Fault diagnosis method, device and electronic equipment

Abstract

The present invention relates to the field of signal processing technology, and specifically to a fault diagnosis method, device and electronic device, wherein the method comprises obtaining a vibration time domain signal of an object to be detected; determining a corresponding signal analysis method based on the category of the object to be detected; analyzing the vibration time domain signal according to the signal analysis method to determine a target spectrum signal; extracting characteristic values ​​from the target spectrum signal, and determining whether the object to be detected has a fault based on the characteristic values. The corresponding signal analysis method is determined according to the category of the object to be detected, that is, there is a corresponding relationship between the category and the signal analysis method, and the vibration time domain signal is analyzed using the corresponding signal analysis method to obtain an accurate target spectrum signal, thereby improving the accuracy of fault diagnosis based on the target spectrum signal.

Description

Technical Field

[0001] The present invention relates to the technical field of signal analysis, and in particular to a fault diagnosis method, device and electronic equipment. Background Art

[0002] Fault diagnosis technology is a technology that detects abnormal conditions of equipment by monitoring the equipment's status parameters, and analyzes and diagnoses the causes of the faults after the abnormal conditions are detected. Its purpose is to use all the current new technologies and achievements to detect potential equipment failures, so as to prevent equipment accidents before they occur. Existing diagnosis is generally performed by collecting vibration signals for analysis. There are many signal analysis methods based on vibration signals, such as spectrum analysis, refined spectrum analysis, inverse spectrum analysis, wavelet transform, etc.

[0003] In the existing fault diagnosis methods, the same signal analysis method is generally used to analyze the vibration signals of different detection objects. For example, the fault diagnosis of shafts and bearings uses the same signal analysis method to extract characteristic values ​​for fault diagnosis. However, since different signal analysis methods have different application scopes, if the same signal analysis method is used to analyze different detection objects, the extracted characteristic values ​​will not meet the requirements of fault diagnosis, resulting in low accuracy of fault diagnosis. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides a fault diagnosis method, device and electronic device to solve the problem of low fault diagnosis accuracy.

[0005] According to a first aspect, an embodiment of the present invention provides a fault diagnosis method, comprising:

[0006] Acquire a vibration time domain signal of the object to be detected;

[0007] Based on the category of the object to be detected, determining a corresponding signal analysis method;

[0008] Analyze the vibration time domain signal according to the signal analysis method to determine the target spectrum signal;

[0009] A feature value is extracted from the target spectrum signal, and based on the feature value, it is determined whether the object to be detected has a fault.

[0010] The fault diagnosis method provided by the embodiment of the present invention determines the corresponding signal analysis method according to the category of the object to be detected, that is, there is a corresponding relationship between the category and the signal analysis method. By analyzing the vibration time domain signal using the corresponding signal analysis method, an accurate target spectrum signal can be obtained, thereby improving the accuracy of fault diagnosis based on the target spectrum signal.

[0011] In combination with the first aspect, in a first implementation manner of the first aspect, analyzing the vibration time domain signal according to the signal analysis method to determine the target spectrum signal includes:

[0012] Acquire attribute parameters of the object to be detected, wherein the attribute parameters include a center frequency and a frequency search bandwidth;

[0013] Performing a qualitative analysis on the vibration time domain signal based on the attribute parameter to obtain a qualitative analysis result to determine whether the object to be detected has a fault, the qualitative analysis result including whether a sideband exists or not;

[0014] An envelope analysis is performed on the vibration time domain signal to determine the target frequency spectrum signal.

[0015] The fault diagnosis method provided in the embodiment of the present invention performs qualitative analysis on the vibration time domain signal to predict the fault, and the quantitative analysis based on envelope analysis can obtain accurate target spectrum signals, thereby combining qualitative analysis with quantitative analysis to ensure the accuracy of fault diagnosis.

[0016] In combination with the first embodiment of the first aspect, in the second embodiment of the first aspect, the qualitative analysis of the vibration time domain signal based on the attribute parameter to obtain a qualitative analysis result to determine whether the object to be detected has a fault includes:

[0017] Calculate the sideband energy value of the vibration time domain signal based on the attribute parameter;

[0018] The sideband energy value is compared with a threshold value to obtain the qualitative analysis result to determine whether the object to be detected has a fault.

[0019] The fault diagnosis method provided by the embodiment of the present invention, since gear eccentricity, tooth meshing wear, broken teeth, cracks or tooth damage in gear faults will cause sidebands to appear at the meshing frequency, therefore, determining whether there are sidebands is an important prerequisite for gear fault diagnosis. By calculating the sideband energy value of the vibration time domain signal, it can be determined whether the object to be detected has a fault, thereby predicting the fault.

[0020] In combination with the second implementation of the first aspect, in the third implementation of the first aspect, the sideband energy value is calculated using the following formula:

[0021]

[0022] In the formula, is the power spectrum density of the vibration time domain signal r(t), r(nT s ) is the time domain sampling value of r(t), T sis the sampling period, f is the center frequency, N is the number of time domain samples, and M is the frequency search bandwidth.

[0023] In combination with the first aspect, in a fourth implementation manner of the first aspect, the vibration time domain signal includes an acceleration signal and an angle signal with equal time intervals, and the analyzing the vibration time domain signal according to the signal analysis method to determine the target spectrum signal includes:

[0024] Determine the relationship between angle and time based on the acceleration signal and the angle signal;

[0025] Based on the relationship between the angle and time, the time corresponding to each interval target angle is calculated to obtain a time series;

[0026] Interpolating the time series, and calculating the acceleration value corresponding to each interval target angle based on the interpolation result;

[0027] The target frequency spectrum signal is determined by performing a transformation based on the acceleration value corresponding to each target angle interval.

[0028] The fault diagnosis method provided by the embodiment of the present invention converts the rotational speed signal into a target spectrum signal corresponding to an angle, and obtains the target spectrum signal corresponding to each interval angle by interpolation. Based on this, every time the object to be detected rotates through an equal angle, a real-time vibration amplitude will be returned, so that the diagnosis method can be applied to occasions with large fluctuations in rotation speed, ensuring that the component information of the rotation frequency and its multiples can be accurately analyzed, thereby improving the accuracy of fault diagnosis.

[0029] In combination with the fourth implementation of the first aspect, in the fifth implementation of the first aspect, interpolating the time series and calculating the acceleration value corresponding to each interval target angle based on the interpolation result includes:

[0030] Interpolating the time series to obtain an interpolated time series;

[0031] Based on the acceleration signal, the acceleration value corresponding to each time in the interpolation time series is determined, so as to obtain the acceleration value corresponding to each interval target angle.

[0032] In combination with the first aspect, in a sixth implementation of the first aspect, extracting a characteristic value of the target spectrum signal and determining whether the object to be detected has a fault based on the characteristic value includes:

[0033] Obtaining a target frequency and a target fault detection model corresponding to the object to be detected;

[0034] Extracting the amplitude corresponding to the target frequency from the target spectrum signal to determine the characteristic value;

[0035] The characteristic value is input into the target fault detection model to determine whether the object to be detected has a fault.

[0036] The fault diagnosis method provided in the embodiment of the present invention uses a target fault detection model corresponding to the object to be detected to perform fault diagnosis. Since the characteristic values ​​input into the target fault detection model have been subjected to signal processing, the reliability of the characteristic values ​​is guaranteed. On this basis, fault detection is performed using the fault diagnosis model, which can improve the accuracy of the diagnosis result.

[0037] According to a second aspect, an embodiment of the present invention provides a fault diagnosis device, including:

[0038] An acquisition module, used for acquiring a vibration time domain signal of an object to be detected;

[0039] A determination module, used to determine a corresponding signal analysis method based on the category of the object to be detected;

[0040] An analysis module, used to analyze the vibration time domain signal according to the signal analysis method to determine a target spectrum signal;

[0041] The detection module is used to extract characteristic values ​​from the spectrum signal and determine whether the object to be detected has a fault based on the characteristic values.

[0042] The fault diagnosis device provided in the embodiment of the present invention determines the corresponding signal analysis method according to the category of the object to be detected, that is, there is a corresponding relationship between the category and the signal analysis method. By analyzing the vibration time domain signal using the corresponding signal analysis method, an accurate target spectrum signal can be obtained, thereby improving the accuracy of fault diagnosis based on the target spectrum signal.

[0043] According to the third aspect, an embodiment of the present invention provides an electronic device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the fault diagnosis method described in the first aspect or any one of the embodiments of the first aspect by executing the computer instructions.

[0044] According to a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the fault diagnosis method described in the first aspect or any one of the embodiments of the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 is a flow chart of a fault diagnosis method according to an embodiment of the present invention;

[0047] Figure 2 is a flow chart of a fault diagnosis method according to an embodiment of the present invention;

[0048] Figure 3a It is a direct spectrum diagram of the vibration time domain signal;

[0049] Figure 3b is a target spectrum diagram of an embodiment of the present invention;

[0050] Figure 4a It is the equal time sampling diagram of the vibration time domain signal;

[0051] Figure 4b yes Figure 4a Schematic diagram of equiangular interpolation of vibration time domain signals;

[0052] Figure 4c yes Figure 4a The equi-angle interpolation order spectrum diagram corresponding to the vibration time domain signal;

[0053] Figure 4d yes Figure 4a Direct spectrum diagram of vibration time domain signal;

[0054] Figure 5 It is a schematic diagram of the three-layer decomposition of wavelet packets;

[0055] Figure 6a as well as Figure 6b They are the spectrum diagrams with and without wavelet decomposition respectively;

[0056] Figure 7 is a flow chart of a fault diagnosis method according to an embodiment of the present invention;

[0057] Figure 8 is a structural block diagram of a fault diagnosis device according to an embodiment of the present invention;

[0058] Fig. 9 It is a schematic diagram of the hardware structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0060] It should be noted that each signal analysis method has its scope of application. For example, wavelet decomposition is applicable to filtering out high-frequency noise interference signals, the equiangular interpolation proposed in the embodiment of the present invention is applicable to eliminating the situation of unstable spindle speed, and the sideband search is applicable to the early fault characteristics of gears and rolling bearings. Based on this, the embodiment of the present invention proposes to determine the corresponding signal analysis method according to the category of the object to be detected.

[0061] The fault diagnosis method provided by the embodiment of the present invention can be applied to rotating equipment, such as shafts, gears, bearings or motors, etc. The specific application scenario can be set accordingly according to actual needs, and no limitation is imposed on it here. For example, many faults of the gearbox can be diagnosed from the rotation frequency or power supply frequency and its multiples. The target spectrum signal is obtained by using the corresponding signal analysis method, and then the characteristic value is extracted to determine whether the gearbox is faulty. Further, the fault type can be determined when a fault exists.

[0062] According to an embodiment of the present invention, a fault diagnosis method embodiment is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0063] In this embodiment, a fault diagnosis method is provided, which can be used for electronic devices such as computers, mobile phones, etc. Figure 1 is a flow chart of a fault diagnosis method according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0064] S11, obtaining a vibration time domain signal of the object to be detected.

[0065] The object to be detected is a rotating device, such as a shaft, a bearing, a gear or a motor. The extraction of the vibration time domain signal can be measured by an acceleration sensor, a velocity sensor or a displacement sensor. For example, a measuring element is set at the detection part of the object to be detected, and its vibration time domain signal is measured. After the vibration time domain signal is measured, it is sent to an electronic device, and accordingly, the electronic device can obtain the vibration time domain signal of the object to be detected.

[0066] The fault diagnosis method can be to perform real-time detection during the working process of the object to be detected, and the measuring element sends the real-time measured vibration time domain signal to the electronic device; the electronic device analyzes it to determine whether it has a fault, so as to timely maintain the object to be detected.

[0067] S12, determining a corresponding signal analysis method based on the type of the object to be detected.

[0068] Different objects to be detected have corresponding signal analysis methods, which are divided according to the categories of the objects to be detected. It should be noted that each category of the object to be detected is not limited to one signal analysis method, but can be two or three, which can be set according to the actual application scenario. The signal analysis methods include wavelet decomposition, equiangular interpolation, sideband search method, etc. Among them, equiangular interpolation is used to convert the speed signal of the object to be detected into an angle signal, so that the fault detection method can be applied to occasions with large fluctuations in speed; the sideband search method is used to perform sideband search on the vibration time domain signal of the object to be detected to predict the fault.

[0069] Furthermore, equi-angle interpolation is used to analyze the frequency components related to the shaft speed; the sideband search method is used for early diagnosis and analysis of rolling bearings and gear wear; wavelet decomposition is used for early filtering and motor fault analysis. Since different types of objects to be detected have different eigenvalues ​​required for different fault forms, the method of extracting eigenvalues ​​is different. For example, Table 1 shows four types of objects to be detected, namely shafts, gears, bearings and motors, and each type of object to be detected has a corresponding signal analysis method.

[0070] Table 1 Objects to be detected and signal analysis methods

[0071]

[0072]

[0073] It should be noted that the data in the frequency column in Table 1 are frequency values ​​used for feature extraction.

[0074] S13, analyzing the vibration time domain signal according to a signal analysis method to determine a target frequency spectrum signal.

[0075] After determining the signal analysis method corresponding to the object to be detected, the electronic device uses the corresponding signal analysis method to analyze the vibration time domain signal to obtain the corresponding target spectrum signal, that is, convert the time domain signal into a spectrum signal.

[0076] For example, wavelet decomposition is to first process the time domain signal and then perform FFT conversion to obtain the corresponding target spectrum signal; equal-angle interpolation is to convert the time domain signal into an angle time domain signal and then perform FFT conversion to obtain the target spectrum signal; the sideband search method is used to perform sideband search on the vibration time domain signal and simultaneously perform envelope analysis on the vibration time domain signal to obtain the target frequency signal.

[0077] This step will be described in detail below.

[0078] S14, extracting characteristic values ​​of the target spectrum signal, and determining whether the object to be detected has a fault based on the characteristic values.

[0079] The target spectrum signal represents the relationship between frequency and amplitude. By using the frequency values ​​in Table 1 to extract the characteristic values ​​of the target spectrum signal and using the extracted characteristic values ​​for fault diagnosis, it can be determined whether the object to be detected has a fault. For example, the extracted characteristic values ​​can be compared with the threshold value; or the characteristic values ​​can be analyzed using a fault detection model to determine whether there is a fault. Of course, other methods can also be used to perform fault diagnosis based on the extracted characteristic values, which can be set accordingly according to actual needs.

[0080] This step will be described in detail below.

[0081] The fault diagnosis method provided in this embodiment determines the corresponding signal analysis method according to the category of the object to be detected, that is, there is a corresponding relationship between the category and the signal analysis method. By analyzing the vibration time domain signal using the corresponding signal analysis method, an accurate target spectrum signal can be obtained, thereby improving the accuracy of fault diagnosis based on the target spectrum signal.

[0082] In this embodiment, a fault diagnosis method is provided, which can be used for electronic devices such as computers, mobile phones, etc. Figure 2 is a flow chart of a fault diagnosis method according to an embodiment of the present invention. Figure 2 As shown, the process includes the following steps:

[0083] S21, obtaining a vibration time domain signal of the object to be detected.

[0084] For details, please see Figure 1 S11 of the illustrated embodiment will not be described in detail here.

[0085] S22, determining a corresponding signal analysis method based on the type of the object to be detected.

[0086] For details, please see Figure 1 S12 of the illustrated embodiment will not be described in detail here.

[0087] S23, analyzing the vibration time domain signal according to a signal analysis method to determine a target frequency spectrum signal.

[0088] When the signal analysis method includes the sideband search method, the specific signal analysis process is as follows, that is, the above S23 includes:

[0089] S231, obtaining attribute parameters of the object to be detected.

[0090] The attribute parameters include a center frequency and a frequency search bandwidth.

[0091] Specifically, when analyzing gear faults, the center frequency is the meshing frequency of the gears, and the gear shaft rotation frequency is used as the frequency search bandwidth; when analyzing bearing faults, the center frequency is the natural frequency of the bearing, and the inner ring rotation frequency is used as the frequency search bandwidth. The natural frequency of the bearing is approximately 1kHz-5kHz, so the frequency with the maximum amplitude corresponding to the horizontal coordinate of 1kHz-5kHz can be found in the frequency deviation as the center frequency.

[0092] S232, performing a qualitative analysis on the vibration time domain signal based on the attribute parameters to obtain a qualitative analysis result to determine whether the object to be detected has a fault.

[0093] The qualitative analysis result includes the presence of sidebands and the absence of sidebands.

[0094] The electronic device uses the acquired attribute parameters to calculate the sideband energy value of the vibration time domain signal, and compares the calculated sideband energy value with the threshold. If the sideband energy value is greater than or equal to the threshold, the qualitative analysis result is that the sideband exists; if the sideband energy value is less than the threshold, the qualitative analysis result is that the sideband does not exist.

[0095] In some optional implementations of this embodiment, the above S232 includes:

[0096] (1) Calculate the sideband energy value of the vibration time domain signal based on the attribute parameters.

[0097] Specifically, the sideband energy value is calculated using the following formula:

[0098]

[0099] In the formula, is the power spectrum density of the vibration time domain signal r(t), r(nT s ) is the time domain sampling value of r(t), T s is the sampling period, f is the center frequency, N is the number of time domain samples, and M is the frequency search bandwidth.

[0100] (2) Compare the sideband energy value with the threshold to obtain a qualitative analysis result to determine whether there is a fault in the object to be detected.

[0101] When the sideband energy value is greater than or equal to the threshold, the qualitative analysis result is determined to be that the sideband exists; otherwise, the qualitative analysis result is determined to be that the sideband does not exist.

[0102] Since gear eccentricity, tooth meshing wear, broken teeth, cracks or tooth damage in gear faults will cause sidebands to appear at the meshing frequency, determining whether there are sidebands is an important prerequisite for gear fault diagnosis. By calculating the sideband energy value of the vibration time domain signal, it is possible to determine whether the object to be detected has a fault, thereby predicting the fault.

[0103] S233, performing envelope analysis on the vibration time domain signal to determine the target frequency spectrum signal.

[0104] When performing envelope analysis on vibration time domain signals, the filter bandwidth is an issue that needs to be considered. Specifically, the bandwidth should be at least three times greater than the highest fault characteristic frequency of the bearing to ensure that at least the third harmonic of the highest fault characteristic frequency can be seen in the resonance demodulation envelope spectrum. Generally, among the fault characteristic frequencies of rolling bearings, the inner ring fault characteristic is the largest, so three times the inner ring fault characteristic frequency is usually used as the minimum value of the bandpass filter bandwidth.

[0105] The method used by electronic equipment to perform envelope analysis on vibration time domain signals can be Hilbert envelope or other methods. There is no limitation on it here, and it can be set accordingly according to actual needs. Taking Hilbert envelope as an example, Hilbert envelope is the envelope of the absolute value of the time domain signal. It can be used to extract amplitude modulation signals (relative to high-frequency natural frequency signals, modulation signals are generally low-frequency signals), which has great advantages in extracting fault features. Assume x(t) is a continuous time signal, and its Hilbert transform is:

[0106]

[0107] From x(t) and The analytical signal is formed, namely:

[0108]

[0109] The amplitude of the analytical signal is:

[0110]

[0111] The phase of the analytical signal is:

[0112]

[0113] From the above transformation, we can see that the Hilbert transform of x(t) is to shift its phase by π / 2, while the others remain unchanged. The real part of the analytical signal is x(t) itself, and the imaginary part is the Hilbert transform. The Hilbert envelope is the amplitude of the analytical signal. By performing FFT on the signal after the Hilbert envelope, a clearer demodulation spectrum can be obtained, which is the target spectrum signal.

[0114] In the rolling bearing vibration test, the shaft rotation frequencies are 158.2Hz, 316.4Hz and 473.1Hz (the three frequencies are the inner ring fault characteristic frequency, its second frequency and third frequency respectively). Figure 3a As shown, the spectrum of the sideband search method is as follows Figure 3b Take the maximum amplitude of 2919Hz as the center frequency, meet the judgment conditions of the sideband, judge that there is a sideband, and then perform bandpass filtering (the frequency band range is around 2919Hz and 500Hz), and finally get the envelope as follows Figure 3b Results shown.

[0115] contrast Figure 3a as well as Figure 3b For each target frequency in the spectrum, the amplitude corresponding to the target frequency of 158.2 Hz, the value of the sideband search method is 10 times that of the direct spectrum diagram, which greatly improves the characteristic value of the signal.

[0116] When the signal analysis method includes equal-angle interpolation, the vibration time domain signal includes acceleration signals and angle signals with equal time intervals, and the above S23 also includes:

[0117] (1) Based on the acceleration signal and the angle signal, determine the relationship between the angle and time.

[0118] Among them, the acceleration signal Acc(t) can be collected by the acceleration sensor, and the angle signal is characterized by the key phase sensor. Every time the object to be detected rotates one circle, the key phase sensor measures a pulse signal, and the angle corresponding to a pulse signal is 360°, so the detection using the key phase sensor is equivalent to equal angle sampling.

[0119] For example, assuming that the relationship between the angle θ and the time t of the object to be detected is expressed by the following formula:

[0120] θ(t)=at 3 +bt 2 +ct+d

[0121] To determine the four parameters a, b, c, and d in the above formula, four known points must be available. For example, 0, 360, 720, and 1080 degrees are used as known points to obtain the relationship between the shaft rotation angle and time. Among them, 360° corresponds to one pulse of the key phase sensor, 720° corresponds to two pulses of the key phase sensor, and 1080° corresponds to three pulses of the key phase sensor. Substituting the above four angles and the corresponding time into the above formula, the four parameters a, b, c, and d can be determined.

[0122] (2) Based on the relationship between angle and time, the time corresponding to each interval target angle is calculated to obtain a time series.

[0123] After the relationship between angle and time is determined, equal angle sampling is performed to obtain the time corresponding to each target angle interval. For example, a circle of 360° is divided into 60 equal parts, and each interval of 6° is substituted into the above formula to obtain the corresponding time t value. Then, the electronic device can obtain the time corresponding to 0°, 6°, 12°, 18°, ..., 360°, respectively, and obtain the time series.

[0124] (3) Interpolate the time series and calculate the acceleration value corresponding to each interval target angle based on the interpolation result.

[0125] The electronic device performs interpolation processing on the time series, for example, using cubic spline interpolation to obtain an interpolated time series; and then determines the acceleration value corresponding to each interval target angle according to the acceleration signal.

[0126] As an optional implementation of this embodiment, step (3) of S23 above may include:

[0127] 3.1) Interpolate the time series to obtain the interpolated time series.

[0128] 3.2) Based on the acceleration signal, determine the acceleration value corresponding to each time in the interpolation time series, so as to obtain the acceleration value corresponding to each interval target angle.

[0129] Specifically, the time series is interpolated to the acceleration signal Acc(t) using cubic spline to obtain θ_Acc(θ).

[0130] (4) Transform the acceleration value corresponding to each target angle interval to determine the target spectrum signal.

[0131] By performing FFT transformation on the obtained sequence θ_Acc(θ), the target spectrum signal can be determined.

[0132] The speed signal is converted into a target spectrum signal corresponding to the angle, and the target spectrum signal corresponding to each interval angle is obtained by interpolation. Based on this, every time the detected object rotates through an equal angle, a real-time vibration amplitude will be returned, so that the diagnostic method can be applied to occasions with large fluctuations in speed, ensuring that the component information of the rotation frequency and its multiples can be accurately analyzed, thereby improving the accuracy of fault diagnosis.

[0133] In a specific example, assume that the angular acceleration of the motor is 180° / s 2 , the sampling frequency is 100Hz, the sampling time is 12s, and the analog signal is: 2sind(90t 2 )+sind(180t 2 ). When collecting data with equal angle interpolation, a high sampling frequency acceleration sensor, a key phase sensor and a data collector are required. Specifically, as shown in Table 2:

[0134] Table 2 Hardware configuration required for equal angle interpolation

[0135]

[0136] By performing equiangular interpolation processing on the measured vibration time domain signal, the following is obtained: Figure 4a The equal time sampling diagram shown in Figure 4b The equi-angle interpolation diagram shown in Figure 4c The equi-angle interpolation order spectrum shown in Figure 4d The direct spectrum plot is shown.

[0137] Since the speed is changing during the motor acceleration process, Figure 4d It is impossible to express its meaning, Figure 4c It can be clearly concluded that the 1 times frequency conversion has an amplitude of 2, and the 2 times frequency conversion has an amplitude of 1, which is consistent with the analog signal information.

[0138] When the signal analysis method includes wavelet decomposition, wavelet decomposition, as a time-frequency processing method for non-stationary signals based on time and scale, has good localization characteristics in both time and frequency domains. The advantages of wavelet decomposition make it suitable for processing non-stationary signals and an effective tool for gearbox fault diagnosis.

[0139] φ j,k (t) = 2 j / 2 φ(2 j tk) (1)

[0140] ψ j,k (t) = 2 j / 2 ψ(2 j tk) (2)

[0141] Formula (1) and formula (2) are wavelet function and scaling function respectively. After the original signal is convolved with them, the low-frequency component and high-frequency component of the original signal can be obtained. In addition, it is necessary to select a suitable wavelet basis function. Different wavelet basis functions have different frequency responses. Since wavelet decomposition is divided into two, when decomposing 3 layers, the decomposition diagram is as follows: Figure 5 Assume that the sampling frequency is 2048 Hz, and eight signal components are obtained after the third layer of decomposition. The bandwidth of each signal component is 128 Hz. At this time, the effective value of each signal component is obtained as the vibration characteristic value.

[0142] During the test, a speed sensor was installed on the reducer housing of the gearbox. The sampling frequency was 2048 Hz. The frequency deviation signals after and without wavelet decomposition were as follows: Figure 6a as well as Figure 6b shown.

[0143] Specifically, the meshing frequency of the gears of the gearbox is shown in Table 3:

[0144] Table 3 Frequency domain characteristic parameters during normal operation

[0145] Characteristic parameters Vertical mill main frequency Cylindrical gear meshing frequency Double frequency Bevel gear meshing frequency Double frequency Frequency value / Hz 48.5 211 422 462.5 925

[0146] contrast Figure 6a and Figure 6b ,The amplitude of the cylindrical gear at 1 times the frequency of 21Hz is 1.25, but at the 1 times the frequency of the bevel gear meshing, the amplitude after wavelet decomposition is 0.43, and the amplitude without wavelet processing is 0.37. Since the amplitudes of 211Hz and 462Hz are both important characteristic values ​​for judging gear faults, for the case of small amplitudes, the amplitude of wavelet decomposition is 16% larger than that of non-wavelet decomposition, which can more realistically show the true value, and the characteristic is for the case of small amplitudes.

[0147] Three signal analysis methods are described above. Specifically, equiangular interpolation is used to analyze the frequency components related to the shaft speed, the sideband search method is used for early diagnosis analysis such as early gear wear of rolling bearings, and wavelet analysis is used for early filtering and motor fault analysis.

[0148] S24, extracting characteristic values ​​of the target spectrum signal, and determining whether the object to be detected has a fault based on the characteristic values.

[0149] For details, please see Figure 1 S14 of the illustrated embodiment will not be described in detail here.

[0150] The fault diagnosis method provided in this embodiment performs qualitative analysis on the vibration time domain signal to predict the fault, and the quantitative analysis based on envelope analysis can obtain accurate target spectrum signals, combining qualitative analysis with quantitative analysis to ensure the accuracy of fault diagnosis.

[0151] In this embodiment, a fault diagnosis method is provided, which can be used for electronic devices such as computers, mobile phones, etc. Figure 7 is a flow chart of a fault diagnosis method according to an embodiment of the present invention. Figure 7 As shown, the process includes the following steps:

[0152] S31, obtaining a vibration time domain signal of the object to be detected.

[0153] For details, please see Figure 1 S11 of the illustrated embodiment will not be described in detail here.

[0154] S32: Determine a corresponding signal analysis method based on the type of the object to be detected.

[0155] For details, please see Figure 1 S12 of the illustrated embodiment will not be described in detail here.

[0156] S33, analyzing the vibration time domain signal according to a signal analysis method to determine a target frequency spectrum signal.

[0157] For details, please see Figure 2 S23 of the illustrated embodiment will not be described in detail here.

[0158] S34, extracting characteristic values ​​of the target spectrum signal, and determining whether the object to be detected has a fault based on the characteristic values.

[0159] Specifically, the above S34 includes:

[0160] S341, obtaining a target frequency and a target fault detection model corresponding to the object to be detected.

[0161] The target frequency is the frequency value that is focused on during fault detection. Different types of objects to be detected correspond to different target frequencies under different faults, as shown in Table 1.

[0162] The target fault detection model is an algorithm based on statistics or machine learning that can optimize the structure and parameters of the predictive model according to the defined objective function, thereby memorizing the information in the data and making judgments on similar data. The information recorded by the fault detection model can be the mode of multidimensional data, its similarity to a certain state, or the correlation between input features.

[0163] S342, extracting the amplitude corresponding to the target frequency from the target spectrum signal to determine the characteristic value.

[0164] After determining the target spectrum signal, the electronic device extracts the amplitude from the target spectrum signal using the target frequency value to obtain the corresponding characteristic value. For example, when the target frequency is 1 times the frequency, the amplitude corresponding to 1 times the frequency is extracted from the target spectrum signal.

[0165] S343, inputting the characteristic value into the target fault detection model to determine whether there is a fault in the object to be detected.

[0166] Specifically, the input of the fault detection model is the feature value, and the output is the predicted fault. As shown in Table 1, Table 1 shows 4 categories of objects to be detected, and each object to be detected has a corresponding target fault detection model, so there are 4 corresponding target fault detection models.

[0167] The fault diagnosis method provided in this embodiment uses a target fault detection model corresponding to the object to be detected to perform fault diagnosis. Since the characteristic values ​​input into the target fault detection model have been subjected to signal processing, the reliability of the characteristic values ​​is guaranteed. On this basis, fault detection is performed using the fault diagnosis model, which can improve the accuracy of the diagnosis result.

[0168] In the present embodiment, a fault diagnosis device is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware of a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware is also possible and conceived.

[0169] This embodiment provides a fault diagnosis device, such as Figure 8 As shown, including:

[0170] An acquisition module 41 is used to acquire a vibration time domain signal of an object to be detected;

[0171] A determination module 42, configured to determine a corresponding signal analysis method based on the type of the object to be detected;

[0172] An analysis module 43 is used to analyze the vibration time domain signal according to the signal analysis method to determine a target spectrum signal;

[0173] The detection module 44 is used to extract characteristic values ​​from the frequency spectrum signal and determine whether the object to be detected has a fault based on the characteristic values.

[0174] The fault diagnosis device provided in this embodiment determines the corresponding signal analysis method according to the category of the object to be detected, that is, there is a corresponding relationship between the category and the signal analysis method. By analyzing the vibration time domain signal using the corresponding signal analysis method, an accurate target spectrum signal can be obtained, thereby improving the accuracy of fault diagnosis based on the target spectrum signal.

[0175] The fault diagnosis device in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC circuit, a processor and a memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0176] The further functional description of each of the above modules is the same as that of the above corresponding embodiments and will not be repeated here.

[0177] The embodiment of the present invention also provides an electronic device having the above Figure 8 The fault diagnosis device shown.

[0178] See also Fig. 9 , Fig. 9 is a schematic diagram of the structure of an electronic device provided by an optional embodiment of the present invention, such as Fig. 9 As shown, the electronic device may include: at least one processor 51, such as a CPU (Central Processing Unit), at least one communication interface 53, a memory 54, and at least one communication bus 52. The communication bus 52 is used to realize the connection and communication between these components. The communication interface 53 may include a display screen (Display) and a keyboard (Keyboard), and the optional communication interface 53 may also include a standard wired interface and a wireless interface. The memory 54 may be a high-speed RAM memory (Random Access Memory) or a non-volatile memory (non-volatile memory), such as at least one disk storage. The memory 54 may optionally be at least one storage device located away from the aforementioned processor 51. The processor 51 may be combined with Figure 8 In the described device, the memory 54 stores an application program, and the processor 51 calls the program code stored in the memory 54 to execute any of the above method steps.

[0179] The communication bus 52 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus. The communication bus 52 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Fig. 9 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0180] Among them, the memory 54 may include a volatile memory (English: volatile memory), such as a random access memory (English: random-access memory, abbreviated: RAM); the memory may also include a non-volatile memory (English: non-volatile memory), such as a flash memory (English: flash memory), a hard disk drive (English: hard disk drive, abbreviated: HDD) or a solid-state drive (English: solid-state drive, abbreviated: SSD); the memory 54 may also include a combination of the above types of memory.

[0181] The processor 51 may be a central processing unit (CPU), a network processor (NP), or a combination of a CPU and a NP.

[0182] The processor 51 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.

[0183] Optionally, the memory 54 is also used to store program instructions. The processor 51 can call the program instructions to implement the present application Figure 1 , 2And the fault diagnosis method shown in 7 embodiments.

[0184] The embodiment of the present invention also provides a non-transitory computer storage medium, wherein the computer storage medium stores computer executable instructions, and the computer executable instructions can execute the fault diagnosis method in any of the above method embodiments. The storage medium can be a disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk (HDD) or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above types of memory.

[0185] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A fault diagnosis method, characterized in that: include: Acquire a vibration time domain signal of the object to be detected; Based on the category of the object to be detected, determining a corresponding signal analysis method; Different objects to be detected have corresponding signal analysis methods, which are divided according to the categories of the objects to be detected, wherein each category of the objects to be detected is not limited to only one signal analysis method; The signal analysis methods include wavelet decomposition, equiangular interpolation and sideband search methods; Among them, equal angle interpolation is used to convert the speed signal of the object to be detected into an angle signal, so that the fault diagnosis method is suitable for occasions with large fluctuations in speed; The sideband search method is used to search the sidebands of the vibration time domain signal of the object to be detected and predict the fault; Equiangular interpolation is used to analyze the frequency components related to the shaft speed; The sideband search method is used for early diagnosis and analysis of rolling bearing and gear wear; Wavelet decomposition is used for early filtering and motor fault analysis; Different types of objects to be detected have different fault forms and require different feature values, and the methods of extracting feature values ​​are different. That is, shafts, gears, bearings and motors, each type of object to be detected has a corresponding signal analysis method; The vibration time domain signal is analyzed according to the signal analysis method to determine the target spectrum signal. The specific steps are as follows: When the signal analysis method includes the sideband search method, the specific signal analysis process is as follows: Acquire attribute parameters of the object to be detected, wherein the attribute parameters include a center frequency and a frequency search bandwidth; Based on the attribute parameters, the vibration time domain signal is qualitatively analyzed to obtain a qualitative analysis result to determine whether the object to be detected has a fault, and the qualitative analysis result includes the presence of a sideband and the absence of a sideband; the specific process is as follows: Calculate the sideband energy value of the vibration time domain signal based on the attribute parameter; Compare the sideband energy value with a threshold value to obtain the qualitative analysis result, and determine whether the object to be detected has a fault; when the sideband energy value is greater than or equal to the threshold value, determine that the qualitative analysis result is that a sideband exists; otherwise, determine that the qualitative analysis result is that no sideband exists; Since gear eccentricity, tooth meshing wear, broken teeth, cracks or tooth damage in gear faults will cause sidebands to appear at the meshing frequency, determining whether there are sidebands is an important prerequisite for gear fault diagnosis. By calculating the sideband energy value of the vibration time domain signal, it can be determined whether the object to be detected has a fault, thereby predicting the fault; Performing envelope analysis on the vibration time domain signal to determine the target frequency spectrum signal; When the signal analysis method includes equal angle interpolation, the vibration time domain signal includes acceleration signals and angle signals with equal time intervals. The vibration time domain signal is analyzed according to the signal analysis method to determine the target spectrum signal. The specific process is as follows: Determine the relationship between angle and time based on the acceleration signal and the angle signal; Based on the relationship between the angle and time, the time corresponding to each interval target angle is calculated to obtain a time series; Interpolating the time series, and calculating the acceleration value corresponding to each interval target angle based on the interpolation result, including: interpolating the time series to obtain an interpolated time series; Based on the acceleration signal, determine the acceleration value corresponding to each time in the interpolation time series, so as to obtain the acceleration value corresponding to each interval target angle; Transforming based on the acceleration value corresponding to each target angle interval to determine the target spectrum signal; Extracting characteristic values ​​of the target spectrum signal and determining whether the object to be detected has a fault based on the characteristic values, the specific process is as follows: Obtaining a target frequency and a target fault detection model corresponding to the object to be detected; Extracting the amplitude corresponding to the target frequency from the target spectrum signal to determine the characteristic value; The characteristic value is input into the target fault detection model to determine whether the object to be detected has a fault.

2. An electronic device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the fault diagnosis method according to claim 1 by executing the computer instructions.

3. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the fault diagnosis method according to claim 1.

Images