Error compensation-based leaf tip timing signal processing method and system

Abstract

The embodiment of the invention provides a blade tip timing signal processing method and system based on error compensation, belongs to the technical field of blade monitoring, and solves the problem of insufficient measurement deviation caused by the fact that a band-pass filter is used for filtering a signal in an existing method, and the method comprises the steps: obtaining a blade timing signal in real time based on a blade timing sensor; correcting a blade end vibration displacement model based on an error compensation mechanism in combination with a blade end timing signal, and filtering and reconstructing a blade end correction signal based on discrete Fourier transform in combination with a least square method; eliminating the measurement deviation by using the sparse feature to obtain a deviation elimination signal after deviation elimination; according to the method, when the leaf tip correction signal is processed, the discrete Fourier transform is combined with the least square method to perform filtering reconstruction processing, and the multi-measurement snapshot matrix is constructed for the leaf tip correction signal after filtering reconstruction processing, so that useful information in the signal can be more effectively mined, measurement deviation is eliminated, and the measurement accuracy of the leaf tip correction signal is improved. Therefore, a deviation elimination signal with higher quality is obtained.

Description

Technical Field

[0001] The embodiments of the present invention relate to the technical field of blade monitoring, and in particular to a blade tip timing signal processing method and system based on error compensation. Background Art

[0002] Monitoring the health of rotor blades is crucial in many rotating machinery applications, including aircraft engines and gas turbines. Tip timing technology, an effective means of monitoring blade vibration, measures the tip signal's duration as the blade passes a sensor to obtain information about the blade's vibration characteristics. However, in practice, the tip timing signal measurement process is susceptible to interference from various factors, leading to errors in the measurement results and thus compromising the accurate assessment of the blade's vibration status.

[0003] Chinese patent CN113565584B discloses a time-frequency filtering method for blade tip timing signals, the method comprising the following steps: using a blade tip timing sensor to obtain a time pulse of a rotating blade, and then converting the actual arrival time and the theoretical arrival time difference Δt into displacement data according to the blade radius R and the rotation speed n; selecting and analyzing data according to the sampling frequency, using a spectrum analysis method to obtain an aliasing frequency amplitude vector, combining the aliasing frequency amplitude vectors at different sampling frequencies to form a two-dimensional amplitude matrix, and obtaining a rotation frequency-aliasing frequency graph from the rotation frequency sequence, the aliasing frequency sequence, and the two-dimensional amplitude matrix; identifying a straight line in the RAF and obtaining an expression for the straight line, i.e., slope and intercept parameters; calculating the aliasing frequency at different sampling frequencies according to the straight line expression, generating a bandpass filter frequency band range according to the aliasing frequency, and filtering the signal using the bandpass filter.

[0004] However, the existing method uses a bandpass filter to filter the signal, but it mainly performs simple filtering on the bandpass filter frequency band range generated by the aliasing frequency, which is insufficient to eliminate the measurement deviation, resulting in a deviation between the final signal and the actual blade tip torsional vibration. Summary of the Invention

[0005] The embodiments of the present invention aim to solve at least one of the technical problems existing in the prior art, and provide a blade tip timing signal processing method and system based on error compensation.

[0006] In a first aspect, an embodiment of the present invention provides a blade tip timing signal processing method based on error compensation, the method comprising:

[0007] The blade timing sensor acquires the blade timing signal in real time, and the simulation software draws the blade Campbell diagram based on the blade timing signal to determine the blade tip vibration displacement model in the blade Campbell diagram;

[0008] Extract the blade tip timing signal corresponding to the blade tip vibration displacement model, and correct the blade tip vibration displacement model based on the error compensation mechanism and the blade tip timing signal to obtain the blade tip torsional vibration correction model.

[0009] Load the blade tip torsional vibration correction model, extract the blade tip correction signal from the blade tip torsional vibration correction model, and filter and reconstruct the blade tip correction signal based on discrete Fourier transform combined with least squares method;

[0010] A multi-measurement snapshot matrix is ​​constructed for the blade tip correction signal after filtering and reconstruction, and the sparse features are used to eliminate the measurement deviation to obtain the deviation-eliminated signal.

[0011] In some possible embodiments, the method further includes:

[0012] Obtain the deviation cancellation signal, establish a multi-rotor transfer function matrix based on the deviation cancellation signal, extract the blade tip vibration characteristics through the iterative reweighting method combined with the multi-rotor transfer function matrix, and calculate the DTW distance between the blade tip vibration characteristics and the fault database based on the dynamic time warping algorithm to determine whether the DTW distance exceeds the distance threshold;

[0013] If the DTW distance does not exceed the distance threshold, the blade tip fault mode is matched based on the DTW distance between the blade tip vibration characteristics and the fault database.

[0014] In some possible embodiments, the method for determining a blade tip vibration displacement model in a Campbell diagram of a blade includes:

[0015] Identify blade timing signals in the Campbell diagram, perform synchronous analysis of blade timing signals in the time and frequency domains, and extract the speed-frequency-amplitude three-dimensional matrix of the signal based on wavelet transform;

[0016] A speed-frequency-amplitude three-dimensional matrix of the loading signal is used to extract at least one set of resonance points in the Campbell diagram of the blade based on the speed-frequency-amplitude three-dimensional matrix;

[0017] The Campbell diagram curve is fitted using the least squares complex fingering method to obtain the modal damping ratio and vibration mode coefficient corresponding to the resonance point;

[0018] The blade tip vibration differential equation is constructed based on the Euler-Bernoulli algorithm. The modal damping ratio, mode coefficient, and resonance frequency corresponding to the resonance point are loaded. The displacement response is solved based on the finite element method, and the blade tip vibration displacement model including the displacement response results is obtained.

[0019] In some possible embodiments, the mode coefficient of the resonance point is calculated using the following formula:

[0020]

[0021] Among them, φ, are the vibration mode coefficient and modal damping ratio of the resonance point, S(n,f,A) represents the speed-frequency-amplitude three-dimensional matrix of the signal, G represents the number of resonance points, A,f,f g Represents amplitude, signal frequency, and resonant frequency respectively, K g ,λ,R g are mechanical power gain factor, blade tip ratio, blade length, R g is the blade adjustment coefficient.

[0022] In some possible embodiments, the method for correcting the blade tip vibration displacement model based on the error compensation mechanism and the blade tip timing signal specifically includes:

[0023] Pre-building an error compensation function based on sensor error compensation, measurement system error compensation, model error compensation, and environmental error compensation, wherein the error compensation function includes a sensor error factor, a measurement error factor, a model error factor, and an environmental error factor, and represents the blade tip vibration displacement error as a combination of multiple error factors;

[0024] A nonconvex sparse regularization model of the displacement response result is established using the nonconvex inverse tangent penalty term, and the nonconvex sparse regularization model is sparsely shrunk using the sensor error factor, measurement error factor, model error factor and environmental error factor as constraints.

[0025] Based on the optimization algorithm, the non-convex sparse regularized model after sparse contraction is solved to obtain the displacement response results after sparse tightening. The blade tip vibration displacement model is multi-parameter coupled by the displacement response results after sparse tightening and the error compensation factor to obtain the blade tip torsional vibration correction model.

[0026] In some possible embodiments, the blade tip vibration displacement error is expressed as:

[0027]

[0028] Among them, μ represents the error compensation factor, μ1, μ2, μ3, μ4 are sensor error factors, measurement error factors, model error factors and environmental error factors respectively, and m represents the number of blade tip vibration displacement modes, φ, are the vibration mode coefficient and modal damping ratio of the resonance point respectively;

[0029] When the blade tip vibration displacement model is multi-parameter coupled using the displacement response results after sparse tightening and the error compensation factor, the multi-parameter coupling is expressed as:

[0030]

[0031] Among them, G new, G0 are the multi-parameter coupling results and the displacement response results after sparse tightening, and κ is the coupling coefficient of the non-convex sparse regularization model.

[0032] In some possible embodiments, the method for filtering and reconstructing the blade tip correction signal based on discrete Fourier transform combined with least squares method includes:

[0033] Obtain the blade tip correction signal, process the blade tip correction signal based on discrete Fourier transform to obtain a discrete transformation matrix, reconstruct the discrete transformation matrix according to Shannon sampling theory, and convert the discrete transformation matrix into the corresponding order range;

[0034] The discrete transformation matrix after matrix reconstruction is subjected to order translation compression to obtain a compressed transformation matrix, and the matrix correlation value between the discrete transformation matrix and the compressed transformation matrix is ​​calculated.

[0035] In some possible embodiments, the method for filtering and reconstructing the blade tip correction signal based on discrete Fourier transform combined with least squares method further includes:

[0036] The matrix cross-correlation values ​​are loaded and iteratively reconstructed using the least squares method to obtain sparse estimation values, forming a signal reconstruction matrix. Based on the signal reconstruction matrix, the leaf end correction signal after filtering and reconstruction is obtained by reverse deduction.

[0037] In a second aspect, an embodiment of the present invention provides a blade-end timing signal processing system based on error compensation, which is used to implement the blade-end timing signal processing method based on error compensation as described above. The blade-end timing signal processing system based on error compensation includes:

[0038] The timing signal acquisition module acquires the blade timing signal in real time based on the blade timing sensor. The simulation software draws the blade Campbell diagram based on the blade timing signal and determines the blade tip vibration displacement model in the blade Campbell diagram.

[0039] A correction and reconstruction module is used to extract the blade tip timing signal corresponding to the blade tip vibration displacement model, correct the blade tip vibration displacement model based on the error compensation mechanism combined with the blade tip timing signal, obtain the blade tip torsional vibration correction model, load the blade tip torsional vibration correction model, extract the blade tip correction signal from the blade tip torsional vibration correction model, and filter and reconstruct the blade tip correction signal based on discrete Fourier transform combined with least squares method;

[0040] The deviation elimination module constructs a multi-measurement snapshot matrix based on the blade tip correction signal after filtering and reconstruction, and uses sparse features to eliminate measurement deviations to obtain a deviation-eliminated signal.

[0041] The fault identification module is used to obtain the deviation elimination signal, establish a multi-rotor transfer function matrix based on the deviation elimination signal, extract the blade tip vibration characteristics through the iterative reweighting method combined with the multi-rotor transfer function matrix, and calculate the DTW distance between the blade tip vibration characteristics and the fault database based on the dynamic time warping algorithm. It is determined whether the DTW distance exceeds the distance threshold. If the DTW distance does not exceed the distance threshold, the blade tip fault mode is matched based on the DTW distance between the blade tip vibration characteristics and the fault database.

[0042] In some possible embodiments, the fault identification module includes:

[0043] a matrix building unit, used for acquiring a deviation elimination signal and building a multi-rotor transfer function matrix based on the deviation elimination signal;

[0044] The feature extraction unit extracts the blade tip vibration features through iterative reweighting combined with the multi-rotor transfer function matrix, and calculates the DTW distance between the blade tip vibration features and the fault database based on the dynamic time warping algorithm;

[0045] The fault judgment unit is used to judge whether the DTW distance exceeds the distance threshold. If the DTW distance does not exceed the distance threshold, the blade tip fault mode is matched based on the blade tip vibration characteristics and the DTW distance in the fault database.

[0046] The blade tip timing signal processing method and system based on error compensation of the embodiment of the present invention, when processing the blade tip correction signal, not only adopts discrete Fourier transform combined with least squares method for filtering and reconstruction processing, but also constructs a multi-measurement snapshot matrix for the blade tip correction signal after filtering and reconstruction processing, and uses sparse features to eliminate measurement deviations, which can more effectively mine useful information in the signal and eliminate measurement deviations, thereby obtaining a higher quality deviation elimination signal, so that the obtained deviation elimination signal can be better used for subsequent blade vibration analysis, fault diagnosis and other tasks, providing a more reliable basis.

[0047] Furthermore, the error-compensated blade tip timing signal processing method and system of the present invention provide a clear, quantitative decision-making basis for blade tip fault risk assessment by determining whether the DTW distance exceeds a distance threshold. When the DTW distance exceeds the threshold, a fault risk warning can be issued in a timely manner, prompting relevant personnel to take appropriate measures to prevent further deterioration of the fault. When the DTW distance does not exceed the threshold, it can also provide strong guarantees for the normal operation of the equipment, reducing the risk of false alarms and missed alarms. Through precise blade tip vibration feature extraction and fault diagnosis, potential blade tip fault risks can be detected early, allowing timely maintenance or adjustment measures to avoid further deterioration of the fault, reduce equipment damage and downtime caused by the fault, and thus improve the operational safety of the equipment.

[0048] Furthermore, the blade tip timing signal processing method and system based on error compensation of the embodiment of the present invention, when determining the blade tip vibration displacement model in the blade Campbell diagram, performs time-frequency domain synchronous analysis on the blade timing signal through wavelet transform, and constructs a three-dimensional matrix of speed-frequency-amplitude, which provides a fine and comprehensive signal data basis for the subsequent resonance point extraction, so that the characteristics of blade vibration can be more accurately identified, and the Campbell diagram curve is fitted with the least squares complex finger method to obtain the modal damping ratio and vibration mode coefficient corresponding to the resonance point, which provides key and precise modal parameters for the subsequent construction of the blade tip vibration differential equation, so that the description of the blade vibration characteristics is more accurate. Combining the error compensation mechanism to correct the blade tip vibration displacement model can effectively reduce the influence of various error factors on the model, further enhance the accuracy and stability of the model, and make it closer to actual working conditions.

[0049] Furthermore, the blade tip timing signal processing method and system based on error compensation of the embodiment of the present invention provides a method for correcting the blade tip vibration displacement model based on the error compensation mechanism combined with the blade tip timing signal, constructs a sparse regularization model based on the non-convex inverse tangent penalty term, and iteratively shrinks the solution space through the proximal gradient method to effectively suppress the frequency conversion harmonic interference. Moreover, through the error compensation mechanism and sparse reconstruction technology, the monitoring accuracy of traditional multi-sensors is achieved under a single sensor layout. The corrected blade tip vibration displacement model can accurately invert crack propagation, and is combined with the dynamic time warping (DTW) algorithm to match the fault database, which significantly reduces the false alarm rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0051] Figure 1 1 is a schematic diagram of an implementation flow of a blade tip timing signal processing method based on error compensation provided by an embodiment of the present invention;

[0052] Figure 2 A schematic diagram of a method for determining a blade tip vibration displacement model in a Campbell diagram of a blade is shown;

[0053] Figure 3 A schematic diagram of the implementation process of a method for correcting the blade tip vibration displacement model based on an error compensation mechanism combined with blade tip timing signals is shown;

[0054] Figure 4 The following is a flow chart showing the implementation of a method for filtering and reconstructing the blade tip correction signal based on discrete Fourier transform combined with least squares method;

[0055] Figure 5 The schematic diagram of the structure of the blade tip timing signal processing system based on error compensation is shown. DETAILED DESCRIPTION

[0056] To enable those skilled in the art to better understand the technical solutions of the present invention, the present invention is further described below in conjunction with the accompanying drawings and specific embodiments. It is apparent that the described embodiments are only a portion of the embodiments of the present invention, rather than all of them. Based on the described embodiments of the present invention, all other embodiments obtained by those skilled in the art without requiring creative effort are within the scope of protection of the present invention.

[0057] Unless otherwise specified, the technical terms or scientific terms used in the embodiments of the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The terms "including" or "comprising" used in the embodiments of the present invention neither limit the shapes, numbers, steps, actions, operations, components, originals and / or their groups mentioned, nor exclude the appearance or addition of one or more other different shapes, numbers, steps, actions, operations, components, originals and / or their groups, or the addition of these. In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number and order of the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0058] Unless otherwise specifically stated, the relative arrangements of the components and steps, numerical expressions, and numerical values ​​set forth in these embodiments do not limit the scope of the present invention. At the same time, it should be understood that, for ease of description, the dimensions of the various parts shown in the drawings are not drawn in accordance with actual proportional relationships, and that the techniques, methods, and devices known to those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, the techniques, methods, and devices shown should be considered part of the authorized specification. In all examples shown and discussed herein, any specific other examples may have different values. It should be noted that similar symbols and letters represent similar items in the following figures, and therefore, once an item is defined in one figure, it does not need to be further discussed in subsequent figures.

[0059] In the description of the embodiments of the present invention, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In the embodiments of the present invention, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in the embodiments of the present invention and the features of different embodiments or examples, unless they are mutually inconsistent.

[0060] Below, the exemplary embodiments according to the present invention will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments of the present invention, and it should be understood that the present invention is not limited to the exemplary embodiments described herein.

[0061] The existing method uses a bandpass filter to filter the signal, but it mainly performs simple filtering on the bandpass filter frequency band range generated by the aliasing frequency, which is insufficient to eliminate the measurement deviation, resulting in a deviation between the final signal and the actual blade tip torsional vibration.

[0062] In response to the above problems, an embodiment of the present invention proposes a blade tip timing signal processing method and system based on error compensation. In short, when the method is implemented, the blade timing signal is first acquired in real time based on the blade timing sensor, and the simulation software draws the blade Campbell diagram based on the blade timing signal to determine the blade tip vibration displacement model in the blade Campbell diagram. Then, based on the error compensation mechanism and the blade tip timing signal, the blade tip vibration displacement model is corrected to obtain a blade tip torsional vibration correction model, and the blade tip correction signal in the blade tip torsional vibration correction model is extracted. The blade tip correction signal is filtered and reconstructed based on the discrete Fourier transform combined with the least squares method, and a multi-measurement snapshot matrix is ​​constructed for the blade tip correction signal after filtering and reconstruction. The measurement deviation is eliminated using sparse features to obtain a deviation-eliminated signal after the deviation is eliminated.

[0063] In an embodiment of the present invention, when processing the blade tip correction signal, not only is discrete Fourier transform combined with the least squares method used for filtering and reconstruction processing, but a multi-measurement snapshot matrix is ​​also constructed for the blade tip correction signal after filtering and reconstruction processing, and sparse features are used to eliminate measurement deviations. This can more effectively mine useful information in the signal and eliminate measurement deviations, thereby obtaining a higher quality deviation elimination signal, so that the obtained deviation elimination signal can be better used for subsequent blade vibration analysis, fault diagnosis and other tasks, providing a more reliable basis.

[0064] The embodiment of the present invention provides a blade tip timing signal processing method based on error compensation. Figure 1 The following is a schematic diagram of a blade tip timing signal processing method based on error compensation. The blade tip timing signal processing method based on error compensation specifically includes:

[0065] Step S10: obtaining a blade timing signal in real time based on a blade timing sensor; drawing a blade Campbell diagram based on the blade timing signal by simulation software; and determining a blade tip vibration displacement model in the blade Campbell diagram.

[0066] It should be noted that blade timing sensors include but are not limited to eddy current sensors, fiber optic sensors, capacitive sensors, and laser sensors, and blade timing signals include but are not limited to time pulse signals, phase signals, frequency signals, and displacement signals. In this embodiment, the blade Campbell diagram intuitively shows the relationship between the blade vibration frequency and the rotational speed, which helps to quickly identify the resonance point and potential vibration problems of the blade, and facilitates a comprehensive evaluation of the vibration performance of the blade.

[0067] Step S20: extract the blade tip timing signal corresponding to the blade tip vibration displacement model, and correct the blade tip vibration displacement model based on the error compensation mechanism combined with the blade tip timing signal. Introducing the error compensation mechanism to correct the blade tip vibration displacement model can effectively reduce the model deviation caused by factors such as sensor error and environmental interference, so that the corrected blade tip torsional vibration correction model is closer to the actual vibration condition of the blade, thereby obtaining the blade tip torsional vibration correction model.

[0068] Step S30 , loading the blade tip torsional vibration correction model, extracting the blade tip correction signal from the blade tip torsional vibration correction model, and filtering and reconstructing the blade tip correction signal based on discrete Fourier transform combined with least squares method.

[0069] In an embodiment of the present invention, discrete Fourier transform combined with least squares method is used to filter and reconstruct the blade tip correction signal, which can not only effectively remove the noise and interference components in the signal, but also retain the useful signal to the greatest extent, improve the signal-to-noise ratio and quality of the signal, and make the signal more clearly and accurately reflect the actual vibration state of the blade.

[0070] Step S40 , constructing a multi-measurement snapshot matrix for the leaf tip correction signal after filtering and reconstruction, eliminating measurement deviations using sparse features, and obtaining a deviation-eliminated signal after deviation elimination.

[0071] Step S50: Obtain a deviation elimination signal, establish a multi-rotor transfer function matrix based on the deviation elimination signal, extract the blade tip vibration characteristics by combining the multi-rotor transfer function matrix through an iterative reweighting method, and calculate the DTW distance between the blade tip vibration characteristics and the fault database based on a dynamic time warping algorithm.

[0072] Step S60: Determine whether the DTW distance exceeds a distance threshold.

[0073] Step S70 : If the DTW distance does not exceed the distance threshold, the blade tip fault mode is matched based on the blade tip vibration signature and the DTW distance in the fault database.

[0074] If the DTW distance exceeds the distance threshold, it is determined that the risk of blade tip failure is small, and a deviation elimination signal after the deviation is eliminated is sent and stored.

[0075] In this embodiment, determining whether the DTW distance exceeds a distance threshold provides a clear, quantitative basis for blade tip fault risk assessment. When the DTW distance exceeds the threshold, a timely fault risk warning is issued, prompting personnel to take appropriate measures to prevent further deterioration of the fault. When the DTW distance does not exceed the threshold, the system effectively safeguards the normal operation of the equipment, reducing the risk of false alarms and missed alarms. Through precise blade tip vibration feature extraction and fault diagnosis, potential blade tip fault risks can be detected early, allowing timely repair or adjustment measures to prevent further deterioration of the fault, reduce equipment damage and downtime caused by the fault, and thus improve equipment operational safety.

[0076] In an embodiment of the present invention, when processing the blade tip correction signal, not only is discrete Fourier transform combined with the least squares method used for filtering and reconstruction processing, but a multi-measurement snapshot matrix is ​​also constructed for the blade tip correction signal after filtering and reconstruction processing, and sparse features are used to eliminate measurement deviations. This can more effectively mine useful information in the signal and eliminate measurement deviations, thereby obtaining a higher quality deviation elimination signal, so that the obtained deviation elimination signal can be better used for subsequent blade vibration analysis, fault diagnosis and other tasks, providing a more reliable basis.

[0077] The embodiment of the present invention provides a method for determining a blade tip vibration displacement model in a blade Campbell diagram. Figure 2 A schematic flow chart of a method for determining a blade tip vibration displacement model in a blade Campbell diagram is shown. The method for determining a blade tip vibration displacement model in a blade Campbell diagram specifically includes:

[0078] Step S101 identifies the blade timing signal in the Campbell diagram and performs time-frequency domain synchronization analysis on the blade timing signal. Using time-frequency analysis techniques, such as short-time Fourier transform (STFT) or wavelet transform, the blade timing signal is processed to obtain synchronization information in the time and frequency domains. This helps understand how the frequency components of blade vibration change over time or speed, and extracts the speed-frequency-amplitude three-dimensional matrix of the signal based on wavelet transform.

[0079] It should be noted that the three-dimensional matrix can intuitively reflect the changing trend of the blade resonance frequency with the rotational speed (such as the slope of the Campbell diagram), providing a data basis for the subsequent resonance point identification.

[0080] Step S102: The speed-frequency-amplitude three-dimensional matrix of the loading signal is used to extract at least one set of resonance points from the blade's Campbell diagram based on the speed-frequency-amplitude three-dimensional matrix. The points where the blade resonates are identified by analyzing the frequency and amplitude data in the three-dimensional matrix. Resonance points are typically the frequencies at which the vibration amplitude reaches a peak at a specific speed. This three-dimensional matrix-based analysis method simultaneously considers speed, frequency, and amplitude, improving the accuracy and reliability of resonance point extraction.

[0081] Step S103 , fitting the Campbell plot curve using the least squares complex fingering method to obtain the modal damping ratio and vibration mode coefficient corresponding to the resonance point; the least squares complex fingering method can provide high-precision curve fitting results and accurately reflect the vibration characteristics of the blade.

[0082] In the embodiment of the present invention, the mode coefficient of the resonance point is calculated by the following formula:

[0083]

[0084] Among them, φ, are the vibration mode coefficient and modal damping ratio of the resonance point, S(n,f,A) represents the speed-frequency-amplitude three-dimensional matrix of the signal, G represents the number of resonance points, A,f,f g Represents amplitude, signal frequency, and resonant frequency respectively, K g ,λ,R g are mechanical power gain factor, blade tip ratio, blade length, R g is the blade adjustment coefficient.

[0085] In step S104, a blade tip vibration differential equation is constructed based on the Euler-Bernoulli algorithm, the modal damping ratio, mode shape coefficient, and resonance frequency corresponding to the resonance point are loaded, and the displacement response is solved based on the finite element method to obtain a blade tip vibration displacement model including the displacement response result.

[0086] In this embodiment, the Euler-Bernoulli algorithm is used to construct the differential equation for blade tip vibration, utilizing the Euler-Bernoulli beam theory to establish the differential equation describing blade tip vibration. The Euler-Bernoulli algorithm assumes that the cross-section of the beam remains planar after deformation and ignores the effects of shear deformation. The finite element method (FEM) is used to numerically solve the differential equation. The FEM discretizes a continuous structure into a finite number of elements, enabling the numerical solution of complex engineering problems.

[0087] In an embodiment of the present invention, when determining the blade tip vibration displacement model in the blade Campbell diagram, the blade timing signal is synchronously analyzed in the time and frequency domains by wavelet transform, and a three-dimensional matrix of speed-frequency-amplitude is constructed, which provides a fine and comprehensive signal data basis for the subsequent extraction of resonance points, so that the characteristics of blade vibration can be more accurately identified. The Campbell diagram curve is fitted by the least squares complex fingering method to obtain the modal damping ratio and vibration mode coefficient corresponding to the resonance point, which provides key and accurate modal parameters for the subsequent construction of the blade tip vibration differential equation, making the description of the blade vibration characteristics more accurate. The correction of the blade tip vibration displacement model in combination with the error compensation mechanism can effectively reduce the influence of various error factors on the model, further enhance the accuracy and stability of the model, and make it closer to the actual working conditions.

[0088] The embodiment of the present invention provides a method for correcting a blade tip vibration displacement model based on an error compensation mechanism combined with a blade tip timing signal. Figure 3 A schematic flow chart of a method for correcting a blade tip vibration displacement model based on an error compensation mechanism combined with a blade tip timing signal is shown. The method for correcting a blade tip vibration displacement model based on an error compensation mechanism combined with a blade tip timing signal specifically includes:

[0089] Step S201 : Pre-constructing an error compensation function based on sensor error compensation, measurement system error compensation, model error compensation, and environmental error compensation. The error compensation function includes a sensor error factor, a measurement error factor, a model error factor, and an environmental error factor, and represents the blade tip vibration displacement error as a combination of multiple error factors.

[0090] In the embodiment of the present invention, the error compensation function comprehensively considers multiple error sources, including the accuracy limitations of the sensor itself, the instability of the measurement system, the incompleteness of the model assumptions, and environmental changes. It can compensate for the blade tip vibration displacement error in an all-round way, which helps to more accurately assess the vibration risk of the blade, take measures to prevent potential failures in advance, and optimize equipment maintenance plans and resource allocation.

[0091] In step S202, a nonconvex sparse regularized model of the displacement response is established using a nonconvex inverse tangent penalty term. The nonconvex sparse regularized model is then subjected to sparse contraction using the sensor error factor, measurement error factor, model error factor, and environmental error factor as constraints. The nonconvex inverse tangent penalty term is then introduced into the displacement response model. The nonconvex penalty term helps better handle sparsity during the optimization process, specifically by highlighting the primary components in the model and suppressing noise and other secondary components. The nonconvex sparse regularized model is then solved using an optimization algorithm to achieve sparse contraction. The purpose of sparse contraction is to reduce redundant information in the model and enhance its stability and generalization capabilities.

[0092] In step S203, the non-convex sparse regularized model after sparse contraction is solved based on the optimization algorithm to obtain the displacement response result after sparse tightening. The blade tip vibration displacement model is multi-parameter coupled through the displacement response result after sparse tightening and the error compensation factor to obtain a blade tip torsional vibration correction model. The displacement response result after sparse tightening obtained by the optimization algorithm, combined with the error compensation factor, can significantly improve the accuracy of the blade tip vibration displacement model. The multi-parameter coupling process can comprehensively consider various error correction terms, so that the corrected model can more realistically reflect the actual vibration condition of the blade tip.

[0093] In this embodiment, the blade tip vibration displacement error is expressed as:

[0094]

[0095] Among them, μ represents the error compensation factor, μ1, μ2, μ3, μ4 are sensor error factors, measurement error factors, model error factors and environmental error factors respectively, and m represents the number of blade tip vibration displacement modes, φ, are the vibration mode coefficient and modal damping ratio of the resonance point respectively.

[0096] When the blade tip vibration displacement model is multi-parameter coupled using the displacement response results after sparse tightening and the error compensation factor, the multi-parameter coupling is expressed as:

[0097]

[0098] Among them, G new , G0 are the multi-parameter coupling results and the displacement response results after sparse tightening, and κ is the coupling coefficient of the non-convex sparse regularization model.

[0099] In an embodiment of the present invention, a method for correcting the blade tip vibration displacement model based on an error compensation mechanism combined with a blade tip timing signal is provided. A sparse regularization model is constructed based on a non-convex inverse tangent penalty term. The solution space is iteratively contracted by the proximal gradient method to effectively suppress the interference of rotational harmonics. Furthermore, through the error compensation mechanism and sparse reconstruction technology, the monitoring accuracy of traditional multi-sensors is achieved under a single-sensor layout. The corrected blade tip vibration displacement model can accurately invert crack propagation, and combined with the dynamic time warping (DTW) algorithm to match the fault database, the false alarm rate is significantly reduced.

[0100] The embodiment of the present invention provides a method for filtering and reconstructing the blade tip correction signal based on discrete Fourier transform combined with least squares method. Figure 4 The following is a flow chart showing a method for filtering and reconstructing the blade tip correction signal based on discrete Fourier transform combined with least squares method. The method for filtering and reconstructing the blade tip correction signal based on discrete Fourier transform combined with least squares method specifically includes:

[0101] Step S301: Obtain the leaf-end correction signal. Process the leaf-end correction signal using discrete Fourier transform to obtain a discrete transformation matrix. Reconstruct the discrete transformation matrix based on Shannon sampling theory and convert the discrete transformation matrix into a corresponding order range. Reconstructing the discrete transformation matrix based on Shannon sampling theory ensures that signal sampling meets requirements, avoids aliasing, and thus ensures signal integrity and accuracy. This lays a reliable data foundation for subsequent signal processing steps.

[0102] Step S302 performs order translation compression on the reconstructed discrete transformation matrix to obtain a compressed transformation matrix. The matrix correlation value between the discrete transformation matrix and the compressed transformation matrix is ​​calculated. This order translation compression is then performed on the reconstructed discrete transformation matrix. This effectively reduces the amount of data and the computational complexity of subsequent processing while preserving the key features and primary frequency components of the signal. This improves signal processing efficiency without losing important information.

[0103] Step S303 loads the matrix cross-correlation values ​​and iteratively reconstructs them using the least squares method to obtain sparse estimates, forming a signal reconstruction matrix. Based on the signal reconstruction matrix, the filtered and reconstructed blade tip correction signal is obtained by reverse engineering. The reconstruction results are represented as sparse estimates and formed into a signal reconstruction matrix. This helps simplify signal representation, highlight key features, and suppress noise and interference. This achieves effective filtering of the blade tip correction signal, improves the signal-to-noise ratio and quality, and provides a clearer and more accurate signal foundation for subsequent vibration analysis and fault diagnosis.

[0104] On the other hand, the embodiment of the present invention further provides a blade tip timing signal processing system based on error compensation. Figure 5 The schematic diagram of the structure of the blade-end timing signal processing system based on error compensation is shown. The blade-end timing signal processing system based on error compensation specifically includes:

[0105] The timing signal acquisition module 100 acquires the blade timing signal in real time based on the blade timing sensor. The simulation software draws the blade Campbell diagram based on the blade timing signal and determines the blade tip vibration displacement model in the blade Campbell diagram.

[0106] The correction and reconstruction module 200 is used to extract the blade tip timing signal corresponding to the blade tip vibration displacement model, correct the blade tip vibration displacement model based on the error compensation mechanism combined with the blade tip timing signal, obtain the blade tip torsional vibration correction model, load the blade tip torsional vibration correction model, extract the blade tip correction signal from the blade tip torsional vibration correction model, and filter and reconstruct the blade tip correction signal based on discrete Fourier transform combined with least squares method.

[0107] The deviation elimination module 300 constructs a multi-measurement snapshot matrix for the blade tip correction signal after filtering and reconstruction, eliminates the measurement deviation using sparse features, and obtains a deviation-eliminated signal after the deviation is eliminated.

[0108] Fault identification module 400 is used to obtain a deviation elimination signal, establish a multi-rotor transfer function matrix based on the deviation elimination signal, extract blade tip vibration characteristics through an iterative reweighting method combined with the multi-rotor transfer function matrix, calculate the DTW distance between the blade tip vibration characteristics and a fault database based on a dynamic time warping algorithm, determine whether the DTW distance exceeds a distance threshold, and if the DTW distance does not exceed the distance threshold, match the blade tip fault mode based on the DTW distance between the blade tip vibration characteristics and the fault database.

[0109] In this embodiment, the fault identification module 400 includes:

[0110] The matrix construction unit 410 is configured to obtain a deviation elimination signal and to establish a multi-rotor transfer function matrix based on the deviation elimination signal.

[0111] The feature extraction unit 420 extracts the blade tip vibration features by combining the iterative reweighting method with the multi-rotor transfer function matrix, and calculates the DTW distance between the blade tip vibration features and the fault database based on the dynamic time warping algorithm.

[0112] The fault judgment unit 430 is configured to judge whether the DTW distance exceeds a distance threshold. If the DTW distance does not exceed the distance threshold, the fault mode of the blade tip is matched based on the blade tip vibration characteristics and the DTW distance in the fault database.

[0113] In summary, the present invention provides a blade tip timing signal processing method and system based on error compensation. In an embodiment of the present invention, when processing the blade tip correction signal, not only is discrete Fourier transform combined with least squares method used for filtering and reconstruction processing, but a multi-measurement snapshot matrix is ​​also constructed for the blade tip correction signal after filtering and reconstruction processing, and sparse features are used to eliminate measurement deviations. This can more effectively mine useful information in the signal and eliminate measurement deviations, thereby obtaining a higher quality deviation elimination signal, so that the obtained deviation elimination signal can be better used for subsequent blade vibration analysis, fault diagnosis and other tasks, providing a more reliable basis.

[0114] In an embodiment of the present invention, when determining the blade tip vibration displacement model in the blade Campbell diagram, the blade timing signal is synchronously analyzed in the time and frequency domains by wavelet transform, and a three-dimensional matrix of speed-frequency-amplitude is constructed, which provides a fine and comprehensive signal data basis for the subsequent extraction of resonance points, so that the characteristics of blade vibration can be more accurately identified. The Campbell diagram curve is fitted by the least squares complex fingering method to obtain the modal damping ratio and vibration mode coefficient corresponding to the resonance point, which provides key and accurate modal parameters for the subsequent construction of the blade tip vibration differential equation, making the description of the blade vibration characteristics more accurate. The correction of the blade tip vibration displacement model in combination with the error compensation mechanism can effectively reduce the influence of various error factors on the model, further enhance the accuracy and stability of the model, and make it closer to the actual working conditions.

[0115] In an embodiment of the present invention, a method for correcting the blade tip vibration displacement model based on an error compensation mechanism combined with a blade tip timing signal is provided. A sparse regularization model is constructed based on a non-convex inverse tangent penalty term. The solution space is iteratively contracted by the proximal gradient method to effectively suppress the interference of rotational harmonics. Furthermore, through the error compensation mechanism and sparse reconstruction technology, the monitoring accuracy of traditional multi-sensors is achieved under a single-sensor layout. The corrected blade tip vibration displacement model can accurately invert crack propagation, and combined with the dynamic time warping (DTW) algorithm to match the fault database, the false alarm rate is significantly reduced.

[0116] It will be understood that the above embodiments are merely exemplary embodiments for illustrating the principles of the present invention, and the present invention is not limited thereto. Those skilled in the art will appreciate that various modifications and improvements can be made without departing from the spirit and substance of the present invention, and such modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A blade tip timing signal processing method based on error compensation, characterized in that: The method comprises: The blade timing sensor acquires the blade timing signal in real time, and the simulation software draws the blade Campbell diagram based on the blade timing signal to determine the blade tip vibration displacement model in the blade Campbell diagram; Extract the blade tip timing signal corresponding to the blade tip vibration displacement model, and correct the blade tip vibration displacement model based on the error compensation mechanism and the blade tip timing signal to obtain the blade tip torsional vibration correction model. Load the blade tip torsional vibration correction model, extract the blade tip correction signal from the blade tip torsional vibration correction model, and filter and reconstruct the blade tip correction signal based on discrete Fourier transform combined with least squares method; A multi-measurement snapshot matrix is ​​constructed for the blade tip correction signal after filtering and reconstruction, and the sparse features are used to eliminate the measurement deviation to obtain the deviation-eliminated signal.

2. The blade tip timing signal processing method based on error compensation according to claim 1, characterized in that: The method further comprises: Obtain the deviation cancellation signal, establish a multi-rotor transfer function matrix based on the deviation cancellation signal, extract the blade tip vibration characteristics through the iterative reweighting method combined with the multi-rotor transfer function matrix, and calculate the DTW distance between the blade tip vibration characteristics and the fault database based on the dynamic time warping algorithm to determine whether the DTW distance exceeds the distance threshold; If the DTW distance does not exceed the distance threshold, the blade tip fault mode is matched based on the DTW distance between the blade tip vibration characteristics and the fault database.

3. The blade tip timing signal processing method based on error compensation according to claim 1, characterized in that: The method for determining the blade tip vibration displacement model in the blade Campbell diagram comprises: Identify blade timing signals in the Campbell diagram, perform synchronous analysis of blade timing signals in the time and frequency domains, and extract the speed-frequency-amplitude three-dimensional matrix of the signal based on wavelet transform; A speed-frequency-amplitude three-dimensional matrix of the loading signal is used to extract at least one set of resonance points in the Campbell diagram of the blade based on the speed-frequency-amplitude three-dimensional matrix; The Campbell diagram curve is fitted using the least squares complex fingering method to obtain the modal damping ratio and vibration mode coefficient corresponding to the resonance point; The blade tip vibration differential equation is constructed based on the Euler-Bernoulli algorithm. The modal damping ratio, mode coefficient, and resonance frequency corresponding to the resonance point are loaded. The displacement response is solved based on the finite element method, and the blade tip vibration displacement model including the displacement response results is obtained.

4. The blade tip timing signal processing method based on error compensation according to claim 3, characterized in that: The mode coefficient of the resonance point is calculated using the following formula: Among them, φ, are the vibration mode coefficient and modal damping ratio of the resonance point, S(n,f,A) represents the speed-frequency-amplitude three-dimensional matrix of the signal, G represents the number of resonance points, A,f,f g Represents amplitude, signal frequency, and resonant frequency respectively, K g ,λ,R g are mechanical power gain factor, blade tip ratio, blade length, R g is the blade adjustment coefficient.

5. The blade tip timing signal processing method based on error compensation according to any one of claims 1 to 4, characterized in that: The method for correcting the blade tip vibration displacement model based on the error compensation mechanism and the blade tip timing signal specifically includes: Pre-building an error compensation function based on sensor error compensation, measurement system error compensation, model error compensation, and environmental error compensation, wherein the error compensation function includes a sensor error factor, a measurement error factor, a model error factor, and an environmental error factor, and represents the blade tip vibration displacement error as a combination of multiple error factors; A nonconvex sparse regularization model of the displacement response result is established using the nonconvex inverse tangent penalty term, and the nonconvex sparse regularization model is sparsely shrunk using the sensor error factor, measurement error factor, model error factor and environmental error factor as constraints. Based on the optimization algorithm, the non-convex sparse regularized model after sparse contraction is solved to obtain the displacement response results after sparse tightening. The blade tip vibration displacement model is multi-parameter coupled by the displacement response results after sparse tightening and the error compensation factor to obtain the blade tip torsional vibration correction model.

6. The blade tip timing signal processing method based on error compensation according to claim 5, characterized in that: The blade tip vibration displacement error is expressed as: Among them, μ represents the error compensation factor, μ1, μ2, μ3, μ4 are sensor error factors, measurement error factors, model error factors and environmental error factors respectively, and m represents the number of blade tip vibration displacement modes, φ, are the vibration mode coefficient and modal damping ratio of the resonance point respectively; When the blade tip vibration displacement model is multi-parameter coupled using the displacement response results after sparse tightening and the error compensation factor, the multi-parameter coupling is expressed as: Among them, G new , G0 are the multi-parameter coupling results and the displacement response results after sparse tightening, and κ is the coupling coefficient of the non-convex sparse regularization model.

7. The blade tip timing signal processing method based on error compensation according to any one of claims 1 to 4, characterized in that: The method for filtering and reconstructing the blade tip correction signal based on discrete Fourier transform combined with least squares method includes: Obtain the blade tip correction signal, process the blade tip correction signal based on discrete Fourier transform to obtain a discrete transformation matrix, reconstruct the discrete transformation matrix according to Shannon sampling theory, and convert the discrete transformation matrix into the corresponding order range; The discrete transformation matrix after matrix reconstruction is subjected to order translation compression to obtain a compressed transformation matrix, and the matrix correlation value between the discrete transformation matrix and the compressed transformation matrix is ​​calculated.

8. The blade tip timing signal processing method based on error compensation according to claim 7, characterized in that: The method for filtering and reconstructing the blade tip correction signal based on discrete Fourier transform combined with least squares method also includes: The matrix cross-correlation values ​​are loaded and iteratively reconstructed using the least squares method to obtain sparse estimation values, forming a signal reconstruction matrix. Based on the signal reconstruction matrix, the leaf end correction signal after filtering and reconstruction is obtained by reverse deduction.

9. A blade tip timing signal processing system based on error compensation, used to implement the blade tip timing signal processing method based on error compensation according to any one of claims 1 to 8, characterized in that: The blade tip timing signal processing system based on error compensation includes: The timing signal acquisition module acquires the blade timing signal in real time based on the blade timing sensor. The simulation software draws the blade Campbell diagram based on the blade timing signal and determines the blade tip vibration displacement model in the blade Campbell diagram; A correction and reconstruction module is used to extract the blade tip timing signal corresponding to the blade tip vibration displacement model, correct the blade tip vibration displacement model based on the error compensation mechanism combined with the blade tip timing signal, obtain the blade tip torsional vibration correction model, load the blade tip torsional vibration correction model, extract the blade tip correction signal from the blade tip torsional vibration correction model, and filter and reconstruct the blade tip correction signal based on discrete Fourier transform combined with least squares method; The deviation elimination module constructs a multi-measurement snapshot matrix based on the blade tip correction signal after filtering and reconstruction, and uses sparse features to eliminate measurement deviations to obtain a deviation-eliminated signal. The fault identification module is used to obtain the deviation elimination signal, establish a multi-rotor transfer function matrix based on the deviation elimination signal, extract the blade tip vibration characteristics through the iterative reweighting method combined with the multi-rotor transfer function matrix, and calculate the DTW distance between the blade tip vibration characteristics and the fault database based on the dynamic time warping algorithm. It is determined whether the DTW distance exceeds the distance threshold. If the DTW distance does not exceed the distance threshold, the blade tip fault mode is matched based on the DTW distance between the blade tip vibration characteristics and the fault database.

10. The blade tip timing signal processing system based on error compensation according to claim 9, characterized in that: The fault identification module includes: a matrix building unit, used for acquiring a deviation elimination signal and building a multi-rotor transfer function matrix based on the deviation elimination signal; The feature extraction unit extracts the blade tip vibration features through iterative reweighting combined with the multi-rotor transfer function matrix, and calculates the DTW distance between the blade tip vibration features and the fault database based on the dynamic time warping algorithm; The fault judgment unit is used to judge whether the DTW distance exceeds the distance threshold. If the DTW distance does not exceed the distance threshold, the blade tip fault mode is matched based on the blade tip vibration characteristics and the DTW distance in the fault database.

Images