BEMD-based correlation screening method and system for purifying hydroelectric generator shaft orbit

Abstract

The application discloses a water-turbine-generator shaft center trajectory purification method and system based on correlation screening BEMD, and the method comprises the following steps: acquiring vibration signals in mutually perpendicular directions of the same measuring point of a water-turbine-generator, constructing the signals into complex signals for adaptive decomposition, obtaining IMF components and a final residual component; calculating a correlation coefficient, and screening m effective IMF components; defining a weight coefficient based on the correlation coefficient of each effective IMF component, adaptively weighting and reconstructing the m effective IMF components and the residual component reserved after screening, and obtaining a purified complex signal matrix; extracting a real part of the complex signal matrix as a purified X-direction sequence, extracting an imaginary part as a purified Y-direction sequence, and drawing a shaft center trajectory diagram. The application has the advantages of strong false component identification and inhibition capacity.

Description

Technical Field

[0001] This invention mainly relates to the field of hydropower technology, specifically to a method and system for purifying the shaft trajectory of a hydropower unit based on relevant screening BEMD. Background Technology

[0002] Hydropower, as the largest-scale clean energy source, occupies a crucial position in the energy structure. Furthermore, with the continuous increase in the capacity of individual hydropower units and the extension of annual power generation time, hydropower units have become the core of peak shaving and frequency regulation in the power system, and their safe and stable operation is directly related to the overall reliability of the power grid. A hydropower unit is a typical complex nonlinear system with "water-machine-electricity" coupling; a shutdown due to an accident will cause huge economic losses and social impact. According to relevant statistics, approximately 80% of faults can be reflected in vibration signals, making vibration signal analysis key to fault diagnosis. Due to the variable operating conditions of hydropower units, their vibration is affected by the interaction of hydraulic, mechanical, and electrical factors, resulting in complex fault mechanisms. This leads to strong non-stationarity and nonlinearity in their vibration signals, posing significant challenges to fault feature extraction.

[0003] The shaft center trajectory of a hydropower unit is a crucial characteristic reflecting the unit's operational stability and fault status. In actual engineering, due to various factors such as electromagnetic interference, eddy current sensor misalignment, surface runout, and environmental noise, the original shaft center trajectory collected often contains a large amount of noise and redundant components, resulting in blurred and overlapping trajectories after purification, which seriously affects the accuracy of fault diagnosis.

[0004] Traditional methods for purifying axis-centered trajectories include low-pass filtering, wavelet thresholding, and principal component analysis (PCA). However, low-pass filtering loses high-frequency details of the trajectory; wavelet denoising faces difficulties in basis function selection; and PCA has limited capabilities in processing nonlinear and non-stationary signals. Two-dimensional empirical mode decomposition (BEMD), as a data-driven adaptive decomposition method, has advantages in nonlinear signal processing, but when directly applied to axis-centered trajectories, it easily generates spurious components without physical meaning, and its decomposition efficiency is significantly affected by the interpolation algorithm. Summary of the Invention

[0005] To address the technical problems existing in the prior art, this invention provides a method and system for purifying the shaft trajectory of hydroelectric generators based on correlation screening BEMD, which has strong ability to identify and suppress false components.

[0006] To solve the above-mentioned technical problems, the technical solution proposed by this invention is as follows: A method for purifying the shaft center trajectory of a hydroelectric generator based on correlation screening BEMD includes the following steps: S1. Obtain mutually perpendicular measurements at the same measuring point of the hydropower unit. direction and Vibration signals in the direction and Simultaneously record the reference signal under normal operating conditions and the vibration signal under fault conditions to be analyzed; S2. Collect data under fault conditions. Vibration signal in direction and Vibration signals in the direction Constructed as a complex signal An adaptive decomposition using a two-dimensional empirical mode decomposition algorithm yields a sequence arranged from high frequency to low frequency. IMF components , …, and the final residual component ; S3. Calculate the original value for each IMF component. Direction signal or correlation coefficient of directional signals And set the correlation coefficient threshold. ;when ≥ When the condition is met, the IMF component is determined to be a valid physical component and is retained; when... < If the result is not found, it is initially determined that the component is a false component or a noise-dominated component and is removed, thus obtaining m valid IMF components. S4. Define the weighting coefficients based on the correlation coefficients of each effective IMF component. The m valid IMF components and residual components were selected and retained. Adaptive weighted reconstruction is performed to obtain the purified complex signal matrix. ; S5. Extract the complex signal matrix The real part is used as the purified X-direction sequence. The imaginary part is extracted as the purified Y-direction sequence. ,by The x-axis is... Use the vertical axis to plot the axis trajectory.

[0007] Preferably, the BEMD adaptive decomposition in step S2 specifically includes the following sub-steps: Set the number of projection directions to Then the first Projection direction angle for:

[0008] in ; Project the complex signal z(t) onto The real-valued projection signal is obtained from the above:

[0009] in This indicates the operation of taking the real part; extract The set of times corresponding to local maxima { }, for set { , Interpolation is performed to obtain the direction. Maximum envelope on ; Calculate the centroid of the maximum envelope in all directions. The calculation expression is as follows:

[0010] Subtract the centroid of the envelope from the original complex signal to obtain the remaining components. :

[0011] judge If the IMF's criteria are met, the process iterates repeatedly until... It continues until the Intrinsic Mode Function (IMF).

[0012] Preferably, in step S3, each IMF component is calculated and compared with the original... Direction signal or correlation coefficient of directional signals The specific process is as follows: Calculate the i-th IMF component Compared with the original X-direction signal Pearson correlation coefficient ; Simultaneously calculate the i-th IMF component Compared with the original Y-direction signal The Pearson correlation coefficient is taken as the larger of the two absolute values. .

[0013] Preferably, the Pearson correlation coefficient The calculation formula is:

[0014] in, express The real part, For covariance, Let Variance be the variance.

[0015] Preferably, in step S3, the correlation coefficient between adjacent IMF components is further calculated. for:

[0016] in, The correlation coefficient is the real part of adjacent IMFs. The correlation coefficient represents the imaginary part of adjacent IMFs.

[0017] Preferably, and The expression is as follows: .

[0018] Preferably, in step S4, the adaptive weighted reconstruction process is as follows: The weighting coefficient of the j-th effective IMF component is defined as follows:

[0019] in The correlation coefficient between the j-th effective IMF component and the original signal; the purified two-dimensional data matrix. for:

[0020] in Let R be the j-th effective IMF component, and R be the final residual component.

[0021] The present invention also discloses a computer program product, comprising a computer program that, when executed by a processor, performs the steps of the method described above.

[0022] The present invention further discloses a computer-readable storage medium having a computer program stored thereon, the computer program executing the steps of the method described above when run by a processor.

[0023] The present invention also discloses a hydroelectric generator shaft trajectory purification system based on correlation screening BEMD, comprising a memory and a processor connected to each other, wherein the memory stores a computer program, and the computer program executes the steps of the method described above when run by the processor.

[0024] Compared with the prior art, the advantages of the present invention are as follows: 1. Strong ability to suppress spurious components: This invention introduces correlation analysis into the post-processing stage of two-dimensional empirical mode decomposition. By quantitatively calculating the correlation coefficient between each IMF component and the original vibration signal, an effective physical component discrimination criterion is established. This overcomes the defect that traditional BEMD decomposition results are difficult to objectively evaluate signal fidelity. It can accurately eliminate spurious components and noise-dominant components caused by interpolation errors and mode mixing.

[0025] 2. High fidelity of nonlinear fault features: This invention employs an adaptive weighted reconstruction strategy based on correlation coefficients, replacing the traditional "hard threshold" method of direct discarding or equal weighting. This strategy ensures that effective components in different frequency bands participate in the reconstruction according to their contribution, preserving high-frequency weak fault features while avoiding excessive amplification of noise components, thus ensuring the complete preservation of key nonlinear fault information in the shaft center trajectory.

[0026] 3. Decomposition efficiency optimization: The relevant screening and weighted reconstruction process described in this invention only requires a small number of vector inner product operations in the post-processing stage, without any modification to the BEMD iteration kernel. Therefore, without increasing the computational burden of the main decomposition process, the purification efficiency is significantly improved, which is convenient for real-time or near-real-time applications in engineering sites.

[0027] 4. Robustness and Adaptability: The correlation coefficient screening threshold in this invention can be adaptively adjusted based on prior operating conditions such as the rated load of the hydropower unit and the clearance of the guide bearing. It can also be dynamically set by using the baseline value of the correlation level of the noise component statistically based on the historical fault-free data of the unit. This achieves adaptive matching for different units and operating conditions, reduces human experience intervention, and ensures the stability and versatility of the purification method in different scenarios. Attached Figure Description

[0028] Figure 1 This is a flowchart of the method for purifying the shaft trajectory of a hydroelectric generator based on correlation screening BEMD, according to an embodiment of the present invention.

[0029] Figure 2 This is a flowchart of the adaptive decomposition process according to an embodiment of the present invention.

[0030] Figure 3 This is a flowchart of the adaptive reconstruction signal according to an embodiment of the present invention. Detailed Implementation

[0031] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0032] like Figure 1 As shown in the figure, the specific implementation steps of the hydroelectric generator shaft trajectory purification method based on correlation screening BEMD provided in this embodiment of the invention are as follows: S1. Orthogonal sampling of vibration signals. Using orthogonal sampling technology, two mutually perpendicular directions are sampled from the same measuring point of the hydropower unit. direction and Vibration signal in direction and It is necessary to collect data from the unit under normal conditions and signal data under fault conditions to be analyzed simultaneously.

[0033] S2. BEMD Decomposition. The vibration signal collected under fault conditions is regarded as a complex signal z(t) and used as input. The two-dimensional empirical mode decomposition (BEMD) algorithm is used for adaptive decomposition to obtain a series of IMF components from high frequency to low frequency.

[0034] like Figure 2 As shown, the adaptive decomposition process is as follows: First, the vibration signals collected under fault conditions in two directions need to be treated as a complex signal. Its expression is as follows:

[0035] in, The imaginary unit is used. This complex signal is used as input to a two-dimensional empirical mode decomposition (BEMD), and a series of intrinsic mode function (IMF) components arranged from high frequency to low frequency are obtained through adaptive decomposition. The specific decomposition process is as follows: Set the number of projection directions to Then the first Projection direction angle for:

[0036] in .

[0037] Project the complex signal z(t) onto The real-valued projection signal is obtained from the above:

[0038] in This indicates the operation of taking the real part.

[0039] extract The set of times corresponding to local maxima { } and record the complex values ​​at these times. (in Then for the set { , Interpolation is performed to obtain the result in the direction. Maximum envelope on .

[0040] Calculate the centroid of the envelope corresponding to the maximum value of the vibration signal in all directions. The calculation expression is as follows:

[0041] Subtract the centroid of the envelope from the original complex signal to obtain the remaining components. :

[0042] judge If the IMF's criteria are met, the process iterates repeatedly until... It continues until the Intrinsic Mode Function (IMF).

[0043] The recorded IMF is removed from the original signal to obtain IMF1: The remaining components are .

[0044] Repeat the above process for the remaining components to obtain a series of IMF components from high frequency to low frequency. , …, .in This represents the total number of IMFs in the decomposition.

[0045] S3. For example Figure 3 As shown, calculate each modal component. Compared with the original X-direction signal correlation coefficient Set threshold ,when ≥ If the condition is met, the IMF is determined to be a valid physical component and is retained; if the condition is not met, the IMF is considered to be a valid physical component and is retained. < If so, it is initially determined to be a spurious component or a noise-dominant component and is removed.

[0046] The correlation coefficient calculation process is as follows: The correlation coefficient, a statistical indicator used to reflect the strength of the correlation between variables, can characterize the degree of linear relationship between X and Y. A larger correlation coefficient indicates a better correlation between X and Y; a smaller correlation coefficient indicates a weaker correlation. Specifically, the i-th IMF component is calculated. Compared with the original X-direction signal Pearson correlation coefficient:

[0047] in, express The real part, For covariance, For variance; Simultaneously calculate the i-th IMF component Compared with the original Y-direction signal The Pearson correlation coefficient is taken as the larger of the two absolute values. .

[0048] For vibration signals of rotating machinery, since signals in different directions may differ, a correlation coefficient can be defined between the i-th IMF and the (i+1)-th IMF. for:

[0049] in, and The expression is as follows:

[0050] in: , The correlation coefficient is the real part of adjacent IMFs. The correlation coefficient represents the imaginary part of adjacent IMFs.

[0051] in This indicates the operation of taking the imaginary part; when If the value is below the preset auxiliary threshold, it is further confirmed as a spurious component and deleted.

[0052] S4. Define the weighting coefficients based on the correlation coefficients of each effective IMF component. The m valid IMF components and residual components were selected and retained. Adaptive weighted reconstruction is performed to obtain the purified complex signal matrix. ; The adaptive weighted reconstruction process is as follows: Keep the filter One effective IMF component and the final residual component Reconstruction is performed, introducing weighting coefficients to enhance the contribution of strongly correlated components and suppress residual noise. The weighting coefficient of the j-th effective IMF component is defined as:

[0053] Purified two-dimensional data matrix for:

[0054] S5. Visualize and extract the axis trajectory. The first column serves as the purified X-direction sequence. The second column represents the purified Y-direction sequence. .by The x-axis is... Using the vertical axis as the ordinate, a high signal-to-noise ratio axis-centric trajectory is plotted in a two-dimensional coordinate system. Furthermore, the geometric features of this trajectory (such as trajectory shape, eccentricity, area, and precession direction) can be extracted for subsequent fault diagnosis. Comparing the purified trajectory with the original trajectory provides a direct verification of the effectiveness of this method in eliminating noise, suppressing spurious components, and preserving nonlinear fault characteristics.

[0055] Compared with the prior art, the present invention has the following significant advantages: 1. Strong ability to suppress spurious components: This invention introduces correlation analysis into the post-processing stage of two-dimensional empirical mode decomposition. By quantitatively calculating the correlation coefficient between each IMF component and the original vibration signal, an effective physical component discrimination criterion is established. This overcomes the defect that traditional BEMD decomposition results are difficult to objectively evaluate signal fidelity. It can accurately eliminate spurious components and noise-dominant components caused by interpolation errors and mode mixing.

[0056] 2. High fidelity of nonlinear fault features: This invention employs an adaptive weighted reconstruction strategy based on correlation coefficients, replacing the traditional "hard threshold" method of direct discarding or equal weighting. This strategy ensures that effective components in different frequency bands participate in the reconstruction according to their contribution, preserving high-frequency weak fault features while avoiding excessive amplification of noise components, thus ensuring the complete preservation of key nonlinear fault information in the shaft center trajectory.

[0057] 3. Decomposition efficiency optimization: The relevant screening and weighted reconstruction process described in this invention only requires a small number of vector inner product operations in the post-processing stage, without any modification to the BEMD iteration kernel. Therefore, without increasing the computational burden of the main decomposition process, the purification efficiency is significantly improved, which is convenient for real-time or near-real-time applications in engineering sites.

[0058] 4. Robustness and Adaptability: The correlation coefficient screening threshold in this invention can be adaptively adjusted based on prior operating conditions such as the rated load of the hydropower unit and the clearance of the guide bearing. It can also be dynamically set by using the baseline value of the correlation level of the noise component statistically based on the historical fault-free data of the unit. This achieves adaptive matching for different units and operating conditions, reduces human experience intervention, and ensures the stability and versatility of the purification method in different scenarios.

[0059] The present invention also discloses a computer program product, comprising a computer program that, when executed by a processor, performs the steps of the method described above.

[0060] The present invention further discloses a computer-readable storage medium having a computer program stored thereon, the computer program executing the steps of the method described above when run by a processor.

[0061] The present invention also discloses a hydroelectric generator shaft trajectory purification system based on correlation screening BEMD, comprising a memory and a processor connected to each other, wherein the memory stores a computer program, and the computer program executes the steps of the method described above when run by the processor.

[0062] The products, media, and systems of the present invention, corresponding to the methods described above, also possess the advantages described above.

[0063] The present invention can implement all or part of the processes in the methods of the above embodiments, or it can be implemented by hardware related to computer program instructions. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, it can implement the steps of the above method embodiments. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form. The computer-readable storage medium includes: any entity or device capable of carrying computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. The memory is used to store computer programs and / or modules. The processor implements various functions by running or executing the computer programs and / or modules stored in the memory, and by calling data stored in the memory. The memory may include high-speed random access memory, as well as non-volatile memory, such as hard disks, RAM, plug-in hard disks, smart media cards (SMC), secure digital (SD) cards, flash cards, at least one disk storage device, flash memory device, or other volatile solid-state storage devices.

[0064] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A method for purifying the shaft center trajectory of a hydroelectric generator based on correlation screening BEMD, characterized in that, Including the following steps: S1. Obtain mutually perpendicular measurements at the same measuring point of the hydropower unit. direction and Vibration signals in the direction and Simultaneously record the reference signal under normal operating conditions and the vibration signal under fault conditions to be analyzed; S2. Collect data under fault conditions. Vibration signal in direction and Vibration signals in the direction Constructed as a complex signal An adaptive decomposition using a two-dimensional empirical mode decomposition algorithm yields a sequence arranged from high frequency to low frequency. IMF components , …, and the final residual component ; S3. Calculate the original value for each IMF component. Direction signal or correlation coefficient of directional signals And set the correlation coefficient threshold. ;when ≥ When the condition is met, the IMF component is determined to be a valid physical component and is retained; when... < If the result is not found, the component is initially identified as a false component or a noise-dominated component and is removed, resulting in m valid IMF components. S4. Define the weighting coefficients based on the correlation coefficients of each effective IMF component. The m valid IMF components and residual components were selected and retained. Adaptive weighted reconstruction is performed to obtain the purified complex signal matrix. ; S5. Extract the complex signal matrix The real part is used as the purified X-direction sequence. The imaginary part is extracted as the purified Y-direction sequence. ,by The x-axis is... Use the vertical axis to plot the axis trajectory.

2. The method for purifying the shaft center trajectory of a hydroelectric generator based on correlation screening BEMD according to claim 1, characterized in that, Step S2, the BEMD adaptive decomposition, specifically includes the following sub-steps: Set the number of projection directions to Then the first Projection direction angle for: in ; Project the complex signal z(t) onto The real-valued projection signal is obtained from the above: in This indicates the operation of taking the real part; extract The set of times corresponding to local maxima { }, for set { , Interpolation is performed to obtain the direction. Maximum envelope on ; Calculate the centroid of the maximum envelope in all directions. The calculation expression is as follows: Subtract the centroid of the envelope from the original complex signal to obtain the remaining components. : judge If the IMF's criteria are met, the process iterates repeatedly until... It continues until the Intrinsic Mode Function (IMF).

3. The method for purifying the shaft center trajectory of a hydroelectric generator based on correlation screening BEMD according to claim 1, characterized in that, In step S3, each IMF component is compared with the original... Direction signal or correlation coefficient of directional signals The specific process is as follows: Calculate the i-th IMF component Compared with the original X-direction signal Pearson correlation coefficient ; Simultaneously calculate the i-th IMF component Compared with the original Y-direction signal The Pearson correlation coefficient is taken as the larger of the two absolute values. .

4. The method for purifying the shaft center trajectory of a hydroelectric generator based on correlation screening BEMD according to claim 3, characterized in that, Pearson correlation coefficient The calculation formula is: in, express The real part, For covariance, Let Variance be the variance.

5. The method for purifying the shaft center trajectory of a hydroelectric generator based on correlation screening BEMD according to claim 4, characterized in that, In step S3, the correlation coefficient between adjacent IMF components is further calculated. for: in, The correlation coefficient is the real part of adjacent IMFs. The correlation coefficient represents the imaginary part of adjacent IMFs.

6. The method for purifying the shaft center trajectory of a hydroelectric generator based on correlation screening BEMD according to claim 5, characterized in that, and The expression is as follows: 。 7. The method for purifying the shaft center trajectory of a hydroelectric generator based on correlation screening BEMD according to any one of claims 1-6, characterized in that, In step S4, the adaptive weighted reconstruction process is as follows: The weighting coefficient of the j-th effective IMF component is defined as follows: in The correlation coefficient between the j-th effective IMF component and the original signal; the purified two-dimensional data matrix. for: in Let R be the j-th effective IMF component, and R be the final residual component.

8. A computer program product, comprising a computer program, characterized in that, The computer program is executed by the processor to perform the steps of the method as described in any one of claims 1-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program, when run by a processor, performs the steps of the method as described in any one of claims 1-7.

10. A system for purifying the shaft center trajectory of a hydroelectric generator based on correlation screening BEMD, comprising an interconnected memory and a processor, wherein the memory stores a computer program, characterized in that, The computer program, when run by a processor, performs the steps of the method as described in any one of claims 1-7.

Images