are as follows:

[0065] Step 1: Calculate the bispectrum of the modulation signal of each sub-band signal, which is expressed as:

[0066]

[0067] In formula (3), M1=max(0,-f2,f2), M2=min(M-1,M-1-f2,M-1+f2), and i represents the number of sub-band signals.

[0068] Step 2: Synchronously average each sub-band signal to generate a local modulation signal bispectrum, which is defined as:

[0069]

[0070] Step 3: By performing amplitude and amplitude-frequency analysis on the signal generated after eliminating random noise and interference frequency components, the rotation frequency of different components of the mechanical equipment (i.e., f sx ,f sy and f sz ) and the modulation frequency (i.e., f rx1 , f rx2 , f ry1 ,f ry2 ,f rz1 and f rz2 ), the result is as follows Figure 2 shown by Figure 2 It can be seen that the proposed method realizes the separation and demodulation of multi-source coupled modulation signals, overcoming the problem that traditional demodulation methods cannot simultaneously demodulate multiple carriers and multiple modulation components.

[0071] The specific steps of step three are as follows:

[0072] Step 1: Considering that there are multiple harmonic components of meshing frequency in each vibration source signal, its signal model y(t) is expressed as:

[0073]

[0074] In formula (5), f z represents the meshing frequency, A n and θ n represents the amplitude and initial phase of the nth meshing frequency component, a n (t) and b n (t) represents the amplitude modulation and frequency modulation components of the nth meshing frequency component, which can be expressed as:

[0075]

[0076] In formulas (6) and (7), A n，k , A n，k and Respectively represent a n (t) and b n The amplitude and phase of the corresponding order in (t), f m Indicates the fault characteristic frequency. To ensure that the phase information does not affect the amplitude modulation, b m (t) is set to zero and its signal model y(t) is redefined as:

[0077]

[0078] Step 2: According to the equivalence of trigonometric functions, perform Fourier transform on y(t), which is defined as:

[0079]

[0080] In formula (9), δ(·) represents the impulse function.

[0081] Step 3: Substitute equation (9) into equation (4), and the amplitude modulation result of the local modulation signal bispectrum at (f, f2) is expressed as:

[0082]

[0083] In formula (10), there are sidebands with fault characteristic frequencies as intervals near the meshing frequency. When a local fault occurs in the mechanical equipment, the sidebands in the spectrum will appear at the position (m, n and k are integers).

[0084] Step 4: Without loss of generality, set the amplitude a n (t) = 0, initial phase θ n =β n，l =0, then the analog signal model in equation (8) is simplified to:

[0085]

[0086] Step 5: For when and When certain conditions are met, its y(t) is expressed as:

[0087]

[0088] Step 6: Use Bessel's formula y(t) is redefined as:

[0089]

[0090] In formula (13), J λ (β) represents the first order λ and parameter β of the Bessel function, and the Fourier transform of y(t) is expressed as:

[0091]

[0092] Step 7: Substitute equation (14) into equation (4), and the frequency modulation result of the local modulation signal bispectrum at (f, f2) is expressed as:

[0093]

[0094] In formula (15), the fault characteristic frequency is at the meshing frequency f m Near nf, the sidebands in the spectrum appear z +λlf m location.

[0095] It is to be understood that the present invention is described by some embodiments, and it is known to those skilled in the art that various changes or equivalent substitutions may be made to these features and embodiments without departing from the spirit and scope of the present invention. In addition, under the teachings of the present invention, these features and embodiments may be modified to adapt to specific circumstances and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the scope of protection of the present invention.

Claims

1. A method for separating and demodulating modulation signals based on multi-vibration source coupling. It is characterized in that The following steps are involved: Step S1: Signal separation The collected mechanical equipment multi-vibration source measurement signals are decomposed into a series of sub-band signals, and the generated sub-band signals are analyzed by using the bi-spectrum of the modulation signal to achieve synchronous separation of the mechanical equipment multi-vibration source measurement signals; Step S2: Interference component elimination Synchronously averaging the sub-band signals generated by the dual-spectrum demodulation of the modulation signal to generate a local modulation signal dual-spectrum, and eliminating random noise and interference frequency in the sub-band signals; Step S3: Demodulation of coupled signal By performing amplitude and amplitude-frequency analysis on the signal generated after eliminating random noise and interference frequency components, the fault characteristics of different components of the mechanical equipment can be demodulated.

2. According to the method for separating and demodulating multi-source coupling modulation signals according to claim 1, It is characterized in that The step S1 specifically includes the following steps: Step S1-1: collect the multi-vibration source measurement signal x of the mechanical equipment q (t) is transformed by Fourier, which is expressed as: In formula (1), X q (f) represents the spectral components of the signal; Step S1-2: Convert the spectrum X q (f) Divide into K sub-band signals, each sub-band signal contains M values, and adjacent sub-band segments are allowed to overlap; when the number of sampling points in the sub-band segment is less than zero, automatic filling can be performed, and K is defined as: K=floor(N-nfft) / (nfft-nlap)+1(2) In formula (2), N represents the original signal length, nfft represents the window function length, nlap represents the overlapped sample length, and floor(·) represents the rounded-down function.

3. According to the method for separating and demodulating multi-source coupling modulation signals according to claim 1, It is characterized in that The step S2 specifically includes the following steps: Step S2-1: Calculate the bispectrum of the modulation signal of each sub-band signal, which is expressed as: In formula (3), M 1 =max(0,-f 2 ,f 2 ), M 2 =min(M-1,M-1-f 2 ,M-1+f 2 ), i represents the number of sub-band signals; Step S2-2: Perform synchronous averaging on each sub-band signal to generate a local modulation signal bispectrum, which is defined as:

4. The method for separating and demodulating modulation signals based on multi-vibration source coupling according to claim 1, It is characterized in that The step S3 specifically includes the following steps: Step S3-1: Considering that there are multiple harmonic components of meshing frequency in each vibration source signal, its signal model y(t) is expressed as: In formula (5), f z represents the meshing frequency, A n and θ n represents the amplitude and initial phase of the nth meshing frequency component, a n (t) and b n (t) represents the amplitude modulation and frequency modulation components of the nth meshing frequency component, which can be expressed as: In formulas (6) and (7), A n,k , A n，k and Respectively represent a n (t) and b n The amplitude and phase of the corresponding order in (t), f m Indicates the fault characteristic frequency; To ensure that the phase information does not affect the amplitude modulation, b m (t) is set to zero and its signal model y(t) is redefined as: Step S3-2: According to the equivalence of trigonometric functions, Fourier transform is performed on y(t), which is defined as: In formula (9), δ(·) represents the impulse function; Step S3-3: Substitute equation (9) into equation (4), the local modulation signal bispectrum is (f, f 2 ) is expressed as: In formula (10), there are sidebands with fault characteristic frequencies as intervals near the meshing frequency; when a local fault occurs in the mechanical equipment, the sidebands in the spectrum will appear at the position (m, n and k are integers); Step S3-4: Without loss of generality, set the amplitude a n (t) = 0, initial phase θ n =β n，l =0, then the analog signal model in equation (8) is simplified to: Step S3-5: For when and When certain conditions are met, its y(t) is expressed as: Step S3-6: Using Bessel formula y(t) is redefined as: In formula (13), J λ (β) represents the first order λ and parameter β of the Bessel function, and the Fourier transform of y(t) is expressed as: Step S3-7: By substituting equation (14) into equation (4), the local modulation signal bispectrum is obtained in (f, f 2 ) is expressed as: In formula (15), the fault characteristic frequency is at the meshing frequency f m Near nf, the sidebands in the spectrum appear z +λlf m location.

Images