Fan bearing fault identification method and system

Abstract

The invention belongs to the technical field of fan bearings, and particularly discloses a fan bearing fault identification method and system.The method comprises the steps that vibration acceleration sensors are arranged at a fan installation base and a motor rear end cover, fan vibration acceleration signals are obtained and filtered, and the cycle frequency alpha of the obtained fan vibration acceleration signals is estimated; finding an optimal filter coefficient; if the convergence condition is met, the kurtosis of the output signal xi is calculated, the optimal filter is obtained by combining the optimal filter coefficient and the optimal filter length, the fan vibration acceleration signal is input into the optimal filter, the filtered deconvolution signal is output, and the envelope amplitude spectrum of the deconvolution signal is obtained based on Hilbert transform demodulation; and comparing with a preset bearing fault characteristic frequency, and diagnosing the fault type of the fan bearing. By adopting the technical scheme, an accurate fan bearing fault diagnosis result is obtained based on improved self-adaptive maximum second-order cyclostationary blind deconvolution.

Description

Technical Field

[0001] The invention belongs to the technical field of fan bearings and relates to a fan bearing fault identification method and system. Background Art

[0002] The cooling fan of the traction motor is one of the key components of high-speed EMUs. Its main function is to provide cooling air for the traction motor and remove the heat generated by the internal windings when the traction motor is running, so as to prevent the traction motor from overheating and damage. During the operation and maintenance process, many cooling fan bearing failures were found, causing the EMU to slow down or stop during the journey, affecting the operation quality of the EMU and the operation order of the trunk railway.

[0003] At present, the inspection and maintenance of the traction system of high-speed EMUs is mainly based on regular maintenance, supplemented by post-maintenance. However, this maintenance method will largely cause some components in the traction system to be "over-maintained" or "under-maintained", resulting in an increase in the operating cost of the traction system or the probability of unexpected failure.

[0004] Therefore, studying the causes of fan bearing failure and taking effective remediation measures are of great significance to improving the quality and safety level of EMUs and ensuring transportation order. Summary of the invention

[0005] The object of the present invention is to provide a fan bearing fault identification method and system to achieve fan bearing fault diagnosis.

[0006] In order to achieve the above object, the basic scheme of the present invention is: a method for identifying a fan bearing fault, comprising the following steps:

[0007] S1, deploy vibration acceleration sensors at the fan installation base and the rear end cover of the motor, and use the vibration acceleration sensors to obtain the fan vibration acceleration signal;

[0008] S2, setting the initial filter length and initial filter coefficient to filter the fan vibration acceleration signal;

[0009] S3, using the envelope product spectrum EHPS, estimating the cycle frequency α of the obtained fan vibration acceleration signal;

[0010] S4, using the cyclic frequency, the second-order cyclostationarity is used as the objective function of deconvolution, and the optimal filter coefficient is found by cyclic iteration; if the convergence condition is met, the output signal x is calculated i kurtosis, and execute step S5, otherwise execute step S3, where x i is the i-th output signal result that meets the convergence condition;

[0011] S5, based on prior knowledge, obtain the energy distribution interval of the fan rotation frequency and its resonance frequency band, set the filter length step value to traverse this interval, and compare the kurtosis value of the signal filtering result of N filters with different lengths. If the filter length is equal to the spectrum interval length, execute step S6, otherwise update the filter length and execute step S3;

[0012] S6, comparing the kurtosis values ​​of the signal filtering results of N filters with different lengths, and finding the filter length corresponding to the maximum kurtosis as the optimal filter length;

[0013] S7, combining the optimal filter coefficient and the optimal filter length to obtain the optimal filter;

[0014] S8, input the fan vibration acceleration signal into the optimal filter, output the filtered deconvolution signal, demodulate based on Hilbert transform to obtain the envelope amplitude spectrum of the deconvolution signal, compare it with the preset bearing fault characteristic frequency, and diagnose the fan bearing fault type.

[0015] The working principle and beneficial effect of this basic scheme are: this technical scheme introduces the kurtosis value to improve the maximum second-order cyclostationary blind deconvolution method of estimating the cyclic frequency of the vibration signal using the envelope product spectrum (EHPS).

[0016] The vibration acceleration sensor is used to obtain the vibration acceleration signal of the fan. Under different length filters, the envelope product spectrum (EHPS) is used to estimate the cyclic frequency of the vibration signal. The second-order cyclostationarity is used as the objective function of deconvolution to iterate and find the optimal filter coefficient. At the same time, the kurtosis value of the signal after filtering with different filter lengths is calculated and compared to find the filter length corresponding to the maximum kurtosis. The global optimal filter coefficient is the optimal filter coefficient under the filter length corresponding to the maximum kurtosis. Then the final signal is output to amplify the fault characteristics of the estimated cyclic frequency, so that the type of fan bearing fault can be diagnosed.

[0017] Furthermore, for filters of different lengths, the envelope product spectrum EHPS is used to estimate the cyclic frequency α of the obtained fan vibration acceleration signal. The specific steps are as follows:

[0018] If a fan vibration acceleration signal is x(t), the harmonic product spectrum is expressed as H(w):

[0019]

[0020] Where F(w) is the amplitude of the Fourier transform of the signal x(t); M is the number of harmonics considered;

[0021] The frequency corresponding to the harmonic product spectrum is the cyclic frequency α. In the harmonic correlation spectrum structure HRSS, the position of the global maximum of EHPS is considered to be the fundamental frequency in HRSS.

[0022] The cycle frequency of vibration acceleration signal is estimated by envelope product spectrum, which is easy to operate.

[0023] Furthermore, the second-order cyclostationarity is used as the objective function of deconvolution, and the method of finding the optimal filter coefficient by cyclic iteration is:

[0024] The recovery of the excitation s0 from the noisy signal x is expressed as:

[0025] s=x*h=(s0*g)*h≈s0

[0026] Where s is the estimated excitation source; g is the unknown impulse response of the transfer path; h is the FIR inverse filter;

[0027] The second-order cyclostationarity level ICS2 is:

[0028]

[0029] Among them, k is the sample index; is the second-order complex statistic for estimating the excitation source; e is a natural constant; j is an imaginary unit, is the second-order statistic for estimating the excitation source; N is the length of the signal, also known as the number of signal points; L is the length of the inverse filter h; T s is the period of fault impact; α is the frequency of fault impact corresponding to the harmonic product spectrum in the fan vibration acceleration signal, that is, the cycle frequency; n is 1, 2, 3, ..., N;

[0030] The criterion h0 is used as the objective function of deconvolution to detect the fault of rotating machinery, and the fault characteristics of the cycle frequency estimated by the envelope product spectrum are amplified, which is expressed as:

[0031]

[0032] The second-order cyclostationarity is used as the objective function of deconvolution to find the optimal filter coefficients for subsequent use.

[0033] Furthermore, the optimal filter coefficients are calculated through the eigenvalue algorithm EVA. The EVA algorithm transforms the unconstrained optimization maximization problem into a generalized eigenvalue problem. The convolution of the discrete time signal is:

[0034]

[0035] Where X is the discretized form of the noise signal x;

[0036] Will and It is expressed as:

[0037]

[0038] E=[e1 … e k … e K ]

[0039] e k =[e -j2πkα(L-1) … e -j2πkα(N-1) ] T

[0040] Where s is the discretized form of the estimated excitation source, and the elements in D are the diagonal elements of the matrix s; s H To estimate the conjugate transpose of the excitation source, diag is a matrix calculation method used to extract or construct the diagonal elements of the matrix; s[l] is the lth element of the diagonal of the matrix s; e k is the matrix composed of all complex exponential forms under the kth sample index; E is the complex statistical matrix; K is the sample index;

[0041] The second-order cyclostationarity is expressed as:

[0042]

[0043] Among them, D H is the conjugate matrix of matrix D, E H is the conjugate matrix of E;

[0044] Substituting s=Xh, we get:

[0045]

[0046] Among them, h H is the transposed form of the filter, R XWX is the weighted correlation matrix, R XX is the correlation matrix, X H The transposed form of the discrete form of the noise signal; W is the weight correlation matrix:

[0047]

[0048] Using the iterative algorithm, we can find the maximum eigenvalue λ and get the optimal filter coefficient h:

[0049] R XWX h=R XX hλ.

[0050] The EVA algorithm transforms the unconstrained optimization maximization problem into a generalized eigenvalue problem, which is convenient for calculation.

[0051] Further, the method for calculating kurtosis is as follows:

[0052] The rotation frequency of the fan shaft is f:

[0053]

[0054] Wherein, v is the rotation speed of the fan shaft;

[0055] Compute kurtosis:

[0056]

[0057] Among them, Kur is the kurtosis; σ x is the standard deviation; x(n) is the input fan vibration acceleration noise signal; is the mean value of the fan vibration acceleration signal; n is 1, 2, 3, …, N, and N is the number of sample points.

[0058] The kurtosis value is introduced to improve the maximum second-order cyclostationary blind deconvolution method for estimating the cyclic frequency of vibration signals using envelope product spectrum (EHPS), which is convenient for use.

[0059] The present invention also provides a fan bearing fault identification system, comprising a data acquisition module and a processing module, wherein the data acquisition module is used to collect the fan vibration acceleration signal, and the output end of the data acquisition module is connected to the input end of the processing module;

[0060] The processing module executes the method of the present invention to identify the fan bearing fault.

[0061] This system uses data acquisition module and processing module to obtain fan vibration acceleration signal, analyze and process it, and realize fan bearing fault identification.

[0062] Furthermore, the data acquisition module includes a vibration acceleration sensor, which is installed on the fan installation base and the rear end cover of the motor.

[0063] Set the installation position of the vibration acceleration sensor to ensure that the collected data is more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 It is a flow chart of the fan bearing fault identification method of the present invention. DETAILED DESCRIPTION

[0065] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0066] In the description of the present invention, it is necessary to understand that the terms "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0067] In the description of the present invention, unless otherwise specified and limited, it should be noted that the terms "installed", "connected" and "connected" should be understood in a broad sense. For example, it can be a mechanical connection or an electrical connection, or it can be the internal connection of two components. It can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to the specific circumstances.

[0068] The present invention discloses a method for identifying fan bearing faults. Based on an improved adaptive maximum second-order cyclostationary blind deconvolution method, the kurtosis value is introduced to improve the maximum second-order cyclostationary blind deconvolution method for estimating the cyclic frequency of a vibration signal using an envelope product spectrum (EHPS), thereby realizing the diagnosis of fan bearing fault types. Figure 1 As shown, the fan bearing fault identification method includes the following steps:

[0069] S1, deploy vibration acceleration sensors at the fan installation base and the rear end cover of the motor, use the vibration acceleration sensors to obtain the fan vibration acceleration signal, and the vibration acceleration sensors transmit the signal through the DEWETRON multi-channel data acquisition device;

[0070] S2, setting the initial filter length and initial filter coefficient (such as the initial filter length L=40; the initial filter coefficient is a column vector containing L zeros), and filtering the fan vibration acceleration signal;

[0071] S3, using the envelope product spectrum EHPS, estimating the cycle frequency α of the obtained fan vibration acceleration signal;

[0072] S4, using the cyclic frequency, the second-order cyclostationarity is used as the objective function of deconvolution, and the optimal filter coefficient is found by cyclic iteration; if the convergence condition is met (convergence error 0.001), the output signal x is calculated i kurtosis, and execute step S5, otherwise execute step S3, where x i is the i-th output signal result that meets the convergence condition;

[0073] S5, based on prior knowledge (calculate the resonant frequency according to the formula: K1 is the elastic coefficient of the system, and m is the mass of the fan system; then take 200 on the left and right as the upper and lower limits respectively; identify the resonance frequency band according to the spectrum of the fan vibration acceleration signal), obtain the fan rotation frequency and its resonance frequency band energy distribution interval, set the filter length step value to traverse this interval, and compare the kurtosis value of the signal filtering result of N filters with different lengths. If the filter length is equal to the spectrum interval length (according to the resonance frequency interval, assume that its interval length is Z, here assume that the filter length increases with ΔL=10, when the filter length increases to L=Z, stop the loop and execute S6), then execute step S6, otherwise update the filter length and execute step S3;

[0074] S6, comparing the kurtosis values ​​of the signal filtering results of N filters with different lengths, and finding the filter length corresponding to the maximum kurtosis as the optimal filter length;

[0075] S7, combining the optimal filter coefficient and the optimal filter length (under the optimal filter length, iteratively selecting the optimal filter coefficient) to obtain the optimal filter;

[0076] S8, input the fan vibration acceleration signal into the optimal filter, output the filtered deconvolution signal, and obtain the envelope amplitude spectrum of the deconvolution signal based on Hilbert transform demodulation (calculated after Hilbert transform demodulation according to the signal processing method), and compare it with the preset bearing fault characteristic frequency to diagnose the fan bearing fault type.

[0077] In a preferred embodiment of the present invention, for filters of different lengths, the envelope product spectrum EHPS is used to estimate the cyclic frequency α of the obtained fan vibration acceleration signal, and the specific steps are:

[0078] If a fan vibration acceleration signal is x(t), the harmonic product spectrum is expressed as H(w):

[0079]

[0080] Where F(w) is the amplitude of the Fourier transform of the signal x(t); M is the number of harmonics considered;

[0081] The frequency corresponding to the harmonic product spectrum is the cyclic frequency α. In the harmonic correlation spectrum structure HRSS, the position of the global maximum of EHPS is considered to be the fundamental frequency in HRSS.

[0082] In a preferred embodiment of the present invention, the second-order cyclostationarity is used as the objective function of deconvolution, and the method of finding the optimal filter coefficient by cyclic iteration is:

[0083] The recovery of the excitation s0 from the noisy signal x is expressed as:

[0084] s=x*h=(s0*g)*h≈s0

[0085] Where s is the estimated excitation source; g is the unknown impulse response of the transfer path; h is the FIR inverse filter;

[0086] The second-order cyclostationarity level ICS2 is:

[0087]

[0088]

[0089] Among them, k is the sample index; is the second-order complex statistic for estimating the excitation source; e is a natural constant; j is an imaginary unit, is the second-order statistic for estimating the excitation source; N is the length of the signal, also known as the number of signal points, n is 1, 2, 3, ..., N; L is the length of the inverse filter h; T s is the period of fault impact; α is the frequency of fault impact corresponding to the harmonic product spectrum in the fan vibration acceleration signal, that is, the cycle frequency;

[0090] The criterion h0 is used as the objective function of deconvolution to detect the fault of rotating machinery, and the fault characteristics of the cycle frequency estimated by the envelope product spectrum are amplified, which is expressed as:

[0091]

[0092] In a preferred embodiment of the present invention, the optimal filter coefficients are calculated by using an eigenvalue algorithm (EVA). The EVA algorithm transforms the unconstrained optimization maximization problem into a generalized eigenvalue problem. The convolution of the discrete time signal is:

[0093]

[0094] Where X is the discretized form of the noise signal x;

[0095] Will and It is expressed as:

[0096]

[0097] E=[e1 … e k … e K ]

[0098] e k =[e -j2πkα(L-1) … e -j2πkα(N-1) ] T

[0099] Where s is the discretized form of the estimated excitation source, and the elements in D are the diagonal elements of the matrix s; s H To estimate the conjugate transpose of the excitation source, diag is a matrix calculation method used to extract or construct the diagonal elements of the matrix; s[l] is the lth element of the diagonal of the matrix s; e k is the matrix composed of all complex exponential forms under the kth sample index; E is the complex statistical matrix; K is the sample index;

[0100] The second-order cyclostationarity is expressed as:

[0101]

[0102] Among them, D H is the conjugate matrix of matrix D, E H is the conjugate matrix of E;

[0103] Substituting s=Xh, we get:

[0104]

[0105] Among them, h H is the transposed form of the filter, R XWX is the weighted correlation matrix, R XX is the correlation matrix, X H The transposed form of the discrete form of the noise signal; W is the weight correlation matrix:

[0106]

[0107] Using the iterative algorithm (each time the eigenvalue is calculated, a corresponding filter coefficient will be output to see whether the eigenvalue meets the convergence condition. If it does, the corresponding filter coefficient will be output, which is the optimal filter coefficient), the maximum eigenvalue λ is found, and the optimal filter coefficient h is obtained:

[0108] R XWX h=R XX hλ.

[0109] In a preferred embodiment of the present invention, the method for calculating the kurtosis is as follows:

[0110] The rotation frequency of the fan shaft is f:

[0111]

[0112] Wherein, v is the rotation speed of the fan shaft, in r / min;

[0113] Compute kurtosis:

[0114]

[0115] Among them, Kur is the kurtosis; σ x is the standard deviation; x(n) is the input fan vibration acceleration noise signal; is the mean value of the fan vibration acceleration signal; n is 1, 2, 3, …, N, and N is the number of sample points.

[0116] For example, the actual measured vibration acceleration signal of the faulty fan bearing is used for verification, where the bearing model is SKF 6310. The fault characteristic frequency at a speed of 3000r / min is calculated according to the corresponding parameters as shown in Table 1:

[0117] Table 1 Characteristic frequency of rolling bearing (speed 3000r / min)

[0118]

[0119] When the filter length L=40, it is combined with the iterated filter coefficients to form an optimal filter, and ACYCBD is performed to verify the inner ring fault of the fan bearing.

[0120] The present invention also provides a fan bearing fault identification system, comprising a data acquisition module and a processing module, wherein the data acquisition module is used to acquire the fan vibration acceleration signal, and the output end of the data acquisition module is electrically connected to the input end of the processing module.

[0121] The processing module executes the method of the present invention to identify the fan bearing fault. The system uses the data acquisition module and the processing module to obtain the fan vibration acceleration signal, and performs analysis and processing to achieve the fan bearing fault identification.

[0122] In a preferred embodiment of the present invention, the data acquisition module includes a vibration acceleration sensor, which is installed (such as welded, bonded, embedded, clamped, etc.) on the fan installation base and the rear end cover of the motor. The installation position of the vibration acceleration sensor is set to ensure that the collected data is more accurate.

[0123] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0124] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

Claims

1. A method for identifying a fan bearing fault, characterized in that: The steps include: S1, deploy vibration acceleration sensors at the fan installation base and the rear end cover of the motor, and use the vibration acceleration sensors to obtain the fan vibration acceleration signal; S2, setting the initial filter length and initial filter coefficient to filter the fan vibration acceleration signal; S3, using the envelope product spectrum EHPS, estimating the cycle frequency α of the obtained fan vibration acceleration signal; S4, using the cyclic frequency, the second-order cyclostationarity is used as the objective function of deconvolution, and the optimal filter coefficient is found by cyclic iteration; if the convergence condition is met, the output signal x is calculated i kurtosis, and execute step S5, otherwise execute step S3, where x i is the i-th output signal result that meets the convergence condition; S5, based on prior knowledge, obtain the fan rotation frequency and its resonance frequency band energy distribution interval, set the filter length step value to traverse this interval, and compare the kurtosis values ​​of the signal filtering results of H filters with different lengths. If the filter length is equal to the spectrum interval length, execute step S6, otherwise update the filter length and execute step S3; S6, comparing the kurtosis values ​​of the signal filtering results of H filters with different lengths, and finding the filter length corresponding to the maximum kurtosis as the optimal filter length; S7, combining the optimal filter coefficient and the optimal filter length to obtain the optimal filter; S8, input the fan vibration acceleration signal into the optimal filter, output the filtered deconvolution signal, demodulate based on Hilbert transform to obtain the envelope amplitude spectrum of the deconvolution signal, compare it with the preset bearing fault characteristic frequency, and diagnose the fan bearing fault type.

2. The fan bearing fault identification method according to claim 1, characterized in that: For filters of different lengths, the envelope product spectrum EHPS is used to estimate the cyclic frequency α of the obtained fan vibration acceleration signal. The specific steps are as follows: If a fan vibration acceleration signal is x(t), the harmonic product spectrum is expressed as H(w): Where F(w) is the amplitude of the Fourier transform of the signal x(t); M is the number of harmonics considered; The frequency corresponding to the harmonic product spectrum is the cyclic frequency α. In the harmonic correlation spectrum structure HRSS, the position of the global maximum of EHPS is considered to be the fundamental frequency in HRSS.

3. The fan bearing fault identification method according to claim 1, characterized in that: Taking the second-order cyclostationarity as the objective function of deconvolution, the method of finding the optimal filter coefficients by cyclic iteration is: The recovery of the excitation s0 from the noisy signal x is expressed as: s=x*h=(s0*g)*h≈s0 Where s is the estimated excitation source; g is the unknown impulse response of the transfer path; h is the FIR inverse filter; The second-order cyclostationarity level ICS2 is: Among them, k is the sample index; is the second-order complex statistic for estimating the excitation source; e is a natural constant; j is an imaginary unit, is the second-order statistic for estimating the excitation source; N is the length of the signal, also known as the number of signal points, n is 1, 2, 3, ..., N; L is the length of the inverse filter h; T s is the period of fault impact; α is the frequency of fault impact corresponding to the harmonic product spectrum in the fan vibration acceleration signal, that is, the cycle frequency; The criterion h0 is used as the objective function of deconvolution to detect the fault of rotating machinery, and the fault characteristics of the cycle frequency estimated by the envelope product spectrum are amplified, which is expressed as:

4. The fan bearing fault identification method according to claim 3, characterized in that: The optimal filter coefficients are calculated by the eigenvalue algorithm EVA. The EVA algorithm transforms the unconstrained optimization maximization problem into a generalized eigenvalue problem. The convolution of the discrete time signal is: s=Xh Where X is the discretized form of the noise signal x; Will and It is expressed as: E=[e1…e k …have been K ] And k =[and -j2πkα(L-1) … And -j2πkα(N-1) ] T Where s is the discretized form of the estimated excitation source, and the elements in D are the diagonal elements of the matrix s; s H To estimate the conjugate transpose of the excitation source, diag is a matrix calculation method used to extract or construct the diagonal elements of the matrix; s[l] is the lth element of the diagonal of the matrix s; e k is the matrix composed of all complex exponential forms under the kth sample index; E is the complex statistical matrix; K is the sample index; The second-order cyclostationarity is expressed as: Among them, D H is the conjugate matrix of matrix D, E H is the conjugate matrix of E; Substituting s=Xh, we get: Among them, h H is the transposed form of the filter, R XWX is the weighted correlation matrix, R XX is the correlation matrix, X H The transposed form of the discrete form of the noise signal; W is the weight correlation matrix: Using the iterative algorithm, we can find the maximum eigenvalue λ and get the optimal filter coefficient h: R XWX h=R XX hλ。 5. The fan bearing fault identification method according to claim 1, characterized in that: The method to calculate kurtosis is as follows: The rotation frequency of the fan shaft is f: Wherein, v is the rotation speed of the fan shaft; Compute kurtosis: Among them, Kur is the kurtosis; σ x is the standard deviation; x(n) is the input fan vibration acceleration noise signal; is the mean value of the fan vibration acceleration signal; n is 1, 2, 3, …, N, and N is the number of sample points.

6. A fan bearing fault identification system, characterized in that: It includes a data acquisition module and a processing module, wherein the data acquisition module is used to collect the vibration acceleration signal of the fan, and the output end of the data acquisition module is connected to the input end of the processing module; The processing module executes the method described in any one of claims 1-5 to identify the fan bearing fault.

7. The fan bearing fault identification system according to claim 6, characterized in that: The data acquisition module comprises a vibration acceleration sensor, which is installed on the fan installation base and the rear end cover of the motor.

Images