Gear fault evolution analysis method and system based on vibration signal morphological characteristics

Abstract

The invention discloses a gear fault evolution analysis method and system based on vibration signal morphological characteristics. The gear fault evolution analysis method comprises the following steps of obtaining a vibration signal in a gear operation process, and obtaining a reference spectral line and a target spectral line through data segmentation and frequency spectrum preprocessing; establishing a modulation spectral line dictionary to reconstruct the target spectral line into a first target spectral line, and correcting the first target spectral line into a second target spectral line by a reference spectral line; and carrying out differential extraction on frequency domain fault component spectral lines of the second target spectral line and the reference spectral line, finding out an optimal frequency domain fault component of gear operation from the frequency domain fault component spectral lines, and converting the optimal frequency domain fault component into time domain fault morphological characteristics. According to the method, the main processing process of the vibration signals is put on the frequency domain, interference components are removed by shifting the non-variable frequency domain, the degree and the type of the fault can be visually obtained only through the extracted fault form characteristics without setting a threshold value, and the method has the advantages of being accurate, visual and simple in process.

Description

Technical Field

[0001] The present invention relates to the technical field of gear transmission and state monitoring, and in particular to a gear fault evolution analysis method and system based on vibration signal morphological characteristics. Background Art

[0002] Gears are key components for transmitting motion and power in mechanical equipment. They have high precision, fast speed response and high transmission efficiency. Therefore, they are widely used in important fields such as aerospace, marine engineering and transportation. However, due to factors such as material defects and abnormal working conditions, gears that have been in service for a long time are often prone to various types of faults, such as pitting on the tooth surface and cracks on the tooth root. These faults seriously affect the operating stability and reliability of the system.

[0003] Vibration signal analysis is one of the main means of gear fault evolution analysis. When a gear tooth fails, the vibration signal shows obvious periodic impact in the time domain, which is reflected in the spectrum as modulation side spectrum lines centered on the meshing frequency and spaced by the rotation frequency of the shaft. Traditional gear fault evolution analysis includes time domain and frequency domain analysis. Time domain analysis often relies on the changes of statistical parameter indicators, such as kurtosis and root mean square. Frequency domain analysis constructs a bandpass filter to extract narrowband components near the meshing frequency, or uses signal decomposition methods to decompose the original vibration signal into sub-signals, and then extracts the fault frequency and analyzes its fault evolution law, such as empirical mode decomposition and variational mode decomposition. However, bandpass filtering and signal decomposition have limitations in eliminating irrelevant interference in the modulation sideband. More importantly, it is difficult to effectively distinguish gear fault types with similar spectral line distributions, and the accuracy is not high. In addition, whether it is time domain analysis or frequency domain analysis, the understanding of the evolution law of gear faults ultimately usually depends on the changes of indicators, which leads to dependence on indicators and lacks the ability to explain the fault mechanism in detail. The process is complicated and cannot be intuitively identified. Summary of the invention

[0004] Purpose of the invention: The purpose of the present invention is to provide an accurate, visual and simple method and system for analyzing gear fault evolution based on vibration signal morphological characteristics.

[0005] Technical solution: The gear fault evolution analysis method based on vibration signal morphological characteristics of the present invention comprises the following steps:

[0006] Obtain the vibration signal during the gear operation process, and obtain the reference spectrum and target spectrum through data segmentation and spectrum preprocessing;

[0007] Establishing a modulation spectrum line dictionary to reconstruct the target spectrum line into a first target spectrum line, and correcting the first target spectrum line into a second target spectrum line by using a reference spectrum line;

[0008] Differentially extract the frequency-domain fault component spectrum lines of the second target spectrum line and the reference spectrum line, find the optimal frequency-domain fault components of the gear operation from them, and convert them into time-domain fault morphological features.

[0009] Further, the vibration signal is acquired under the working conditions of the gear system, and the parameters of the working conditions include sampling frequency, effective frequency range, number of gear teeth, gear rotation speed, rotation frequency of the rotating shaft, and gear meshing frequency; among them, the rotation frequency of the gear operation is represented by the rotation frequency of the rotating shaft.

[0010] Further, the modulation spectrum line dictionary characterizes the fault frequency with the attenuation coefficient as a variable, and the specific formula is as follows:

[0011]

[0012] In the formula, δ(·) represents the impulse function; α represents the modulation spectrum line attenuation coefficient, 0 < α < 1; κ represents the number of spectrum lines, 1 < κ < K, and according to the distribution characteristics of the gear system meshing frequency and rotation frequency, the upper limit K is not greater than 0.25*z1, where z1 represents the number of teeth of the target gear; ω represents the independent variable; R + represents the set of positive real numbers; N + represents the set of positive integers; f z represents the rotation frequency of the rotating shaft where the gear is located; D(ω) represents the constructed spectrum line dictionary.

[0013] Further, in the process of generating the first target spectrum line by the modulation spectrum line dictionary, shift-invariant sparsity is introduced, and the specific process is as follows:

[0014] First, establish a reconstruction formula according to the modulation spectrum line dictionary:

[0015] X an (ω) = D*s + ε

[0016] In the formula, ε represents the error term; D represents the modulation spectrum line dictionary; s represents the sparse coefficient; X an (ω) represents the target spectrum line;

[0017] Secondly, introduce L1 regularization, take the regularization coefficient as a variable, and solve the reconstruction formula with the minimum of the error term as the goal to obtain the first target spectrum line. The specific process is as follows:

[0018]

[0019] X ar (ω) = D*s

[0020] In the formula, argmin(·) represents taking the minimum value of the objective function; ||·||1 represents the L1 norm of the s vector; λ > 0 represents the regularization parameter, which is used to control the trade-off between sparsity and the error term; Xar (ω) represents the first target spectrum line; m and n represent the position of the element; N and L represent the number of rows and columns of the dictionary.

[0021] Furthermore, the process of obtaining the optimal frequency domain fault component is as follows:

[0022] The acquired frequency domain fault component spectrum includes multiple frequency domain fault components, each of which corresponds to a spectrum line interval, and the spectrum line interval of the frequency domain fault component spectrum is extracted;

[0023] The frequency domain fault components corresponding to the spectrum line intervals distributed near the rotation frequency are used as the optimal frequency domain fault components.

[0024] Furthermore, the process of obtaining the optimal frequency domain fault component by enumeration is as follows:

[0025] sequentially calculating the variance between the spectral line spacing and the rotation frequency;

[0026] With the goal of minimizing the variance, the optimal attenuation coefficient and regularization coefficient are found through enumeration method, thereby obtaining the optimal first target spectrum line, second target spectrum line, frequency domain fault component spectrum line, and spectrum line interval, and the frequency domain fault component corresponding to the optimal spectrum line interval distributed near the rotation frequency is taken as the optimal frequency domain fault component.

[0027] Further, correcting the first target spectral line by the reference spectral line specifically includes:

[0028] Extracting peak values ​​of the reference spectrum line and the first target spectrum line;

[0029] Constructing a window function with the length of the rotation frequency as the minimum window length;

[0030] The window function is made to traverse the spectrum range by aligning the spectrum lines, and the frequency of the first target spectrum line is corrected based on the reference spectrum line; when the frequencies under the peaks in each window are the same, the second target spectrum line is obtained.

[0031] Furthermore, the data segmentation process is as follows:

[0032] Processing the vibration signal into time series data with a certain signal length;

[0033] The stable time series data are divided and aggregated into a reference area signal set, and the unstable time series data are divided and aggregated into a target area signal set; wherein, after each division, the length of the time series data at least includes one gear rotation cycle.

[0034] Furthermore, the spectrum preprocessing process is as follows:

[0035] Based on fast Fourier transform, the reference area signal set and the target area signal set are converted from the time domain to the frequency domain to obtain a reference area spectrum set and a target area spectrum set;

[0036] The spectral line value sets corresponding to the reference area spectrum set and the target area spectrum set obey normal distribution. The expectation of each set of spectral line values ​​is used as the amplitude. After normalization, the estimated value of the frequency corresponding to each set of spectral line values ​​is obtained, and the reference spectral lines and target spectral lines are obtained.

[0037] The gear fault evolution analysis system based on vibration signal morphological characteristics of the present invention comprises:

[0038] The spectrum acquisition module is used to obtain the vibration signal during the operation of the gear, and obtain the reference spectrum and target spectrum through data segmentation and spectrum preprocessing;

[0039] A target spectrum line optimization module, used for establishing a modulation spectrum line dictionary to reconstruct the target spectrum line as a first target spectrum line, and correcting the first target spectrum line as a second target spectrum line by using a reference spectrum line;

[0040] The fault feature extraction module is used to differentially extract the frequency domain fault component spectrum of the second target spectrum and the reference spectrum, find out the optimal frequency domain fault component of the gear operation, and convert it into a time domain fault morphological feature.

[0041] Beneficial effects: The present invention has the following significant effects: 1. Accuracy: The present invention first converts the vibration signal into the frequency domain, and performs a shift-invariant frequency domain reconstruction process by constructing a modulation spectrum line dictionary, thereby eliminating the interference components (including meshing frequency and noise) in the fault frequency band, and obtaining a purer gear fault signal, wherein the frequency domain sparse representation technology is only a part of removing interference in the process of reconstructing the target spectrum line into the first target spectrum line; in addition, the present invention also adaptively performs a variety of processing on the vibration signal, such as data segmentation, spectrum preprocessing, and correcting the first target spectrum line from the reference spectrum line to the second target spectrum line. The above means all make the fault morphological characteristics of the final output closer to the real vibration signal, thereby improving the accuracy of feature extraction; 2. Visualization: The present invention finally converts the frequency domain fault components near the rotation frequency of the gear operation into the time domain, and then obtains intuitive fault morphological characteristics, which is different from the traditional bandpass Compared with the filter and signal decomposition method, the time domain morphological changes of the fault signal can be observed intuitively. The extraction process can quickly and accurately identify different fault types without setting a threshold, overcoming the problem of poor visualization of gear fault signals by traditional technology. At the same time, for gears with different fault degrees, it can not only realize the diagnosis of a single type of fault, but also identify and analyze the evolution of faults of different degrees, accurately reflect its unique fault characteristics, and provide an intuitive and effective tool for fault diagnosis and prediction; 3. Simple process: The present invention sets all technical processes in the frequency domain, does not need to build an element library in the time domain, and does not need to build an element library in advance, which simplifies the processing process; in addition, the present invention finds the optimal frequency domain fault component near the rotation frequency of the gear from the frequency domain fault component spectrum as the fault morphological feature, and the extraction process does not need to set a threshold, which is also easy to process and more efficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a flow chart of the fault evolution analysis method;

[0043] Figure 2 It is the dynamic response diagram under pitting failure;

[0044] Figure 3 This is the characteristic diagram of pitting failure morphology;

[0045] Figure 4 It is the vibration signal diagram of different pitting degrees;

[0046] Figure 5 The fault morphology characteristic diagram of different pitting degrees;

[0047] Figure 6 Schematic diagram of the dynamic response under crack failure;

[0048] Figure 7 Schematic diagram of crack failure morphology characteristics;

[0049] Figure 8 It is the vibration signal diagram of different crack degrees;

[0050] Fig. 9 Fault morphology characteristic diagram of different crack degrees. DETAILED DESCRIPTION

[0051] The present invention is further explained below in conjunction with the accompanying drawings and specific implementation methods.

[0052] See also Figure 1 As shown, the gear fault evolution analysis method based on vibration signal morphological characteristics of the present invention comprises the following steps:

[0053] Obtain the vibration signal during the gear operation process, and obtain the reference spectrum and target spectrum through data segmentation and spectrum preprocessing;

[0054] Establishing a modulation spectrum line dictionary to reconstruct the target spectrum line into a first target spectrum line, and correcting the first target spectrum line into a second target spectrum line by using a reference spectrum line;

[0055] The frequency domain fault component spectra of the second target spectrum and the reference spectrum are differentially extracted to find out the optimal frequency domain fault component that can best characterize the gear operation fault and convert it into a time domain fault morphological feature.

[0056] The gear fault evolution analysis system based on vibration signal morphological characteristics of the present invention comprises:

[0057] The spectrum acquisition module is used to obtain the vibration signal during the operation of the gear, and obtain the reference spectrum and target spectrum through data segmentation and spectrum preprocessing;

[0058] A target spectrum line optimization module, used for establishing a modulation spectrum line dictionary to reconstruct the target spectrum line as a first target spectrum line, and correcting the first target spectrum line as a second target spectrum line by using a reference spectrum line;

[0059] The fault feature extraction module is used to differentially extract the frequency domain fault component spectrum of the second target spectrum and the reference spectrum, find out the optimal frequency domain fault component that can best characterize the gear operation fault, and convert it into a time domain fault morphological feature.

[0060] The vibration signal is obtained under the working condition of the gear system, and the parameters of the working condition include sampling frequency, effective frequency range, number of gear teeth, gear speed, rotation frequency of the shaft, and gear meshing frequency; wherein the rotation frequency of the shaft represents the rotation frequency of the gear operation. Specifically, the sampling frequency f is specified according to the requirements of the signal acquisition system. s According to the Nyquist sampling theorem, the effective frequency range is [0, 0.5*f s]; determine the number of gear teeth z1 and z2, record the speed n of the gear system during the test, and further calculate the shaft speed f z =n / 60 and gear meshing frequency f m =f z *z1. Install acceleration sensors at key locations of gearboxes or gear transmission systems, and use signal acquisition systems to preset sampling frequencies f s Collect vibration signals during gear operation.

[0061] After the vibration signal is collected, in order to facilitate subsequent processing and improve the accuracy of the signal, the collected vibration signal is segmented. The process is as follows:

[0062] First, the vibration signal is processed into time series data with a certain signal length. Record a long segment of the vibration signal time series data x(t), whose signal length is N x For the convenience of explanation, it is assumed that the signal sequence data x(t) contains healthy signals and fault signals.

[0063] Afterwards, the stable time series data is divided and aggregated into a reference area signal set, and the unstable time series data is divided and aggregated into a target area signal set; wherein, after each division, the length of the time series data contains at least one gear rotation cycle. Specifically, the vibration signal number x(t) in sequence form is divided into a reference area and a target area, wherein the reference area represents the stable part of the vibration signal x(t) in a healthy state, and the target area represents the signal sequence of the vibration signal x(t) after the reference area. The reference area and target area signals are respectively divided into lengths equal to N p The signal length is N p It should contain at least one gear rotation period, that is, N p ≥z1 / f z Therefore, the base region signal set x e (t) and the target area signal set x a (t) can be expressed as:

[0064] x e (t) = [x e,1 (t),x e,2 (t),x e,3 (t),...,x e,i (t)] T (1)

[0065] x a (t) = [x a,1 (t),x a,2 (t),x a,3 (t),...,x a,j (t)]T (2)

[0066] In the formula, i and j represent x e (t) and x a (t) is the number of signal segments, and i ≥ j. In order to ensure the continuity of the signal, it is recommended that the signal segments in the reference area and the signal segments in the target area maintain a certain overlap when dividing.

[0067] After data segmentation, in order to further improve the accuracy of the signal, spectrum preprocessing is performed. The process is as follows:

[0068] First, based on the fast Fourier transform, the reference area signal set and the target area signal set are converted from the time domain to the frequency domain to obtain the reference area spectrum set X e (ω) and the target region spectrum set X a (ω), and retain its corresponding reference region phase spectrum set F e and the target region phase spectrum set F a .

[0069] After that, the spectrum line value sets corresponding to the reference area spectrum set and the target area spectrum set obey the normal distribution, and the expectation of each set of spectrum line values ​​is taken as the amplitude. After normalization, the estimated value of the frequency corresponding to each set of spectrum line values ​​is obtained, and the reference spectrum line and the target spectrum line are obtained. Specifically, under the premise of ensuring the continuity of the time domain signal, it can be considered that the amplitude spectrum set X e (ω) and X a (ω) The spectral line values ​​corresponding to the same frequency obey the normal distribution, that is, A e,k =[A e,k,1 ,A e,k,2 ,…,A e,k,i ] T and A a,k =[A a,k,1 ,A a,k,2 ,…,A a,k,j ] T , where k is the number of frequencies, and A e,k ~N(μ e,k ,σ 2 e,k ), A a,k ~N(μ a,k ,σ 2 a,k ). The expected frequency μ e,k and μ a,k As the amplitude spectrum A e,k and A a,k The estimated value of k frequencies can be obtained by normalizing them, and the reference spectrum line X can be obtained according to equations (3) and (4). en (ω) and the target spectral line X an (ω ),The reference phase spectrum F is further obtained by the same method en and the target phase spectrum F an .

[0070] X en (ω) = nor[μ e,1 , μ e,2 , μ e,3 ,..., μ e,k T (3)

[0071] X an (ω) = nor[μ a,1 , μ a,2 , μ a,3 ,..., μ a,k T (4)

[0072] Based on the vibration mechanism of the faulty gear system, the present invention constructs a modulation spectral line dictionary, and characterizes the fault frequency with the attenuation coefficient as a variable. The specific formula is as follows:

[0073]

[0074] wherein, δ(·) represents an impulse function; α represents a modulation spectral line attenuation coefficient, 0 < α < 1; κ represents the number of spectral lines, 1 < κ < K. According to the distribution characteristics of the meshing frequency and the rotation frequency of the gear system, the upper limit K is not greater than 0.25*z1, where z1 represents the number of teeth of the target gear; ω represents the independent variable; R + represents the set of positive real numbers; N + represents the set of positive integers; f z represents the rotation frequency of the shaft where the gear is located; D(ω) represents the constructed spectral line dictionary.

[0075] In order to remove the interference in the fault frequency band, during the process of generating the first target spectral line from the modulation spectral line dictionary, shift-invariant sparsity is introduced, and the target spectrum is reconstructed in the shift-invariant frequency domain according to the modulation characteristics of the gear fault. The specific process is as follows:

[0076] First, a reconstruction formula is established according to the modulation spectral line dictionary:

[0077] X an (ω) = D*s + ε (6)

[0078] wherein, ε represents an error term; D represents the modulation spectral line dictionary; s represents the sparse coefficient; X an (ω) represents the target spectral line;

[0079] ​​Secondly, in order to ensure the sparsity of s, L1 regularization is introduced. Taking the regularization coefficient as a variable, the reconstruction formula is solved with the minimum error term as the goal to obtain the first target spectrum line. By selecting the maximum λ value that makes the number of non-zero elements of the solved s weight vector, the solution method can choose the fast soft threshold iterative algorithm (F-ISTA). The specific process is as follows:

[0080]

[0081] X ar (ω)=D*s (8)

[0082] Where argmin(·) represents the minimum value of the objective function; ||·||1 represents the L1 norm of the s vector; λ>0 represents the regularization parameter, which is used to control the trade-off between sparsity and error term; X ar (ω) represents the first target spectrum line; m and n represent the position of the element; N and L represent the number of rows and columns of the dictionary.

[0083] After obtaining the first target spectrum line, considering that there are working condition changes or shutdown inspections during the gear meshing process, it will cause slight fluctuations in the rotation speed, resulting in slight deviations in the effective spectrum line. Therefore, it is necessary to perform spectrum line offset correction before feature extraction. The present invention corrects the first target spectrum line by the reference spectrum line, specifically including:

[0084] First, the peak values ​​of the reference spectrum line and the first target spectrum line are extracted respectively. Taking the target spectrum amplitude Aar as an example, its specific expression is:

[0085]

[0086] Secondly, a window function is constructed with the length of the rotation frequency as the minimum window length. In this step, the reconstructed target spectrum X ar The effective frequency of (ω) includes the meshing frequency and the modulation frequency, so based on the reference spectrum X en (ω) for the reconstructed target spectrum X ar (ω) The spectrum line is shifted. The rotation frequency f is calculated z The corresponding length, and the window length is not greater than N fz To construct the window function, the calculation formula is:

[0087]

[0088] By aligning the spectrum, the window function traverses the spectrum range, and the frequency of the first target spectrum line is corrected based on the reference spectrum line; when the frequencies under the peaks in each window are the same, the second target spectrum line is obtained. Specifically, the entire spectrum range is traversed, and the X en (ω) and X ar(ω) spectrum line amplitude, when the frequencies corresponding to the maximum values ​​of the two are different, the target spectrum X ar (ω) frequency is corrected to obtain the second target spectrum line X as (ω).

[0089] After obtaining the second target spectrum line, the optimal frequency domain fault component is obtained based on it, and the process is as follows:

[0090] First, the frequency domain fault component spectrum of the second target spectrum and the reference spectrum is differentially extracted. Specifically, based on the reference spectrum X en (ω) and the second target spectrum line X after correction as (ω), extract the frequency domain fault component spectrum by differential extraction, the obtained frequency domain fault component spectrum includes multiple frequency domain fault components, each of which corresponds to a spectrum interval. The specific expression for extracting the frequency domain fault component spectrum is as shown in formula (11), where the fault component X Fault (ω) can be expressed as Its interference component spectrum (including meshing components and noise) X Inter (ω) can be expressed as

[0091] X f (ω) = X as (ω)-X en (ω) (11)

[0092] Afterwards, the spectral line interval of the frequency domain fault component spectrum is extracted. Fault The effectiveness of fault component extraction is evaluated by the spectral line spacing Δf of (ω), and the spectral line spacing can be expressed as:

[0093] △f=[△f1,△f2,△f3,...,△f eff ] (12)

[0094] Where the subscript eff represents the number of spectral lines with non-zero amplitude.

[0095] Finally, the frequency domain fault component corresponding to the spectral line interval distributed near the rotation frequency is used as the optimal frequency domain fault component. The present invention obtains the optimal frequency domain fault component by enumeration method, and the process is as follows:

[0096] The variance between the spectral line spacing and the rotation frequency is calculated in turn, and the variance between the spectral line spacing and the rotation frequency is calculated by (Δf-f z ) is used to describe whether the spectral line spacing is distributed around the rotation frequency.

[0097]

[0098] With the goal of minimizing the variance V, the optimal attenuation coefficient and regularization coefficient are found through enumeration to establish an optimal modulation spectrum line dictionary, and the optimal first target spectrum line, second target spectrum line, frequency domain fault component spectrum line, and spectrum line interval are generated with the optimal modulation spectrum line dictionary, and the frequency domain fault component corresponding to the spectrum line interval distributed near the rotation frequency is used as the optimal frequency domain fault component. Specifically, since the amount of parameter calculation is very small, the attenuation coefficient α and the regularization coefficient λ of the first target spectrum line are traversed and generated through enumeration to minimize the variance V. Based on the optimal attenuation coefficient α opt and the regularization coefficient λ opt Obtain the optimal frequency domain fault component X Fault_opt (ω), the inverse Fourier transform is used to further calculate the time domain fault morphological characteristics, and the final time domain fault morphological characteristics are obtained and output. According to the fault morphological characteristics x extracted in different time periods Fault (t), the fault evolution process of the gear can be observed intuitively. Its expression is as follows:

[0099]

[0100] See also Figures 2 to 9 In order to verify the effect of the method proposed in the present invention, spur gears with different degrees of tooth surface pitting and tooth root cracks were selected for experiments, and the simulation signal data and actual vibration data were analyzed. Figures 2 to 5 The data under pitting fault include the actual vibration signal obtained, the fault morphological characteristics obtained by the method of the present invention, and the established dynamic response. Figures 6 to 9 The data under crack fault include the actual vibration signal obtained, the fault morphological characteristics obtained by the method of the present invention, and the established dynamic response. Among them, the gear faults all occurred on the pinion, the shaft rotation speed was 1500r / min, and the sampling frequency was 12800Hz. According to the Nyquist sampling theorem, the effective frequency range is [0, 6400]; the number of gear teeth z1=29 and z2=100 are specified, the speed of the gear system n=1500r / min is recorded during the test, and the shaft rotation frequency f is further calculated. z =25Hz and gear meshing frequency f m =f z *z1=725Hz. Time series data x(t), whose signal length is N x =500000.

[0101] The main process of obtaining fault morphological characteristics is as follows:

[0102] Step 1: Data segmentation. Divide the vibration signal data x(t) into a reference area and a target area, and divide the reference area and target area signals into segments of length N. p=Signal segment of 100000.

[0103] Step 2: Spectrum preprocessing.

[0104] Step 3: Reconstruct the target spectrum line as the first target spectrum line. In the construction of the modulation spectrum line dictionary, the attenuation coefficient α ranges from 0.2 to 0.8. K is 5. In the shift-invariant frequency domain reconstruction of the target spectrum, the solution method selects the fast soft threshold iterative algorithm (F-ISTA).

[0105] Step 4: Spectral line shift correction: The first target spectral line is converted into the second target spectral line.

[0106] Step 5: Extract fault morphological features. Figure 3 , 5 , 7, 9, according to the fault morphological features x extracted in different time periods Fault (t), the fault evolution process of the gear can be observed intuitively. By comparing the simulation results with the actual signals, it can be obtained that the synchronization is very strong and the actual vibration signal can be well simulated. For fault types in different states, the fault forms obtained are different, which can distinguish the fault types more accurately, conveniently and intuitively.

Claims

1. A gear fault evolution analysis method based on vibration signal morphological characteristics, characterized in that: The method comprises the following steps: Obtain the vibration signal during the gear operation process, and obtain the reference spectrum and target spectrum through data segmentation and spectrum preprocessing; Establishing a modulation spectrum line dictionary to reconstruct the target spectrum line into a first target spectrum line, and correcting the first target spectrum line into a second target spectrum line by using a reference spectrum line; The frequency domain fault component spectra of the second target spectrum and the reference spectrum are differentially extracted to find out the optimal frequency domain fault component of the gear operation and convert it into a time domain fault morphological feature.

2. The gear fault evolution analysis method based on vibration signal morphological characteristics according to claim 1 is characterized in that: The vibration signal is obtained under the working condition of the gear system, and the parameters of the working condition include sampling frequency, effective frequency range, number of gear teeth, gear speed, rotation frequency of the shaft, and gear meshing frequency; wherein the rotation frequency of the gear operation is represented by the rotation frequency of the shaft.

3. The gear fault evolution analysis method based on vibration signal morphological characteristics according to claim 2 is characterized in that: The modulation spectrum dictionary uses the attenuation coefficient as a variable to characterize the fault frequency. The specific formula is as follows: Where, δ(·) represents the impulse function; α represents the modulation spectral line attenuation coefficient, 0 < α < 1; κ represents the number of spectral lines, 1 < κ < K. According to the distribution characteristics of the meshing frequency and rotational frequency of the gear system, the upper limit K is not greater than 0.25*z1, where z1 represents the number of teeth of the target gear; ω represents the independent variable; R + represents the set of positive real numbers; N + represents the set of positive integers; f z represents the rotational frequency of the shaft where the gear is located; D(ω) represents the constructed spectral line dictionary.

4. The gear fault evolution analysis method based on vibration signal morphological characteristics according to claim 3 is characterized in that: In the process of generating the first target spectral line from the modulation spectral line dictionary, shift-invariant sparsity is introduced. The specific process is as follows: First, a reconstruction formula is established based on the modulation line dictionary: X an (ω)=D*s+ε Where ε represents the error term; D represents the modulation spectrum dictionary; s represents the sparse coefficient; X an (ω) represents the target spectral line; Secondly, L1 regularization is introduced, and the regularization coefficient is used as a variable. The reconstruction formula is solved with the minimum error term as the goal to obtain the first target spectrum line. The specific process is as follows: X ar (ω)=D*s Where argmin(·) represents the minimum value of the objective function; ||·||1 represents the L1 norm of the s vector; λ>0 represents the regularization parameter, which is used to control the trade-off between sparsity and error term; X ar (ω) represents the first target spectrum line; m and n represent the position of the element; N and L represent the number of rows and columns of the dictionary.

5. The gear fault evolution analysis method based on vibration signal morphological characteristics according to claim 4 is characterized in that: The process of obtaining the optimal frequency domain fault component is as follows: The acquired frequency domain fault component spectrum includes multiple frequency domain fault components, each of which corresponds to a spectrum line interval, and the spectrum line interval of the frequency domain fault component spectrum is extracted; The frequency domain fault components corresponding to the spectrum line intervals distributed near the rotation frequency are used as the optimal frequency domain fault components.

6. The gear fault evolution analysis method based on vibration signal morphological characteristics according to claim 5 is characterized in that: The process of obtaining the optimal frequency domain fault component by enumeration method is as follows: sequentially calculating the variance between the spectral line spacing and the rotation frequency; With the goal of minimizing the variance, the optimal attenuation coefficient and regularization coefficient are found through enumeration method, thereby obtaining the optimal first target spectrum line, second target spectrum line, frequency domain fault component spectrum line, and spectrum line interval, and the frequency domain fault component corresponding to the optimal spectrum line interval distributed near the rotation frequency is taken as the optimal frequency domain fault component.

7. The gear fault evolution analysis method based on vibration signal morphological characteristics according to claim 1 is characterized in that: Correcting the first target spectral line by using the reference spectral line specifically includes: Extracting peak values ​​of the reference spectrum line and the first target spectrum line; Constructing a window function with the length of the rotation frequency as the minimum window length; The window function is made to traverse the spectrum range by aligning the spectrum lines, and the frequency of the first target spectrum line is corrected based on the reference spectrum line; when the frequencies under the peaks in each window are the same, the second target spectrum line is obtained.

8. The gear fault evolution analysis method based on vibration signal morphological characteristics according to claim 1 is characterized in that: The process of data segmentation is as follows: Processing the vibration signal into time series data with a certain signal length; The stable time series data are divided and aggregated into a reference area signal set, and the unstable time series data are divided and aggregated into a target area signal set; wherein, after each division, the length of the time series data at least includes one gear rotation cycle.

9. The gear fault evolution analysis method based on vibration signal morphological characteristics according to claim 8 is characterized in that: The process of spectrum preprocessing is as follows: Based on fast Fourier transform, the reference area signal set and the target area signal set are converted from the time domain to the frequency domain to obtain a reference area spectrum set and a target area spectrum set; The spectral line value sets corresponding to the reference area spectrum set and the target area spectrum set obey normal distribution. The expectation of each set of spectral line values ​​is used as the amplitude. After normalization, the estimated value of the frequency corresponding to each set of spectral line values ​​is obtained, and the reference spectral lines and target spectral lines are obtained.

10. A gear fault evolution analysis system based on vibration signal morphological characteristics, characterized in that: include: The spectrum acquisition module is used to obtain the vibration signal during the operation of the gear, and obtain the reference spectrum and target spectrum through data segmentation and spectrum preprocessing; A target spectrum line optimization module, used for establishing a modulation spectrum line dictionary to reconstruct the target spectrum line as a first target spectrum line, and correcting the first target spectrum line as a second target spectrum line by using a reference spectrum line; The fault feature extraction module is used to differentially extract the frequency domain fault component spectrum of the second target spectrum and the reference spectrum, find out the optimal frequency domain fault component of the gear operation, and convert it into a time domain fault morphological feature.

Images