Bearing fault diagnosis device and bearing fault diagnosis method

Abstract

The invention provides a bearing fault diagnosis device and a bearing fault diagnosis method. A bearing fault diagnosis device is provided with: an acoustic signal acquisition unit that acquires operational array acoustic signals by means of a microphone array; the high-pass filtering part is used for carrying out high-pass filtering to obtain an array sound signal after high-pass filtering; a spatial filtering unit that extracts an array sound signal in a target bearing direction by using a beam-forming spatial filter; a cyclic filtering unit that performs cyclic frequency domain filtering on the array sound signal in the target bearing direction and extracts a cyclic autocorrelation function at a bearing fault cyclic frequency; the envelope demodulation part is used for carrying out envelope demodulation on the cyclic autocorrelation function at the fault cyclic frequency of the bearing to obtain the cyclic autocorrelation function at the fault cyclic frequency of the bearing after envelope demodulation; and a fault diagnosis unit that performs spectral analysis on the cycle autocorrelation function at the envelope-demodulated bearing fault cycle frequency, extracts a bearing fault feature frequency, and diagnoses a bearing fault.

Description

Technical Field

[0001] The present invention relates to a bearing fault diagnosis device and a bearing fault diagnosis method, and in particular to a bearing fault diagnosis device and a bearing fault diagnosis method for a scroll compressor. Background Art

[0002] As a new type of positive displacement compressor, scroll compressors offer high energy efficiency, compact size, and high reliability. Bearings are a key component of scroll compressors, and their operating condition directly impacts the compressor's operation and performance. Therefore, monitoring and troubleshooting scroll compressor bearings are crucial for ensuring stable operation.

[0003] For example, Patent Document 1 records a variable frequency scroll compressor fault diagnosis method based on improved VMD and SVM, Patent Document 2 records a reciprocating compressor bearing fault diagnosis method based on improved local mean decomposition, and Patent Document 3 records a rolling bearing fault acoustic diagnosis method based on the time domain equivalent source method, all of which are bearing fault diagnosis methods based on vibration signals.

[0004] Prior art literature

[0005] Patent Literature

[0006] Patent Document 1: China Invention Patent Publication No. CN112733603A

[0007] Patent Document 2: Chinese Invention Patent Publication No. CN105628381A

[0008] Patent Document 3: Chinese Invention Patent Publication No. CN112924176A Summary of the Invention

[0009] Problems to be solved by the invention

[0010] The above existing technical documents all diagnose scroll compressors by measuring the machine vibration signal, performing signal noise reduction and fault feature extraction on the obtained signal, and then making a diagnosis. However, the actual working environment of the scroll compressor is complex, and it is difficult to access the compressor housing to collect the vibration signal. At the same time, it is impossible to determine the fault location and the specific faulty part through the vibration signal. Existing studies regard the bearing fault signal as the main component of the compressor signal, ignoring the problem that various other rotating machines in the working environment of the scroll compressor generate a lot of noise during operation. These noises are difficult to suppress and remove with existing methods, resulting in the inability of existing methods to effectively diagnose bearing faults.

[0011] The purpose of the present invention is to overcome the defects of the above-mentioned prior art that the use of vibration sensors to obtain the operating signals of the scroll compressor bearings requires direct contact with the compressor being tested, which is difficult to collect in a complex production environment. A bearing fault diagnosis device and a bearing fault diagnosis method are provided, which can effectively diagnose scroll compressor bearing faults and determine the specific location of the faulty bearing.

[0012] Technical means to solve the problem

[0013] The bearing fault diagnosis device of the present invention comprises: an acoustic signal acquisition unit, which uses a microphone array to acquire a running array acoustic signal; a high-pass filtering unit, which performs high-pass filtering on the acquired running array acoustic signal to obtain a running array acoustic signal after high-pass filtering; a spatial filtering unit, which uses a beamforming spatial domain filter to extract an array acoustic signal in the target bearing direction from the running array acoustic signal after high-pass filtering; a cyclic filtering unit, which performs cyclic frequency domain filtering on the array acoustic signal in the target bearing direction to extract a cyclic autocorrelation function at the cyclic frequency of the bearing fault; an envelope demodulation unit, which performs envelope demodulation on the cyclic autocorrelation function at the cyclic frequency of the bearing fault to obtain the cyclic autocorrelation function at the cyclic frequency of the bearing fault after envelope demodulation; and a fault diagnosis unit, which performs spectral analysis on the cyclic autocorrelation function at the cyclic frequency of the bearing fault after envelope demodulation to extract the characteristic frequency of the bearing fault and diagnose the bearing fault.

[0014] The bearing fault diagnosis device of the present invention has a cyclic filtering unit that uses the array sound signal in the target bearing direction to calculate the cyclic autocorrelation function of the array sound signal in the target bearing direction, calculates the bearing fault characteristic frequency as the bearing fault cyclic frequency according to the bearing fault type, and uses the cyclic autocorrelation function of the array sound signal in the target bearing direction in combination with the bearing fault cyclic frequency to extract the cyclic autocorrelation function at the bearing fault cyclic frequency.

[0015] The bearing fault diagnosis device of the present invention has a loop filter unit which

[0016]

[0017] Calculate the cyclic autocorrelation function of the array acoustic signal in the target bearing direction, where represents the cyclic autocorrelation function of the signal of the i-th microphone in the microphone array, α represents the cyclic frequency, τ represents the time delay, T represents the cycle period, and p i (t) represents the sound pressure signal collected by the i-th microphone, t represents time, the symbol * represents conjugate transpose, j represents an imaginary unit, and e represents a natural constant.

[0018] The bearing fault diagnosis device of the present invention has a loop filter unit that detects the bearing fault type by

[0019]

[0020]

[0021]

[0022]

[0023] Calculate the bearing fault characteristic frequency, where f o Indicates the bearing outer ring fault frequency, f i Indicates the bearing inner ring fault frequency, f r Indicates the bearing roller failure frequency, f c represents the cage failure frequency, d represents the rolling element diameter, D represents the raceway pitch diameter, α represents the bearing contact angle, z represents the number of rolling elements, and f represents the rotational frequency.

[0024] The bearing fault diagnosis device of the present invention has a beamforming spatial domain filter that calculates the array steering vector of the microphone array based on the array parameters of the microphone array, calculates the signal correlation matrix based on the array acoustic signal after high-pass filtering, calculates the optimal weight vector under linear constrained minimum variance based on the signal correlation matrix and the array steering vector, performs weighted summation based on the optimal weight vector, and obtains the beamformed signal as the array acoustic signal in the target bearing direction.

[0025] The bearing fault diagnosis device of the present invention has a beamforming spatial domain filter.

[0026]

[0027] Calculate the array steering vector, where d is the element spacing, θ is the incident angle, λ is the wavelength, N is the number of elements, the superscript T indicates transpose, j indicates the imaginary unit, and e indicates a natural constant.

[0028] The bearing fault diagnosis device of the present invention has the following optimization criteria for the linear constraint minimum variance:

[0029]

[0030] Among them, R x is the signal correlation matrix, w is the optimal weight vector, and the superscript H represents the conjugate transpose.

[0031] The bearing fault diagnosis device of the present invention has a beamforming spatial domain filter.

[0032]

[0033] Find the optimal weight vector.

[0034] The bearing fault diagnosis device of the present invention has an envelope demodulation unit that calculates the Hilbert transform pair of the cyclic autocorrelation function at the bearing fault cyclic frequency, takes the cyclic autocorrelation function at the bearing fault cyclic frequency as the real part, and takes the Hilbert transform pair as the imaginary part to form an analytical signal, modulo the analytical signal, and obtains the envelope of the cyclic autocorrelation function at the bearing fault cyclic frequency. The envelope is the cyclic autocorrelation function at the bearing fault cyclic frequency after envelope demodulation.

[0035] The bearing fault diagnosis method of the present invention can be used for the above-mentioned bearing fault diagnosis device, and comprises: an acoustic signal acquisition step, which uses a microphone array to acquire a running array acoustic signal; a high-pass filtering step, which performs high-pass filtering on the acquired running array acoustic signal to obtain a running array acoustic signal after high-pass filtering; a spatial filtering step, which uses a beamforming spatial domain filter to extract an array acoustic signal in the target bearing direction for the running array acoustic signal after high-pass filtering; a cyclic filtering step, which performs cyclic frequency domain filtering on the array acoustic signal in the target bearing direction to extract a cyclic autocorrelation function at the bearing fault cyclic frequency; an envelope demodulation step, which performs envelope demodulation on the cyclic autocorrelation function at the bearing fault cyclic frequency to obtain the cyclic autocorrelation function at the bearing fault cyclic frequency after envelope demodulation; and a fault diagnosis step, which performs spectral analysis on the cyclic autocorrelation function at the bearing fault cyclic frequency after envelope demodulation to extract the bearing fault characteristic frequency and diagnose the bearing fault.

[0036] Effects of the Invention

[0037] According to the present invention, there is no need to directly contact the compressor under test, and diagnosis can be performed in a complex compressor working environment; it can suppress the signals received by the microphone array from non-target directions, increase the signals from the target direction, extract the bearing operation sound signal from the spatial domain, and can determine the specific location of the faulty bearing based on the spatial domain information; it can exclude the cyclostationary noise generated by other rotating machinery during operation in a complex working environment, significantly reduce noise interference and improve the signal-to-noise ratio, thereby effectively extracting the fault information component of the target bearing from the collected signal. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 This is a diagram showing the configuration of the bearing fault diagnosis device according to this embodiment.

[0039] Figure 2 4 is a flowchart of the process of the bearing fault diagnosis method that can be used in this embodiment.

[0040] Figure 3 is a schematic diagram of an application example of this embodiment.

[0041] Figure 4 This is the time domain diagram of a single-channel acoustic signal for the application example.

[0042] Figure 5 is the envelope spectrum of the application example. DETAILED DESCRIPTION

[0043] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments described are only a portion of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort should fall within the scope of protection of the present invention.

[0044] Figure 1 The bearing fault diagnosis device 1 includes an acoustic signal acquisition unit 10, a high-pass filter unit 11, a spatial filter unit 12, a loop filter unit 13, an envelope demodulator 14, and a fault diagnosis unit 15.

[0045] The acoustic signal acquisition unit 10 uses a microphone array to collect the running array acoustic signal; the high-pass filtering unit 11 performs high-pass filtering on the collected running array acoustic signal to obtain the running array acoustic signal after high-pass filtering; the spatial filtering unit 12 uses a beamforming spatial domain filter to extract the array acoustic signal in the target bearing direction based on the running array acoustic signal after high-pass filtering; the cyclic filtering unit 13 performs cyclic frequency domain filtering on the array acoustic signal in the target bearing direction to extract the cyclic autocorrelation function at the bearing fault cyclic frequency; the envelope demodulation unit 14 performs envelope demodulation on the cyclic autocorrelation function at the bearing fault cyclic frequency to obtain the cyclic autocorrelation function at the bearing fault cyclic frequency after envelope demodulation; and the fault diagnosis unit 15 performs spectral analysis on the cyclic autocorrelation function at the bearing fault cyclic frequency after envelope demodulation to extract the bearing fault characteristic frequency and diagnose the bearing fault.

[0046] The acoustic signal collection unit 10 comprises a microphone array composed of multiple microphones. The microphone array is oriented toward the scroll compressor and parallel to the main axis of the scroll compressor. The microphone array collects the operating array acoustic signal of the scroll compressor when the scroll compressor is operating at a constant speed. The microphone array is preferably a linear array.

[0047] This embodiment overcomes the defect that the use of vibration sensors to obtain the scroll compressor bearing operation signal requires direct contact with the compressor under test, which is difficult to collect in a complex production environment. This embodiment collects signals from the compressor bearings through acoustic methods, without the need for direct contact with the compressor under test, which greatly relaxes the conditions of use. On the other hand, this embodiment overcomes the defect that the use of microphones to collect bearing operation sound signals is easily affected by irrelevant noises such as environmental noise and mechanical noise, resulting in the bearing operation sound signals in the collected signals being submerged by irrelevant noise, making it difficult to judge the bearing status. This embodiment uses microphone array technology for signal collection, which enhances the signal in the target bearing direction in the collected array sound signal while suppressing noise signals from other directions, effectively improving the signal-to-noise ratio and enhancing the diagnostic effect.

[0048] The high-pass filtering unit 11 performs high-pass filtering on the running array acoustic signal collected by the acoustic signal collecting unit 10 to obtain the running array acoustic signal after high-pass filtering.

[0049] In a compressor's operating environment, ambient noise, mechanical noise, and other noise sources are typically low-frequency. Therefore, low-frequency bearing fault information can be overwhelmed by these noises. Furthermore, due to the bearing's natural frequency, bearing fault information can be modulated to higher frequencies. High-pass filtering the collected operating array acoustic signals can remove low-frequency noise and improve the signal-to-noise ratio.

[0050] The spatial filter unit 12 uses a beamforming spatial domain filter to extract the array acoustic signal in the direction of the target bearing from the high-pass filtered array acoustic signal. The beamforming spatial domain filter uses a linearly constrained minimum variance beamforming algorithm to extract the acoustic signal in the direction of the target bearing from the high-pass filtered array acoustic signal. Linearly constrained minimum variance beamforming is an adaptive beamforming method. This method minimizes the array's output power while maintaining the target signal power, effectively suppressing the power of signals in other directions.

[0051] The beamforming spatial filter is based on the array parameters of the microphone array.

[0052]

[0053] Calculate the array steering vector a(θ) of the microphone array, where d is the element spacing, θ is the angle of incidence, λ is the wavelength, N is the number of elements, the superscript T denotes transpose, j denotes the imaginary unit, and e denotes a natural constant.

[0054] Secondly, the beamforming spatial domain filter runs the array acoustic signal after high-pass filtering to obtain the signal correlation matrix R x .

[0055] Then, the beamforming spatial filter is applied according to the signal correlation matrix Rx and the array steering vector a(θ), calculate the optimal weight vector w under the linear constrained minimum variance. The optimization criterion for the linear constrained minimum variance is

[0056]

[0057] Among them, R x is the signal correlation matrix, w is the optimal weight vector, and the superscript H represents the conjugate transpose. The beamforming spatial domain filter is

[0058]

[0059] Obtain the optimal weight vector w. And according to the optimal weight vector w through

[0060] y(t)=w H x(t)

[0061] Perform weighted summation to obtain the beamformed signal as the array acoustic signal in the target bearing direction, where x(t) = [x1(t), x2(t), …, x N (t)] T , is the array acoustic signal.

[0062] In the above, the beamforming spatial domain filter is used to separate the signals in each direction in the spatial domain for the array acoustic signal after high-pass filtering.

[0063] The cyclic filter unit 13 performs cyclic frequency domain filtering on the array acoustic signal in the target bearing direction to extract the cyclic autocorrelation function at the bearing fault cyclic frequency. The cyclic filter unit 13 performs cyclic frequency domain filtering on the array acoustic signal in the target bearing direction by calculating the cyclic autocorrelation function at the bearing fault cyclic frequency.

[0064] The loop filter unit 13 uses the array acoustic signal in the target bearing direction to

[0065]

[0066] Calculate the cyclic autocorrelation function of the array acoustic signal in the target bearing direction, where represents the cyclic autocorrelation function of the signal of the i-th microphone in the microphone array, α represents the cyclic frequency, τ represents the time delay, T represents the cycle period, and p i (t) represents the sound pressure signal collected by the i-th microphone, t represents time, the symbol * represents conjugate transpose, j represents an imaginary unit, and e represents a natural constant.

[0067] Secondly, the loop filter unit 13 is based on the bearing fault type.

[0068]

[0069]

[0070]

[0071]

[0072] Calculate the bearing fault characteristic frequency as the bearing fault cycle frequency, where f o Indicates the bearing outer ring fault frequency, f i Indicates the bearing inner ring fault frequency, f r Indicates the bearing roller failure frequency, f c represents the cage failure frequency, d represents the rolling element diameter, D represents the raceway pitch diameter, α represents the bearing contact angle, z represents the number of rolling elements, and f represents the rotational frequency.

[0073] Next, the cyclic filter unit 13 uses the cyclic autocorrelation function of the array acoustic signal in the target bearing direction in combination with the bearing fault cyclic frequency to extract the cyclic autocorrelation function at the bearing fault cyclic frequency.

[0074] Bearing fault signals are typical second-order cyclostationary signals. Cyclostationary signals from other rotating machinery in the compressor's operating environment can significantly interfere with the accuracy of fault diagnosis. The cyclic frequency domain filtering used in this embodiment removes interfering noise from the same direction from a cyclic frequency domain perspective, separating cyclostationary signals from different sources and effectively extracting the bearing fault signal in a complex noisy environment.

[0075] The envelope demodulation unit 14 calculates the Hilbert transform pair of the cyclic autocorrelation function at the bearing fault cyclic frequency, takes the cyclic autocorrelation function at the bearing fault cyclic frequency as the real part and the Hilbert transform pair as the imaginary part to form an analytical signal, modulates the analytical signal, and obtains the envelope of the cyclic autocorrelation function at the bearing fault cyclic frequency. The envelope is the cyclic autocorrelation function at the bearing fault cyclic frequency after envelope demodulation.

[0076] The fault diagnosis unit 15 performs fast Fourier transform on the envelope to obtain the envelope spectrum of the cyclic autocorrelation function at the bearing fault cyclic frequency, and extracts the modulation frequency and the higher harmonics of the modulation frequency that meet the bearing fault characteristic frequency as the bearing fault characteristic frequency.

[0077] In the above, the cyclic autocorrelation function is demodulated by envelope analysis, and the bearing fault related information contained in the cyclic autocorrelation function at the bearing fault cyclic frequency is analyzed, so as to diagnose the bearing fault.

[0078] Figure 2FIG2 is a flowchart of a bearing fault diagnosis method that can be used in this embodiment. The bearing fault diagnosis method includes an acoustic signal acquisition step S20, a high-pass filtering step S21, a spatial filtering step S22, a loop filtering step S23, an envelope demodulation step S24, and a fault diagnosis step S25.

[0079] The bearing fault diagnosis method of this embodiment can be used in the above-mentioned bearing fault diagnosis device 1. The bearing fault diagnosis method comprises: an acoustic signal acquisition step S20, which uses a microphone array to acquire a running array acoustic signal; a high-pass filtering step S21, which performs high-pass filtering on the acquired running array acoustic signal to obtain a running array acoustic signal after high-pass filtering; a spatial filtering step S22, which uses a beamforming spatial domain filter to extract an array acoustic signal in the target bearing direction from the running array acoustic signal after high-pass filtering; a cyclic filtering step S23, which performs cyclic frequency domain filtering on the array acoustic signal in the target bearing direction to extract a cyclic autocorrelation function at the cyclic frequency of the bearing fault; an envelope demodulation step S24, which performs envelope demodulation on the cyclic autocorrelation function at the cyclic frequency of the bearing fault to obtain the cyclic autocorrelation function at the cyclic frequency of the bearing fault after envelope demodulation; and a fault diagnosis step S25, which performs spectral analysis on the cyclic autocorrelation function at the cyclic frequency of the bearing fault after envelope demodulation to extract the bearing fault characteristic frequency and diagnose the bearing fault.

[0080] Figure 3 is a schematic diagram of an application example of this embodiment. Figure 4 This is the time domain diagram of a single-channel acoustic signal for the application example. Figure 5 is the envelope spectrum of the application example.

[0081] like Figure 3 As shown, the test device includes a microphone array 33, a microphone bracket 34, and the bearing to be tested is the upper bearing of the scroll compressor 31 mounted on the scroll compressor base 32. The number of array elements 331 of the microphone linear array used in this example is 10, the spacing between each array element 331 is 0.025m, and the distance between the microphone array and the scroll compressor is 0.5m. During the experiment, the recorded running noise of the rotating machinery was played as a simulation of the actual environmental noise, and the collected single microphone time domain sound signal was as follows Figure 4 As shown. The collected array acoustic signal is processed using the device or method in the above embodiment: first, high-pass filtering is performed with a cutoff frequency of 5000Hz, and the optimal weight of the linear constrained minimum variance beamforming is calculated based on the array parameters and relative coordinates. The signal is spatially filtered, and the cyclic autocorrelation function and the bearing fault characteristic frequency of the filtered signal are calculated. The bearing fault characteristic frequency is taken as the cyclic frequency, and the cyclic autocorrelation function at the cyclic frequency is extracted. The cyclic autocorrelation function is envelope demodulated and the result is spectrally analyzed. The envelope spectrum result is as shown below. Figure 5 As shown. Figure 5 Obvious characteristic frequencies of bearing faults can be seen in the image, indicating that the scroll compressor bearing fault was successfully diagnosed in a complex noise environment.

[0082] The bearing fault diagnosis device and bearing fault diagnosis method described above are characterized in that a microphone linear array is placed in parallel near the scroll compressor to be tested to collect the compressor operation sound signal; the collected array sound signal is high-pass filtered to improve the signal-to-noise ratio; the optimal weight of the linear constrained minimum variance beamformer is solved according to the microphone array parameters used, the array output is calculated using the weight, and the signal is spatially filtered; the cyclic autocorrelation function and the bearing fault characteristic frequency of the filtered signal are calculated, and the cyclic autocorrelation function at the cyclic frequency is calculated using the bearing fault characteristic frequency as the cyclic frequency; the cyclic autocorrelation function is envelope demodulated to obtain an envelope spectrum; and the envelope spectrum is spectrally analyzed to extract fault characteristics and diagnose the condition of the tested bearing.

[0083] The bearing fault diagnosis device and bearing fault diagnosis method recorded above use envelope analysis including cyclic frequency domain filtering and beamforming spatial domain filtering to achieve acoustic diagnosis of scroll compressor bearings. They can effectively eliminate irrelevant noise interference in complex noise environments, extract fault feature information, and realize scroll compressor bearing fault diagnosis under non-contact conditions.

[0084] The bearing fault diagnosis device and bearing fault diagnosis method described above can be applied not only to bearing fault diagnosis of scroll compressors, but also to other rotating machines such as motors and generators.

[0085] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and such modifications or substitutions are intended to be within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be subject to the scope of protection of the claims.

[0086] Explanation of symbols

[0087] Bearing fault diagnosis device…1, acoustic signal acquisition unit…10, high-pass filter unit…11, spatial filter unit…12, loop filter unit…13, envelope demodulation unit…14, fault diagnosis unit…15, acoustic signal acquisition step…20, high-pass filter step…21, spatial filter step…22, loop filter step…23, envelope demodulation step…24, fault diagnosis step…25, scroll compressor…31, compressor base…32, microphone array…33, microphone bracket…34, array element…331.

Claims

1. A bearing fault diagnosis device, characterized in that: have: An acoustic signal collection unit, which uses a microphone array to collect operating array acoustic signals; a high-pass filtering unit for performing high-pass filtering on the collected running array acoustic signal to obtain a high-pass filtered running array acoustic signal; A spatial filtering unit, which uses a beamforming spatial domain filter to extract the array acoustic signal in the direction of the target bearing from the array acoustic signal after the high-pass filtering; a cyclic filtering unit, which performs cyclic frequency domain filtering on the array acoustic signal in the direction of the target bearing to extract the cyclic autocorrelation function at the cyclic frequency of the bearing fault; An envelope demodulation unit performs envelope demodulation on the cyclic autocorrelation function at the bearing fault cyclic frequency to obtain the cyclic autocorrelation function at the bearing fault cyclic frequency after envelope demodulation; as well as The fault diagnosis unit performs spectrum analysis on the cyclic autocorrelation function at the bearing fault cyclic frequency after the envelope demodulation, extracts the bearing fault characteristic frequency, and diagnoses the bearing fault.

2. The bearing fault diagnosis device according to claim 1, characterized in that: The cyclic filtering unit calculates the cyclic autocorrelation function of the array acoustic signal in the target bearing direction by using the array acoustic signal in the target bearing direction. Calculate the bearing fault characteristic frequency as the bearing fault cycle frequency according to the bearing fault type, The cyclic autocorrelation function of the array acoustic signal in the target bearing direction is combined with the bearing fault cyclic frequency to extract the cyclic autocorrelation function at the bearing fault cyclic frequency.

3. The bearing fault diagnosis device according to claim 2, characterized in that: The loop filter unit is Calculating the cyclic autocorrelation function of the array acoustic signal in the target bearing direction, in, represents the cyclic autocorrelation function of the signal of the i-th microphone in the microphone array, α represents the cyclic frequency, τ represents the time delay, T represents the cycle period, and p i (t) represents the sound pressure signal collected by the i-th microphone, t represents time, the symbol * represents conjugate transpose, j represents an imaginary unit, and e represents a natural constant.

4. The bearing fault diagnosis device according to claim 2, characterized in that: The loop filter unit is based on the bearing fault type. Calculate the bearing fault characteristic frequency, Among them, f o Indicates the bearing outer ring fault frequency, f i Indicates the bearing inner ring fault frequency, f r Indicates the bearing roller failure frequency, f c represents the cage failure frequency, d represents the rolling element diameter, D represents the raceway pitch diameter, α represents the bearing contact angle, z represents the number of rolling elements, and f represents the rotational frequency.

5. The bearing fault diagnosis device according to claim 1, characterized in that: The beamforming spatial domain filter calculates the array steering vector of the microphone array according to the array parameters of the microphone array, According to the high-pass filtering, the array acoustic signal is run to obtain the signal correlation matrix. Calculating the optimal weight vector under linear constrained minimum variance according to the signal correlation matrix and the array steering vector, A weighted sum is performed according to the optimal weight vector to obtain a beamformed signal as an array acoustic signal in the target bearing direction.

6. The bearing fault diagnosis device according to claim 5, characterized in that: The beamforming spatial filter is Calculate the array steering vector, Where d is the element spacing, θ is the incident angle, λ is the wavelength, N is the number of elements, the superscript T indicates transpose, j indicates the imaginary unit, and e indicates a natural constant.

7. The bearing fault diagnosis device according to claim 6, characterized in that: The optimization criterion for the linear constrained minimum variance is Among them, R x is the signal correlation matrix, w is the optimal weight vector, and the superscript H represents the conjugate transpose.

8. The bearing fault diagnosis device according to claim 7, characterized in that: The beamforming spatial filter is The optimal weight vector is obtained.

9. The bearing fault diagnosis device according to claim 1, characterized in that: The envelope demodulation unit calculates a Hilbert transform pair of a cyclic autocorrelation function at the bearing fault cyclic frequency. The cyclic autocorrelation function at the bearing fault cycle frequency is used as the real part, and the Hilbert transform pair is used as the imaginary part to form an analytical signal. The analytical signal is modulated to obtain the envelope of the cyclic autocorrelation function at the bearing fault cycle frequency, where the envelope is the cyclic autocorrelation function at the bearing fault cycle frequency after envelope demodulation.

10. A bearing fault diagnosis method applicable to the bearing fault diagnosis device of claims 1-9, characterized in that: have: an acoustic signal collecting step, which uses a microphone array to collect an operating array acoustic signal; a high-pass filtering step of performing high-pass filtering on the collected running array acoustic signal to obtain a high-pass filtered running array acoustic signal; A spatial filtering step, which utilizes a beamforming spatial domain filter to extract the array acoustic signal in the direction of the target bearing from the array acoustic signal after the high-pass filtering; a cyclic filtering step of performing cyclic frequency domain filtering on the array acoustic signal in the direction of the target bearing to extract a cyclic autocorrelation function at the cyclic frequency of the bearing fault; An envelope demodulation step of performing envelope demodulation on the cyclic autocorrelation function at the bearing fault cyclic frequency to obtain the cyclic autocorrelation function at the bearing fault cyclic frequency after envelope demodulation; as well as The fault diagnosis step is to perform spectrum analysis on the cyclic autocorrelation function at the bearing fault cyclic frequency after the envelope demodulation, extract the bearing fault characteristic frequency, and diagnose the bearing fault.

Images