Bearing fault recognition method based on improved one-dimensional local binary pattern and information fusion

Abstract

The application provides a bearing fault recognition method based on improved one-dimensional local binary pattern and information fusion, comprising: obtaining a vibration signal of a device where the bearing is located; calculating an autocorrelation function of the vibration signal; quantizing the autocorrelation function and reconstructing the signal by using an improved one-dimensional local binary pattern method to obtain a one-dimensional texture feature signal, wherein the improved one-dimensional local binary pattern method takes the global root mean square of the signal as a quantization threshold; fusing the vibration signal and the one-dimensional texture feature signal to obtain a fusion signal; performing spectral analysis on the fusion signal and calculating a bearing fault characteristic frequency, and judging the bearing state and the fault type through the obtained frequency component and the bearing fault characteristic frequency. The bearing fault diagnosis method can accurately extract prominent and rich characteristic frequencies corresponding to bearing faults in a complex noise environment, and realizes accurate judgment of bearing compound fault types.

Description

Technical Field

[0001] This invention belongs to the field of fault diagnosis technology, and specifically relates to a method for identifying rolling bearing faults, which is a bearing fault identification method based on improved one-dimensional local binary pattern and information fusion. Background Technology

[0002] Rolling bearings are critical components of rotating machinery, and their operating condition directly affects the safety and reliability of the equipment. Therefore, health monitoring and fault diagnosis of bearings are of great significance. When a bearing fails, it generates periodic impact signals. Different faults correspond to different characteristic frequencies. By analyzing the signals and extracting the fault characteristic frequencies, the fault type can be identified.

[0003] Existing one-dimensional local binary pattern (1D-LBP) typically uses the local center value of the window as the quantization threshold, focusing only on local signal information and failing to fully consider the overall characteristics and impact patterns of the fault signal, resulting in insufficient fault feature extraction. Furthermore, 1D-LBP research is concentrated on pattern recognition, with a significant lack of studies combining it with other signal processing methods. This leads to incomplete information representation in signal analysis, making it difficult to meet the demands of high-precision diagnosis.

[0004] Therefore, how to improve the one-dimensional local binary model method to meet the engineering requirements of efficient bearing fault identification has become an urgent problem to be solved. Summary of the Invention

[0005] In view of this, the present invention proposes a bearing fault identification method based on improved one-dimensional local binary pattern and information fusion to solve the problems existing in the prior art.

[0006] The bearing fault identification method based on improved one-dimensional local binary pattern and information fusion provided by this invention includes:

[0007] Obtain vibration signals from the equipment where the bearing is located. ],in, The vibration signal Length;

[0008] Calculate the vibration signal The autocorrelation function is obtained to obtain the autocorrelation function. ],in, The autocorrelation function Length;

[0009] The autocorrelation is analyzed using an improved one-dimensional local binary model method. Quantization and signal reconstruction are performed to obtain the vibration signal. Corresponding one-dimensional texture feature signal The improved one-dimensional local binary mode method uses the global root mean square of the signal as the quantization threshold.

[0010] The vibration signal With the one-dimensional texture feature signal Information fusion is performed to obtain a fused signal with more significant fault characteristics;

[0011] The fused signal is subjected to spectral analysis, and the bearing fault characteristic frequency is calculated based on the bearing geometric parameters and rotational speed. The bearing condition and fault type are determined by the obtained frequency components and the bearing fault characteristic frequency.

[0012] Preferably, the autocorrelation function The calculation formula is as follows:

[0013] ;

[0014] In the formula, The vibration signal Length, For time delay, .

[0015] Further optimization, .

[0016] Further optimization involves using an improved one-dimensional local binary model method to analyze the autocorrelation. The steps for quantization and signal reconstruction are as follows:

[0017] Set the sliding window parameters, which include the length of the sliding window and the sliding step size;

[0018] Calculate the autocorrelation function global root mean square value ;

[0019] Using the sliding window, the autocorrelation function is traversed point by point with a step size of 1. With the global root mean square The signal within each window is quantized and reconstructed using a threshold value to obtain the local texture feature value of the signal within each window.

[0020] Arrange the local texture feature values ​​corresponding to all signals within the window in the order in which the sliding window slides, to obtain the vibration signal. Corresponding one-dimensional texture feature signal .

[0021] Further optimization, the first The methods for quantization and signal reconstruction of the signal within a window are as follows, where, , The length of the sliding window:

[0022] Calculate the first Each element in the signal within the window is related to the global root mean square. The difference;

[0023] Calculate the first based on the difference. The binary code corresponding to each element in the signal within each window is used to obtain a binary code sequence, where the first... The first of the signals in the window The binary code corresponding to each element The calculation formula is as follows:

[0024] ;

[0025] or ;

[0026] in, Indicates the first The first of the signals in the window The elements and the global root mean square The difference;

[0027] The binary encoded sequence is converted into decimal feature values. As the first Local texture feature values ​​of the signal within a window.

[0028] Further optimization was performed using the binarization method for information fusion, the specific method of which is as follows:

[0029] The vibration signals were respectively and one-dimensional texture feature signals Perform a discrete Fourier transform to obtain the corresponding complex spectrum. and complex spectrum ;

[0030] The complex spectrum and complex spectrum Align in the frequency domain and compute the complex spectrum. Complex spectrum The amplitude and phase of the complex spectral values ​​at each frequency component;

[0031] Based on complex spectrum and complex spectrum The amplitude and phase of the complex spectral values ​​at each frequency component are used to reconstruct the time-domain discrete sequence for vector-bispectral analysis. and ;

[0032] Calculate the time-domain discrete sequence respectively and Third-order cumulative and Among them, fixed If the value is 0, extract along the edge. One-dimensional third-order cumulant characteristics of direction and ;

[0033] The extracted one-dimensional third-order cumulant and Perform one-dimensional discrete Fourier transforms on each to obtain the vibration signal. arrow double spectrum and one-dimensional texture feature signals arrow double spectrum ;

[0034] From the vibration signal arrow double spectrum and one-dimensional texture feature signals arrow double spectrum The signal with the most significant fault characteristics is selected as the final fused signal.

[0035] Further optimization, complex spectrum and complex spectrum In the The methods for calculating the amplitude and phase of the complex spectrum values ​​at each frequency component are as follows:

[0036] Obtain the complex spectrum respectively In the Complex spectrum values ​​of each frequency component and complex spectrum In the Complex spectrum values ​​of each frequency component Calculate the complex spectrum values ​​respectively. real part and the virtual part and complex spectrum values real part and the virtual part ;

[0037] Based on complex spectrum values Calculate the complex spectrum value using the real and imaginary parts. amplitude With phase According to the complex spectrum value Calculate the complex spectrum value using the real and imaginary parts. amplitude With phase .

[0038] Further preferably, the time-domain discrete sequence and The reconstruction method is as follows:

[0039] Based on the complex spectrum and complex spectrum Calculate the total vector fusion amplitude of the corresponding frequency component based on the amplitude of the complex spectrum value at the corresponding frequency component;

[0040] By combining the phase of the total vector fusion amplitude and the complex spectrum value on the corresponding frequency component, the complex spectrum value corresponding to the corresponding frequency component is reconstructed.

[0041] The reconstructed complex spectrum value and Indexed by frequency component By combining the elements in ascending order, the reconstructed complex spectrum is obtained. and ;

[0042] The reconstructed complex spectrum and Performing an inverse Fourier transform yields the discrete-time sequence. and .

[0043] The bearing fault identification method based on improved one-dimensional local binary pattern and information fusion provided by this invention achieves preliminary noise suppression and enhancement of periodic characteristics by calculating the autocorrelation function of the original vibration signal; by changing the threshold for quantizing the signal in the one-dimensional local binary pattern algorithm from a local intermediate value to the global root mean square value of the signal, an effective measurement of the signal impact characteristics is achieved while considering the overall characteristics of the signal; and by fusing the original signal with the texture feature signal obtained by the improved one-dimensional local binary pattern, fault information is further highlighted while suppressing noise. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 A flowchart of the bearing fault identification method based on improved one-dimensional local binary pattern and information fusion provided by the present invention;

[0046] Figure 2 The following diagram shows the results of the combined fault analysis of the bearing outer ring and inner ring using the method proposed in this invention: where (a) represents the vibration signal. (a) Time domain plot; (b) Vibration signal (c) is the spectrum; autocorrelation function (d) represents the one-dimensional texture feature signal obtained by the proposed method. (e) is the time-domain plot; (e) is the one-dimensional texture feature signal. (f) is the time-domain plot of the fused signal; (g) is the spectrum of the fused signal. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, to avoid obscuring the invention with unnecessary details, only processing steps closely related to the solution of this invention are shown in the drawings, while other details not closely related to this invention are omitted.

[0048] like Figure 1 As shown, the bearing fault identification method based on improved one-dimensional local binary pattern and information fusion provided by the present invention includes the following steps:

[0049] S1: Obtain the vibration signal of the equipment where the bearing is located. ],in, The vibration signal Length;

[0050] S2: Calculate the vibration signal The autocorrelation function is obtained to obtain the autocorrelation function. ],in, The autocorrelation function Length;

[0051] Wherein, the autocorrelation function The calculation formula is as follows:

[0052] ;

[0053] In the formula, The vibration signal Length, For time delay, ;

[0054] Preferred, At this point, the autocorrelation function ];

[0055] S3: Utilize the improved one-dimensional local binary model method to analyze the autocorrelation. Quantization and signal reconstruction are performed to obtain the vibration signal. Corresponding one-dimensional texture feature signal The improved one-dimensional local binary mode method uses the global root mean square of the signal as the quantization threshold.

[0056] Among them, the improved one-dimensional local binary pattern method is used to analyze the autocorrelation. The steps for quantization and signal reconstruction are as follows:

[0057] S31: Set the sliding window parameters, wherein the sliding window parameters include the length of the sliding window and the sliding step size. Preferably, the length of the sliding window... for The sliding step size is 1;

[0058] S32: Calculate the autocorrelation function global root mean square value The calculation formula is as follows:

[0059] ;

[0060] In the formula, The autocorrelation function Length, , The autocorrelation function The Middle One element;

[0061] S33: Using the sliding window, traverse the autocorrelation function point by point with a step size of 1. With the global root mean square The signal within each window is quantized and reconstructed using a threshold value to obtain the local texture feature value of the signal within each window.

[0062] Among them, the ( , The method for quantizing and reconstructing the signal within a window (where the length of the sliding window is 1) is as follows:

[0063] S331: Calculate the... Each element in the signal within the window is related to the global root mean square. The difference, where the first The first of the signals in the window The elements and the global root mean square The difference The calculation formula is as follows:

[0064] ;

[0065] In the formula, For the first The first of the signals in the window One element, ;

[0066] S332: Calculate the first based on the difference. The binary code corresponding to each element in the signal within each window is used to obtain a binary code sequence, where the first... The first of the signals in the window The binary code corresponding to each element The calculation formula is as follows:

[0067] ;

[0068] or ;

[0069] S333: Convert the binary encoded sequence into decimal feature values. As the first The local texture feature value of the signal within a window is calculated using the following formula:

[0070] ;

[0071] S34: Arrange the local texture feature values ​​corresponding to all signals within the window in the order in which the sliding window slides, to obtain the vibration signal. Corresponding one-dimensional texture feature signal ;

[0072] S4: The vibration signal [and the one-dimensional texture feature signal] Information fusion is performed to obtain a fused signal with more significant fault characteristics. Information fusion can employ methods such as cross-correlation function of signals, loss of bispectrum, DS evidence theory, and principal component analysis.

[0073] Preferably, the bispectral loss method is used for information fusion, and the specific method is as follows:

[0074] S41: The vibration signals are respectively... and one-dimensional texture feature signals Perform a discrete Fourier transform to obtain the corresponding complex spectrum. and complex spectrum ;

[0075] S42: The complex spectrum and complex spectrum Align in the frequency domain and compute the complex spectrum. Complex spectrum The amplitude and phase of the complex spectrum values ​​at each frequency component, where the complex spectrum... and complex spectrum In the The methods for calculating the amplitude and phase of the complex spectrum values ​​at each frequency component are as follows:

[0076] S421: Obtain the complex spectrum respectively In the Complex spectrum values ​​of each frequency component and complex spectrum In the Complex spectrum values ​​of each frequency component Calculate the complex spectrum values ​​respectively. real part and the virtual part and complex spectrum values real part and the virtual part ;

[0077] S422: Based on the complex spectrum value Calculate the complex spectrum value using the real and imaginary parts. amplitude With phase According to the complex spectrum value Calculate the complex spectrum value using the real and imaginary parts. amplitude With phase Among them, complex spectrum value amplitude With phase The calculation formula is as follows:

[0078] ;

[0079] Among them, complex spectrum value amplitude With phase The calculation formula is as follows:

[0080] ;

[0081] S43: Based on complex spectrum and complex spectrum The amplitude and phase of the complex spectral values ​​at each frequency component are used to reconstruct the time-domain discrete sequence for vector-bispectral analysis. and ;

[0082] Among them, time-domain discrete sequence and The reconstruction method is as follows:

[0083] S431: Based on the complex spectrum and complex spectrum Calculate the total vector fusion amplitude of the corresponding frequency component based on the amplitude of the complex spectral value at the corresponding frequency component, where the first... The total vector fusion amplitude corresponding to each frequency component The calculation formula is as follows:

[0084] ;

[0085] S432: By combining the total vector fusion amplitude and the phase of the complex spectrum value on the corresponding frequency component, the complex spectrum value corresponding to the corresponding frequency component is reconstructed, wherein the reconstructed first... Complex spectrum value corresponding to each frequency component and The calculation formula is as follows:

[0086] ;

[0087] In the formula, The imaginary unit;

[0088] S433: Reconstruct the complex spectrum value and Indexed by frequency component By combining the elements in ascending order, the reconstructed complex spectrum is obtained. ;

[0089] S434: For the reconstructed complex spectrum... and Performing an inverse Fourier transform yields the discrete-time sequence. and ;

[0090] S44: Calculate the time-domain discrete sequences respectively. and Third-order cumulative and ,fixed A constant value (default is 0), extracting along the edge One-dimensional third-order cumulant characteristics of direction and The calculation formula is as follows:

[0091] ;

[0092] in, , As variables, For quantitative purposes, , discrete sequence The Middle One element, , discrete sequence The Middle One element, ;

[0093] S45: Extracted one-dimensional third-order cumulative quantity and Perform one-dimensional discrete Fourier transforms on each to obtain the vibration signal. arrow double spectrum and one-dimensional texture feature signals arrow double spectrum ,in, and The calculation formula is as follows:

[0094] ;

[0095] In the formula, Angular frequency, ;

[0096] S46: From the vibration signal arrow double spectrum and one-dimensional texture feature signals arrow double spectrum The signal with the most significant fault characteristics is selected as the final fused signal.

[0097] S5: Perform spectral analysis on the fused signal and calculate the bearing fault characteristic frequency based on the bearing geometric parameters and rotational speed. Determine the bearing condition and fault type by using the obtained frequency components and bearing fault characteristic frequencies.

[0098] The effectiveness of the method proposed in this invention will be verified through specific embodiments below. In the embodiments, the data sampling frequency is 10000 Hz.

[0099] For the sake of simplicity, let's use Represents the rotation frequency, in Inner ring fault characteristic frequency, in Outer ring fault characteristic frequency, in ,by Cage fault characteristic frequencies. Using random bearing composite fault data as an example, the method proposed in this invention is verified. The characteristic frequencies corresponding to each typical data point are calculated according to the bearing fault characteristic frequency calculation formula.

[0100] Example

[0101] To demonstrate the effectiveness of the method proposed in this invention, data from a combined outer and inner ring failure of the bearing is analyzed as an example. The corresponding rotational speed in this state is... =1530r / min, the characteristic frequencies of the bearing are: =25.51Hz, =113.1Hz, =65.47Hz, =44.43Hz, =9.35Hz.

[0102] Figure 2 The following diagram shows the results of the combined fault analysis of the bearing outer ring and inner ring using the method proposed in this invention: where (a) represents the vibration signal. (a) Time domain plot; (b) Vibration signal (c) is the spectrum; autocorrelation function (d) represents the one-dimensional texture feature signal obtained by the proposed method. (e) is the time-domain plot; (e) is the one-dimensional texture feature signal. (f) is the time-domain plot of the fused signal; (g) is the spectrum of the fused signal.

[0103] (1) Analysis Figure 2 (b) The fault features in the original signal spectrum were overwhelmed by noise, and no typical frequency components corresponding to bearing faults could be found. In other words, bearing faults could not be determined based on the spectrum of the original signal.

[0104] (2) Analysis Figure 2 (e) After processing by the proposed method of enhancing one-dimensional local binary mode, the impulse component in the spectrum of the one-dimensional texture feature signal is significantly increased, and the noise in the signal is also suppressed.

[0105] (3) Analysis Figure 2 (g) Furthermore, by analyzing the prominent frequency components in the spectrum, it can be found that the signal spectrum obtained by the method proposed in this invention has the following characteristics:

[0106] 1) Noise components are significantly reduced, achieving effective noise suppression;

[0107] 2) Typical and prominent characteristic frequencies corresponding to bearing inner ring failures exist, including 289.3Hz ((289.3+2fr) / 3=113.48Hz) and 730.1Hz ((703.1-fr) / 6=112.9Hz);

[0108] 3) There are typical and prominent frequency components corresponding to the bearing outer ring fault, including 1233Hz (1233 / 19=64.89Hz).

[0109] It can be observed that the method proposed in this invention has significant noise suppression capabilities. It can accurately extract prominent and abundant fault characteristic frequencies of the outer and inner rings of the bearing corresponding to bearing faults in complex noise environments, thereby achieving accurate judgment of bearing composite fault types.

[0110] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of this application can be implemented in various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0111] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0112] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0113] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0114] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0115] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A bearing fault identification method based on improved one-dimensional local binary pattern and information fusion, characterized in that, include: Obtain vibration signals from the equipment where the bearing is located. ],in, The vibration signal Length; Calculate the vibration signal The autocorrelation function is obtained. ],in, The autocorrelation function Length; The autocorrelation is analyzed using an improved one-dimensional local binary model method. Quantization and signal reconstruction are performed to obtain the vibration signal. Corresponding one-dimensional texture feature signal The improved one-dimensional local binary mode method uses the global root mean square of the signal as the quantization threshold. The vibration signal With the one-dimensional texture feature signal Information fusion is performed to obtain a fused signal with more significant fault characteristics; The fused signal is subjected to spectral analysis, and the bearing fault characteristic frequency is calculated based on the bearing geometric parameters and rotational speed. The bearing condition and fault type are determined by the obtained frequency components and the bearing fault characteristic frequency.

2. The bearing fault identification method based on improved one-dimensional local binary pattern and information fusion as described in claim 1, characterized in that, The autocorrelation function The calculation formula is as follows: ； In the formula, The vibration signal Length, For time delay, .

3. The bearing fault identification method based on improved one-dimensional local binary pattern and information fusion according to claim 2, characterized in that, 。 4. The bearing fault identification method based on improved one-dimensional local binary pattern and information fusion according to claim 1, characterized in that, The autocorrelation is analyzed using an improved one-dimensional local binary model method. The steps for quantization and signal reconstruction are as follows: Set the sliding window parameters, which include the length of the sliding window and the sliding step size; Calculate the autocorrelation function global root mean square value ; Using the sliding window, the autocorrelation function is traversed point by point with a step size of 1. With the global root mean square The signal within each window is quantized and reconstructed using a threshold value to obtain the local texture feature value of the signal within each window. Arrange the local texture feature values ​​corresponding to all signals within the window in the order in which the sliding window slides, to obtain the vibration signal. Corresponding one-dimensional texture feature signal .

5. The bearing fault identification method based on improved one-dimensional local binary pattern and information fusion according to claim 4, characterized in that, No. The methods for quantization and signal reconstruction of the signal within a window are as follows, where, , The length of the sliding window: Calculate the first Each element in the signal within the window is related to the global root mean square. The difference; Calculate the first based on the difference. The binary code corresponding to each element in the signal within each window is used to obtain a binary code sequence, where the first... The first of the signals in the window The binary code corresponding to each element The calculation formula is as follows: ； or ; in, Indicates the first The first of the signals in the window The elements and the global root mean square The difference; The binary encoded sequence is converted into decimal feature values. As the first Local texture feature values ​​of the signal within a window.

6. The bearing fault identification method based on improved one-dimensional local binary pattern and information fusion according to claim 1, characterized in that, Information fusion is performed using the binarization method, and the specific method is as follows: The vibration signals were respectively and one-dimensional texture feature signals Perform a discrete Fourier transform to obtain the corresponding complex spectrum. and complex spectrum ; The complex spectrum and complex spectrum Align in the frequency domain and compute the complex spectrum. Complex spectrum The amplitude and phase of the complex spectral values ​​at each frequency component; Based on complex spectrum and complex spectrum The amplitude and phase of the complex spectral values ​​at each frequency component are used to reconstruct the time-domain discrete sequence for vector-bispectral analysis. and ; Calculate the time-domain discrete sequence respectively and Third-order cumulative and Among them, fixed If the value is 0, extract along the path. One-dimensional third-order cumulant characteristics of direction and ; The extracted one-dimensional third-order cumulant and Perform one-dimensional discrete Fourier transforms on each to obtain the vibration signal. arrow double spectrum and one-dimensional texture feature signals arrow double spectrum ; From the vibration signal arrow double spectrum and one-dimensional texture feature signals arrow double spectrum The signal with the most significant fault characteristics is selected as the final fused signal.

7. The bearing fault identification method based on improved one-dimensional local binary pattern and information fusion according to claim 6, characterized in that, Complex spectrum and complex spectrum In the The methods for calculating the amplitude and phase of the complex spectrum values ​​at each frequency component are as follows: Obtain the complex spectrum respectively In the Complex spectrum values ​​of each frequency component and complex spectrum In the Complex spectrum values ​​of each frequency component Calculate the complex spectrum values ​​respectively. real part and the virtual part and complex spectrum values real part and the virtual part ; Based on complex spectrum values Calculate the complex spectrum value using the real and imaginary parts. amplitude With phase According to the complex spectrum value Calculate the complex spectrum value using the real and imaginary parts. amplitude With phase .

8. The bearing fault identification method based on improved one-dimensional local binary pattern and information fusion according to claim 6, characterized in that, The time-domain discrete sequence and The reconstruction method is as follows: Based on the complex spectrum and complex spectrum Calculate the total vector fusion amplitude of the corresponding frequency component based on the amplitude of the complex spectrum value at the corresponding frequency component; By combining the phase of the total vector fusion amplitude and the complex spectrum value on the corresponding frequency component, the complex spectrum value corresponding to the corresponding frequency component is reconstructed. The reconstructed complex spectrum value and Indexed by frequency component By combining the elements in ascending order, the reconstructed complex spectrum is obtained. and ; The reconstructed complex spectrum and Performing an inverse Fourier transform yields the discrete-time sequence. and .

Images