A method for diagnosing faults of a multi-channel signal of a rolling bearing and a related device

Abstract

The application provides a rolling bearing multi-channel signal fault diagnosis method and related equipment, which comprises the following steps: collecting original multi-channel vibration signals of a target rolling bearing; performing decomposition and noise reduction on the original multi-channel vibration signals through adaptive multi-variable feature modal decomposition to obtain multi-channel intrinsic modal components; calculating multi-channel comprehensive indexes of the decomposed multi-channel intrinsic modal components, and screening out multi-channel intrinsic modal components with the largest multi-channel comprehensive indexes; analyzing the screened multi-channel intrinsic modal components through full-vector square envelope spectrum, thereby performing fault diagnosis on the rolling bearing and obtaining a rolling bearing fault diagnosis result; compared with the prior art, the application can accurately extract fault characteristic frequencies of the rolling bearing under the interference of the rotation frequency and its multiple frequencies.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of fault diagnosis, in particular to a rolling bearing multi-channel signal fault diagnosis method and related equipment. BACKGROUND

[0002] As one of the most critical and most prone to failure parts in industrial equipment mechanical systems, rolling bearings are long-term in high-load environments. When rolling bearings fail, it may cause accidents and even cause huge economic losses. Therefore, monitoring the running state and diagnosing the fault of rolling bearings is an important task in modern industry. Common rolling bearing faults include inner ring faults, rolling element faults and outer ring faults.

[0003] At present, traditional rolling bearing fault diagnosis methods have good diagnosis results to some extent and have made progress. For example, the application of modal decomposition methods and envelope spectrum technology in rolling bearing fault diagnosis has achieved more significant results, and the rolling bearing multi-channel signal fault diagnosis method combining multivariate modal decomposition methods with holographic spectrum, full spectrum and full vector spectrum has also attracted the attention of many scholars.

[0004] Traditional rolling bearing fault diagnosis methods are mainly based on single-channel vibration signal processing methods such as frequency domain analysis, time domain analysis and wavelet analysis. The method of analyzing and diagnosing single-channel rolling bearing vibration signals through frequency spectrum and envelope spectrum is widely used. However, in the actual working condition of rolling bearings, there may be a large amount of noise, and due to the different positions of the vibration sensors, the collected vibration signals may also have some differences, and the rolling bearing multi-channel vibration signal contains more information. Therefore, it is difficult to detect the running state of the rolling bearing and accurately judge the fault type of the rolling bearing by only using the time-frequency domain analysis method of the single-channel vibration signal of the rolling bearing. SUMMARY

[0005] The present application provides a rolling bearing multi-channel signal fault diagnosis method and related equipment, which aims to improve the accuracy of rolling bearing fault diagnosis.

[0006] To achieve the above purpose, the present application provides a rolling bearing multi-channel signal fault diagnosis method, comprising:

[0007] Step 1, collecting the original multi-channel vibration signal of the target rolling bearing, and pre-processing the original multi-channel vibration signal;

[0008] Step 2, decomposing and denoising the original multi-channel vibration signal by adaptive multivariate characteristic modal decomposition to obtain multi-channel intrinsic modal components;

[0009] Step 3, calculating the multi-channel comprehensive index of the decomposed multi-channel proper mode component, and screening out the multi-channel proper mode component with the largest multi-channel comprehensive index;

[0010] Step 4, analyzing the screened multi-channel proper mode component through the full vector square envelope spectrum, thereby performing fault diagnosis on the rolling bearing and obtaining a rolling bearing fault diagnosis result.

[0011] Further, step 1 comprises:

[0012] By arranging vibration sensors in different directions of the rolling bearing to collect original multi-channel vibration signals thereof;

[0013] The collected data is preprocessed by rounding, and the results are kept to four significant figures.

[0014] Further, the adaptive multi-variable modal decomposition comprises:

[0015] The input multi-channel signal is:

[0016] X(n)=[x1(n),x2(n),…,x o (n)],n=1,2,…,N

[0017] Wherein, x o (n) represents the oth channel signal;

[0018] The basic parameters (f s , C n , I) of the adaptive multi-variable modal decomposition are set, the parameters (L, K) of the adaptive multi-variable modal decomposition are determined by the improved zebra optimization algorithm, f s is the sampling frequency, L is the filter length, C n is the shear number of the oth channel signal, K is the maximum modal number extracted, and I is the maximum iteration number;

[0019] Define the initial iteration i=1 and establish an adaptive FIR filter set, for each channel signal, use M Hanning windows to construct a set of uniformly distributed adaptive FIR filters;

[0020] Obtain the decomposed multi-channel modal component, that is:

[0021]

[0022] Wherein, represents the mth modal component of the oth channel signal decomposed at the ith iteration; X o represents an M-row matrix with the same filter length L; denotes the mth FIR filter of the oth channel signal at the ith iteration, m = 1, 2, …, M; denotes the convolution operator.

[0023] estimate the fault period and update the filter coefficients, the autocorrelation spectrum is defined as follows:

[0024]

[0025] Multivariate EMD determines the fault period by using the autocorrelation spectrum Then, according to the estimated fault period and the decomposed modal components above update the filter coefficients.

[0026] determine whether the maximum number of iterations I is reached. If i = I, terminate the iteration process and continue to the next step; otherwise, return to the decomposition process.

[0027] select the final multichannel modal decomposition component, calculate the correlation coefficient CC between all two adjacent multichannel modal components; by constructing a multichannel correlation coefficient matrix MCC (M x M x o) and locking the two multichannel modal components with the largest correlation coefficient; calculate the correlation kurtosis of the two multichannel modal components; remove the multichannel modal component with smaller correlation kurtosis and make M = M - 1.

[0028] Further, the formula for calculating the correlation coefficient is as follows:

[0029]

[0030] where CC m× ( m-1 ) ×o denotes the correlation coefficient between the mth and m-1th modal components of the oth channel, and denote the mean of the mth and m-1th modal components of the oth channel, respectively.

[0031] When the maximum number of modes K is reached, terminate the iteration process and output the final multivariate signal decomposition result, the K-order multichannel modal U (k) (n) has the following mathematical expression:

[0032]

[0033] where denotes the kth decomposed modal component of the oth channel.

[0034] Further, the formula for calculating the multichannel comprehensive index in step 3 is as follows:

[0035]

[0036] wherein and are respectively k-th modal component of the o-th channel; N o denotes the total number of channels of the multivariate signal; N denotes the data quantity of the multivariate signal; σ is the standard deviation of the k-th modal component; denotes the mean value of the k-th decomposed modal component of the o-th channel; f denotes the fault characteristic frequency; A(f), A(2f), A(3f) are the amplitudes of the first three orders of the fault characteristic frequency at the envelope spectrum; A total denotes the total amplitude of the envelope spectrum.

[0037] The application further provides a fault diagnosis device for a rolling bearing multichannel signal, comprising:

[0038] a collection module, configured to collect multichannel original vibration signals of a target rolling bearing under normal working and different fault states;

[0039] a decomposition module, configured to decompose the multichannel original vibration signals through adaptive multivariate characteristic modal decomposition to obtain multichannel intrinsic modal components;

[0040] a screening module, configured to calculate a multichannel comprehensive index of the decomposed multichannel intrinsic modal components, and to screen an optimal multichannel intrinsic modal component by taking the multichannel comprehensive index as an evaluation index;

[0041] a diagnosis module, configured to generate a full-vector square envelope spectrum of the multichannel intrinsic modal components to perform fault diagnosis and obtain a fault diagnosis result of the target rolling bearing.

[0042] The above scheme of the application has the following beneficial effects:

[0043] The application collects original multichannel vibration signals of a target rolling bearing, and pre-processes the original multichannel vibration signals; the original multichannel vibration signals are decomposed and denoised through adaptive multivariate characteristic modal decomposition to obtain a plurality of multichannel intrinsic modal components; a multichannel comprehensive index of the decomposed multichannel intrinsic modal components is calculated, and a multichannel intrinsic modal component with the largest multichannel comprehensive index is screened out; the screened multichannel intrinsic modal component is analyzed through a full-vector square envelope spectrum, so that the rolling bearing is diagnosed for fault and a fault diagnosis result of the rolling bearing is obtained; compared with the prior art, the adaptive multivariate characteristic modal decomposition can adaptively determine parameters to avoid the problem of manually setting parameters; the full-vector square envelope spectrum can enhance the readability of the frequency spectrum and effectively distinguish the rotation frequency multiple from the fault characteristic frequency, thereby improving the accuracy of fault diagnosis.

[0044] Other beneficial effects of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0045] Figure 1 This is a flowchart illustrating an embodiment of the present invention;

[0046] Figure 2 The diagram shows the results of adaptive multivariable eigenmode decomposition in this embodiment of the invention and the total vector envelope spectrum in the prior art.

[0047] Figure 3 This is a diagram showing the results of adaptive multivariable eigenmode decomposition and total vector squared envelope spectrum in an embodiment of the present invention. Detailed Implementation

[0048] To make the technical problems, solutions, and advantages of this invention clearer, a detailed description will be provided below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0049] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation. Constructing and operating in a specific orientation is not a limitation of this invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0050] Furthermore, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0051] This invention addresses existing problems by providing a method and related equipment for diagnosing multi-channel signal faults in rolling bearings.

[0052] like Figure 1 As shown, an embodiment of the present invention provides a method for diagnosing multi-channel signal faults in rolling bearings, including:

[0053] Step 1: Acquire the raw multi-channel vibration signal of the target rolling bearing and preprocess the raw multi-channel vibration signal;

[0054] Step 2: Decompose and denoise the original multi-channel vibration signal by adaptive multivariate eigenmode decomposition to obtain multi-channel intrinsic mode components;

[0055] Step 3, calculating the multi-channel comprehensive index of the decomposed multi-channel modal component, and screening out the multi-channel modal component with the maximum multi-channel comprehensive index;

[0056] Step 4, analyzing the screened multi-channel modal component through the full vector square envelope spectrum, thereby performing fault diagnosis on the rolling bearing and obtaining the rolling bearing fault diagnosis result.

[0057] Specifically, the original vibration signals of the rolling bearing in the normal state, inner ring fault, rolling element fault, outer ring fault and composite fault and the like working conditions in the "up" and "front" directions are collected by the vibration sensor.

[0058] Specifically, step 1 comprises:

[0059] The original multi-channel vibration signals of the rolling bearing are collected by arranging vibration sensors in different directions of the rolling bearing respectively;

[0060] The collected data is preprocessed by rounding, and the results are kept to four significant digits.

[0061] Specifically, the adaptive multivariate modal decomposition comprises:

[0062] The input multi-channel signal is:

[0063] X(n)=[x1(n),x2(n),…,x o (n)],n=1,2,…,N

[0064] Wherein, x o (n) represents the oth channel signal;

[0065] The basic parameters (f s , C n , I) of the adaptive multivariate modal decomposition are set, the parameters (L, K) of the adaptive multivariate modal decomposition are determined by the improved zebra optimization algorithm, f s is the sampling frequency, L is the filter length, C n is the shear number of the oth channel signal, K is the maximum modal number extracted, and I is the maximum iteration number;

[0066] Define the initial iteration i=1 and establish an adaptive FIR filter set, for the signal of each channel, use M Hanning windows to construct a set of uniformly distributed adaptive FIR filters;

[0067] The decomposed multi-channel modal component is obtained, that is:

[0068]

[0069] where, represents the mth modal component of the oth channel signal decomposition at the ith iteration; X o represents an M row matrix with the same filter length L; represents the mth FIR filter of the oth channel signal at the ith iteration, m = 1, 2,..., M; represents the convolution operation symbol.

[0070] Estimate the fault period and update the filter coefficients, autocorrelation spectrum is defined as follows:

[0071]

[0072] Multivariate empirical mode decomposition determines the fault period by using the autocorrelation spectrum Then, according to the estimated fault period and the above decomposed modal components Update the filter coefficients.

[0073] Determine whether the maximum number of iterations I is reached. If i = I, terminate the iteration process and continue to the next step; otherwise, return to the decomposition process.

[0074] Select the final multichannel modal decomposition component, calculate the correlation coefficient CC between all two adjacent multichannel modal components; by constructing a multichannel correlation coefficient matrix MCC (M x M x o) and locking the two multichannel modal components with the largest correlation coefficient; calculate the correlation kurtosis of the two multichannel modal components; remove the multichannel modal component with smaller correlation kurtosis and make M = M - 1.

[0075] Further, the formula for calculating the correlation coefficient is as follows:

[0076]

[0077] where CC m× ( m-1 ) ×o represents the correlation coefficient between the mth and m-1th modal components of the oth channel, and respectively represent the mean of the mth and m-1th modal components of the oth channel.

[0078] When the maximum number of modes K is reached, terminate the iteration process and output the final multivariate signal decomposition result, the K-order multichannel modal U (k) (n) has the following mathematical expression:

[0079]

[0080] where The kth decomposed modal component of the oth channel.

[0081] Specifically, the multi-channel comprehensive index calculation formula in step 3 is as follows:

[0082]

[0083] Wherein And Respectively The kurtosis and characteristic energy ratio of the kth modal component of the oth channel; N o The total number of channels of the multivariate signal; N represents the data quantity of the multivariate signal; sigma is the standard deviation of the kth modal component; The mean of the kth decomposed modal component of the oth channel; f represents the fault characteristic frequency; A(f), A(2f), A(3f) are the amplitudes of the front three orders of fault characteristic frequencies at the envelope spectrum; A total The total amplitude of the envelope spectrum.

[0084] Specifically, after screening the multi-channel intrinsic modal components obtained by the adaptive multivariate characteristic modal decomposition through the multi-channel comprehensive index in step 3, the analysis result of the optimal multi-channel intrinsic modal component through the full vector square envelope spectrum is as shown in Figure 2 In order to make the experimental results more convincing, the same data is analyzed by the full vector envelope spectrum in the embodiment of the application, and the result is as shown in Figure 3 From Figure 2 And Figure 3 It can be seen that the full vector square envelope spectrum in the embodiment of the application can extract the 5 times frequency (i.e. 5f r ) of the rotation frequency and accurately extract the fault characteristic frequency f i , while the prior art full vector envelope spectrum can only extract the 5 times frequency of the rotation frequency and cannot accurately extract the fault characteristic frequency.

[0085] The embodiment of the application collects the original multi-channel vibration signal of the target rolling bearing; the original multi-channel vibration signal is decomposed and denoised through the adaptive multivariate characteristic modal decomposition to obtain multi-channel intrinsic modal components; the multi-channel comprehensive index of the decomposed multi-channel intrinsic modal components is calculated, and the multi-channel intrinsic modal component with the largest multi-channel comprehensive index is screened out; the screened multi-channel intrinsic modal component is analyzed through the full vector square envelope spectrum, so as to perform fault diagnosis on the rolling bearing and obtain the rolling bearing fault diagnosis result; compared with the prior art, the adaptive multivariate characteristic modal decomposition can adaptively determine the parameters to avoid the problem of manually setting the parameters; the full vector square envelope spectrum can enhance the readability of the frequency spectrum and effectively distinguish the rotation frequency multiple and the fault characteristic frequency, thereby improving the accuracy of fault diagnosis.

[0086] The application further provides a fault diagnosis device for a rolling bearing multi-channel signal, comprising:

[0087] a collection module, configured to collect multi-channel original vibration signals of a target rolling bearing in normal working and different fault states;

[0088] a decomposition module, configured to decompose the multi-channel original vibration signals by adaptive multi-variable feature modal decomposition to obtain multi-channel intrinsic modal components;

[0089] a screening module, configured to calculate multi-channel comprehensive indexes of the decomposed multi-channel intrinsic modal components, and to screen optimal multi-channel intrinsic modal components by taking the multi-channel comprehensive indexes as evaluation indexes;

[0090] a diagnosis module, configured to generate full-vector square envelope spectra of the multi-channel intrinsic modal components to perform fault diagnosis and obtain a fault diagnosis result of the target rolling bearing.

[0091] It should be noted that the information interaction and execution process between the above devices / units are based on the same concept as the method embodiments of the application, and the specific functions and technical effects brought by the method embodiments can be referred to the method embodiments part, which will not be repeated here.

[0092] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above functional units and modules is exemplified, and in actual application, the above functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or software. In addition, the specific names of each functional unit and module are only for easy distinction, and do not limit the protection scope of the application. The specific working process of the units and modules in the system can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0093] The above is the preferred embodiment of the application. It should be noted that for those skilled in the art, without departing from the principles of the application, several improvements and refinements can be made, which should also be considered within the protection scope of the application.

Claims

1. A method for diagnosing multi-channel signal faults in rolling bearings, characterized in that, include: Step 1: Acquire the raw multi-channel vibration signal of the target rolling bearing and preprocess the raw multi-channel vibration signal; Step 2: Decompose and denoise the original multi-channel vibration signal by adaptive multivariate eigenmode decomposition to obtain multi-channel intrinsic mode components; Step 3: Calculate the multi-channel composite index of the decomposed multi-channel intrinsic mode components, and select the multi-channel intrinsic mode components with the largest multi-channel composite index. Step 4: Analyze the selected multi-channel intrinsic mode components using the full vector square envelope spectrum to diagnose the rolling bearing fault and obtain the rolling bearing fault diagnosis results. The adaptive multivariate feature mode decomposition in step 2 includes: Input multi-channel signals, i.e.: X(n)=[x1(n),x2(n),…,x o (n)],n=1,2,…,N Where, x o (n) represents the signal of the o-th channel; Set the basic parameters (f) for adaptive multivariate mode decomposition. s C n The parameters (L, K) of adaptive multivariable mode decomposition are determined by an improved zebra optimization algorithm. s Where L is the sampling frequency, C is the filter length, and L is the sampling frequency. n K is the cutoff number of the o-th channel signal, K is the maximum number of extracted modes, and I is the maximum number of iterations. Define an initial iteration i=1 and establish an adaptive FIR filter bank. For each channel signal, use M Hanning windows to construct a set of uniformly distributed adaptive FIR filters. Obtain the decomposed multi-channel modal components, i.e.: in, X represents the m-th modal component of the o-th channel signal decomposed in the i-th iteration. o This represents an M-row matrix composed of filters of the same length L. This represents the m-th FIR filter for the o-th channel signal in the i-th iteration, where m = 1, 2, ..., M. This represents the convolution operator; Estimating the fault period and updating the filter coefficients, the definition of the autocorrelation spectrum. as follows: Multivariate eigenmode decomposition determines the fault period using autocorrelation spectrum. Then, based on the estimated failure period and the modal components decomposed above... Update the filter coefficients; Determine if the maximum number of iterations I has been reached. If i = I, terminate the iteration process and continue to the next step; otherwise, return to the decomposition process. Select the final multi-channel mode decomposition components, calculate the correlation coefficient CC between all two adjacent multi-channel mode components, construct the multi-channel correlation coefficient matrix MCC (M×M×o) and lock the two multi-channel mode components with the largest correlation coefficient, calculate the correlation kurtosis of the two multi-channel mode components, remove the multi-channel mode components with smaller correlation kurtosis and make M = M-1; Furthermore, the formula for calculating the correlation coefficient is as follows: Among them, CC m×(m-1)×o This represents the correlation coefficient between the m-th and m-1-th modal components of the o-th channel. and These represent the mean values ​​of the m-th and (m-1)-th modal components of the o-th channel, respectively. The iteration process terminates when the maximum number of modes K is reached, and the final multivariable signal decomposition result is output, representing the K-order multichannel mode U. (k) The mathematical expression for (n) is as follows: in, This represents the k-th decomposed mode component of the o-th channel; The expression for the multi-channel composite index (MCI) in step 3 is: in, and They are The ratio of kurtosis to eigenenergy of the k-th modal component in the o-th channel, N o The total number of channels in the multivariable signal is represented by σ, where N represents the number of data points in the multivariable signal, and σ is the standard deviation of the k-th modal component. Let A(f), A(2f), and A(3f) represent the mean of the k-th decomposed mode component of the o-th channel, f represent the fault characteristic frequency, and A(f), A(2f), and A(3f) are the amplitudes of the first three fault characteristic frequencies at the envelope spectrum. total This represents the total amplitude of the envelope spectrum.

2. A fault diagnosis device employing the multi-channel signal fault diagnosis method for rolling bearings as described in claim 1, characterized in that, include: The acquisition module is used to acquire multi-channel raw vibration signals of the target rolling bearing under normal operation and different fault conditions; The decomposition module is used to decompose the multi-channel original vibration signal through adaptive multivariate eigenmode decomposition to obtain multi-channel intrinsic mode components; The filtering module is used to calculate the multi-channel composite index of the decomposed multi-channel intrinsic mode components, and use it as an evaluation index to filter out the optimal multi-channel intrinsic mode components. The diagnostic module is used to generate the full vector square envelope spectrum of multi-channel intrinsic mode components for fault diagnosis, and obtain the fault diagnosis results of the target rolling bearing.

Images