Method and device for obtaining blade vibration modal parameters

Abstract

The embodiment of the present invention discloses a method and device for obtaining blade vibration modal parameters. The method includes: obtaining an undersampled pulse response signal of the blade; constructing a compressed sensing dictionary using the real part projection of the Laplace wavelet, wherein there is a similarity relationship between the real part projection and the undersampled pulse response signal; determining a measurement matrix according to the installation position of the target sensor; determining a non-undersampled pulse response signal according to the dictionary, the measurement matrix and the undersampled pulse response signal; and processing the non-undersampled pulse response signal using a matrix beam method to obtain the vibration modal parameters of the blade. The present invention solves the problem of poor processing effect caused by the compressed sensing algorithm using the Fourier dictionary in the related art, and achieves the technical effect of effectively improving the probability of complete signal recovery, reducing the number of required sensors, and thereby improving the accurate estimation of modal parameters.

Description

Technical Field

[0001] The present invention relates to the technical field of blade vibration measurement, and in particular to a method and device for acquiring blade vibration modal parameters. Background Art

[0002] In aircraft engines, blade breakage often occurs due to the harsh working environment, which in turn leads to engine structural failure and irreparable losses. Effectively monitoring the vibration of rotating blades is an important guarantee for the safety and performance of aircraft engines. In response to this problem, domestic and foreign scholars have conducted a number of studies. Among them, blade tip timing technology (BTT) can monitor blade vibration in real time online without contact, and is a promising blade vibration measurement technology.

[0003] like Figure 1 As shown in the figure, the BTT method requires that several optical sensors be arranged circumferentially on the casing, and a synchronous speed sensor be installed on the shaft to obtain a reference signal and monitor speed changes. The principle of BTT is to reconstruct the blade vibration signal by obtaining the arrival time series of the blade tip passing through the sensor through a series of data processing methods.

[0004] According to Shannon's sampling theorem, when sampling a signal, if you want to get the original signal without distortion, the sampling frequency must be twice the maximum frequency of the components contained in the original signal. However, since the BTT measurement signal is limited by the number of sensors and the impeller speed, it is a typical under-sampled signal, that is, the sampled signal is a distorted blade vibration signal. Many processing methods for blade tip timing signals have emerged to identify parameters such as blade vibration frequency and amplitude from the under-sampled blade tip timing signal.

[0005] According to the theory of compressed sensing (CS), as long as the signal is sparse in a certain domain, the complete signal can be accurately reconstructed from the undersampled data. At present, compressed sensing technology has been applied to the recovery of BTT signals. The vibration signal is considered as a multimodal signal model, and the Fourier dictionary is selected to fit the vibration signal. Figure 2 As shown in the figure, one signal is a simulated blade vibration signal synthesized by a sine wave and two cosine waves, which is sparse in the frequency domain. The other signal is a reconstructed signal restored by the Fourier dictionary through the compressed sensing algorithm. It can be seen that the original signal is basically consistent with the reconstructed signal.

[0006] However, the feasibility of Fourier dictionary in application depends largely on the similarity between the dictionary and the original signal characteristics. When the vibration characteristics of the blade vibration signal are greatly different from those of the sine and cosine signals, or the blade vibration signal is not sparse enough in the frequency domain, the signal reconstructed by the compressed sensing algorithm is often far from the original signal. Figure 3 As shown in the figure, when a pulse excitation is given to the blade, the impulse response signal of the signal can be obtained. The signal is attenuated in the time domain and is not sparse in the frequency domain. After undersampling, the signal reconstructed using the Fourier dictionary compressed sensing algorithm can be seen to be very different from each other.

[0007] It can be seen that the prior art has the following defects:

[0008] 1) When the blade vibration signal is not sparse in the frequency domain, the compressed sensing algorithm using the Fourier dictionary has a poor reconstruction effect and cannot recover the complete signal from the undersampled BTT signal.

[0009] 2) In the blade vibration displacement signal reconstructed by the Fourier dictionary compressed sensing algorithm, only information such as frequency and amplitude can be extracted, and it is impossible to analyze and obtain the vibration damping information.

[0010] In the related art, the blade vibration displacement signal reconstructed by the compressed sensing algorithm using the Fourier dictionary results in poor processing effect, and no effective solution has been proposed yet. Summary of the invention

[0011] The embodiments of the present invention provide a method, device, electronic device and computer-readable storage medium for processing a business request, so as to at least solve the technical problem in the related art that the blade vibration displacement signal is reconstructed by the compressed sensing algorithm using the Fourier dictionary, resulting in poor processing effect.

[0012] According to one aspect of an embodiment of the present invention, a method for acquiring vibration modal parameters of a blade is provided, comprising: acquiring an undersampled pulse response signal of a blade; constructing a compressed sensing dictionary using a real part projection of a Laplace wavelet, wherein a similarity relationship exists between the real part projection and the undersampled pulse response signal; determining a measurement matrix according to an installation position of a target sensor; determining a non-undersampled pulse response signal according to the dictionary, the measurement matrix and the undersampled pulse response signal; and processing the non-undersampled pulse response signal using a matrix bundle method to obtain vibration modal parameters of the blade.

[0013] Optionally, constructing a compressed sensing dictionary includes: acquiring the number of vibration modes and initial vibration mode parameters of the blade; and determining the dictionary according to the number of vibration modes and the vibration mode parameters.

[0014] Optionally, the dictionary is determined according to the number of vibration modes and the vibration mode parameters, including: determining the same number of sub-dictionaries according to the number of vibration modes, wherein the real part projection of Laplace wavelet is used as the atoms of the sub-dictionary, the frequencies in the vibration mode parameters are within the frequency range of the atoms, and the damping in the vibration mode parameters are within the damping range of the atoms; and merging the same number of sub-dictionaries into the dictionary.

[0015] Optionally, determining a non-undersampled impulse response signal according to the dictionary, the measurement matrix and the undersampled impulse response signal includes: calculating an optimal sparse solution according to the dictionary, the measurement matrix and the undersampled impulse response signal; and calculating the non-undersampled impulse response signal according to the optimal sparse solution.

[0016] According to another aspect of an embodiment of the present invention, a device for acquiring vibration modal parameters of a blade is also provided, including: an acquisition module for acquiring an undersampled pulse response signal of a blade; a construction module for constructing a compressed sensing dictionary using a real part projection of a Laplace wavelet, wherein there is a similarity relationship between the real part projection and the undersampled pulse response signal; a first determination module for determining a measurement matrix according to an installation position of a target sensor; a second determination module for determining a non-undersampled pulse response signal according to the dictionary, the measurement matrix and the undersampled pulse response signal; and a processing module for processing the non-undersampled pulse response signal using a matrix bundle method to obtain the vibration modal parameters of the blade.

[0017] Optionally, the construction module includes: an acquisition unit, used to acquire the number of vibration modes and initial vibration mode parameters of the blade; and a determination unit, used to determine the dictionary according to the number of vibration modes and the vibration mode parameters.

[0018] Optionally, the determination unit includes: a determination sub-unit, used to determine the same number of sub-dictionaries according to the number of vibration modes, wherein the real part projection of Laplace wavelet is used as the atoms of the sub-dictionary, the frequency in the vibration mode parameters is within the frequency range of the atom, and the damping in the vibration mode parameters is within the damping range of the atom; and a merging unit, used to merge the same number of sub-dictionaries into the dictionary.

[0019] Optionally, the second determination module includes: a first calculation unit, used to calculate an optimal sparse solution based on the dictionary, the measurement matrix and the undersampled impulse response signal; and a second calculation unit, used to calculate the non-undersampled impulse response signal based on the optimal sparse solution.

[0020] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is further provided, wherein the computer-readable storage medium includes a stored program, wherein the program executes any of the methods described above when running.

[0021] According to another aspect of an embodiment of the present invention, there is further provided an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute any of the above methods through the computer program.

[0022] In an embodiment of the present invention, an undersampled impulse response signal of a blade is obtained; a compressed sensing dictionary is constructed using the real part projection of the Laplace wavelet, wherein a similarity relationship exists between the real part projection and the undersampled impulse response signal; a measurement matrix is ​​determined according to the installation position of the target sensor; a non-undersampled impulse response signal is determined according to the dictionary, the measurement matrix and the undersampled impulse response signal; and the non-undersampled impulse response signal is processed using a matrix bundle method to obtain vibration modal parameters of the blade. That is, the embodiment of the present invention uses compressed sensing theory and blade tip timing technology to obtain the impulse response signal of the blade, realizes a comprehensive estimation of the vibration modal parameters of the blade under the working conditions of the impeller, and constructs a compressed sensing dictionary using the real part projection of the Laplace wavelet, thereby improving the sparsity of the projection of the impulse response signal on the dictionary, thereby solving the problem of poor processing effect caused by the compressed sensing algorithm using the Fourier dictionary in the related technology to effectively improve the probability of complete signal recovery, reduce the number of sensors required, and thereby improve the technical effect of accurate estimation of modal parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0024] Figure 1 is a schematic diagram of a blade tip timing measurement device provided in the prior art;

[0025] Figure 2 It is a schematic diagram of an original signal and a reconstructed signal provided in the prior art;

[0026] Figure 3 is another schematic diagram of an original signal and a reconstructed signal provided in the prior art;

[0027] Figure 4 is a flow chart of a method for acquiring blade vibration modal parameters according to an embodiment of the present invention;

[0028] Figure 5 is a schematic diagram of the principle of blade tip timing measurement technology according to an embodiment of the present invention;

[0029] Figure 6 is a schematic diagram of the real part projection of the Laplace wavelet according to an embodiment of the present invention;

[0030] Figure 7 is a schematic diagram of a device for acquiring blade vibration modal parameters according to an embodiment of the present invention. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0032] It should be noted that the terms "first", "second" and the like in the specification, claims and drawings of the present invention are used to distinguish different objects rather than to limit a specific order.

[0033] According to one aspect of an embodiment of the present invention, a method for obtaining blade vibration modal parameters is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0034] Figure 4 is a flow chart of a method for obtaining blade vibration modal parameters according to an embodiment of the present invention. Figure 4 As shown, the method comprises the following steps:

[0035] Step S402, obtaining an under-sampled impulse response signal of the blade;

[0036] It should be noted that, in the past, in order to obtain the blade pulse response signal, a strain gauge is often attached to the blade, the blade is stimulated by hammering or other methods, and the vibration signal of the strain gauge is obtained by using a sensor in a static state. The present invention can achieve the acquisition of the under-sampled pulse response signal of the blade when the impeller is working. The above blades include but are not limited to blades of aircraft engines.

[0037] Step S404, constructing a compressed sensing dictionary using the real part projection of the Laplace wavelet, wherein there is a similarity relationship between the real part projection and the under-sampled impulse response signal;

[0038] It should be noted that, based on the theory of compressed sensing, the stronger the sparsity of the signal projection on the dictionary, the higher the probability of reconstructing the original signal, and the number of sensors required to fully restore the signal is positively correlated with the sparsity.

[0039] Step S406, determining a measurement matrix according to the installation position of the target sensor;

[0040] The above-mentioned target sensor is a synchronous speed sensor actually installed on the rotating shaft, wherein the installation quantity and installation position of the target sensor can be set according to the needs of the application scenario.

[0041] Step S408, determining a non-undersampled impulse response signal according to the dictionary, the measurement matrix and the undersampled impulse response signal;

[0042] Step S410: Process the non-undersampled impulse response signal using a matrix bundle method to obtain vibration modal parameters of the blade.

[0043] In an embodiment of the present invention, an undersampled impulse response signal of a blade is obtained; a compressed sensing dictionary is constructed using the real part projection of the Laplace wavelet, wherein a similarity relationship exists between the real part projection and the undersampled impulse response signal; a measurement matrix is ​​determined according to the installation position of the target sensor; a non-undersampled impulse response signal is determined according to the dictionary, the measurement matrix and the undersampled impulse response signal; and the non-undersampled impulse response signal is processed using a matrix bundle method to obtain vibration modal parameters of the blade. That is, the embodiment of the present invention uses compressed sensing theory and blade tip timing technology to obtain the impulse response signal of the blade, realizes a comprehensive estimation of the vibration modal parameters of the blade under the working conditions of the impeller, and constructs a compressed sensing dictionary using the real part projection of the Laplace wavelet, thereby improving the sparsity of the projection of the impulse response signal on the dictionary, thereby solving the problem of poor processing effect caused by the compressed sensing algorithm using the Fourier dictionary in the related technology to effectively improve the probability of complete signal recovery, reduce the number of sensors required, and thereby improve the technical effect of accurate estimation of modal parameters.

[0044] In an optional implementation, constructing a compressed sensing dictionary includes: acquiring the number of vibration modes and initial vibration mode parameters of the blade; and determining a dictionary according to the number of vibration modes and the vibration mode parameters.

[0045] In the above implementation, the dictionary is set using the number of vibration modes of the blade and the initial parameters of the vibration modes, that is, the dictionary of compressed sensing theory is constructed using the real part function of the Laplace wavelet to increase the sparsity of the projection of the impulse response signal on the dictionary, thereby increasing the probability of completely recovering the signal and reducing the number of required sensors.

[0046] In an optional embodiment, a dictionary is determined based on the number of vibration modes and vibration modal parameters, including: determining the same number of sub-dictionaries based on the number of vibration modes, wherein the real part projection of Laplace wavelet is used as the atoms of the sub-dictionary, the frequencies in the vibration modal parameters are within the frequency range of the atoms, and the damping in the vibration modal parameters is within the damping range of the atoms; and merging the same number of sub-dictionaries into a dictionary.

[0047] Optionally, the same number of sub-dictionaries is set according to the number of vibration modes, and the real part projection of Laplace wavelet is used as the atom of the sub-dictionary to ensure that the frequency in the vibration mode parameter is within the frequency range of the atom, and the damping in the vibration mode parameter is within the damping range of the atom; and then the above sub-dictionaries are merged into a dictionary. In the above implementation, the real part projection of Laplace wavelet is used to construct the basis matrix in the compression sense to ensure the sparsity of the impulse response signal of the blade under the basis matrix.

[0048] It should be noted that when designing the dictionary, it is necessary to determine the number of blade vibration modes and the approximate range of each vibration mode parameter, and to divide the frequency domain into multiple frequency bands to add atoms, which can reduce the number of atoms and improve the calculation time.

[0049] In an optional embodiment, a non-undersampled impulse response signal is determined based on a dictionary, a measurement matrix, and an undersampled impulse response signal, including: calculating an optimal sparse solution based on the dictionary, the measurement matrix, and the undersampled impulse response signal; and calculating the non-undersampled impulse response signal based on the optimal sparse solution.

[0050] Optionally, the dictionary, the measurement matrix, and the undersampled impulse response signal can be used as input values ​​of the optimal sparse solution solution, and the optimal sparse solution can be calculated, and the optimal sparse solution can be used as the input value of the non-undersampled impulse response signal solution to calculate the non-undersampled impulse response signal. The non-undersampled impulse response signal can be accurately calculated in the above manner.

[0051] An optional implementation manner of the present invention is described in detail below.

[0052] In order to realize the recovery of under-sampled blade pulse excitation signals based on BTT technology, this scheme uses the real part signal of Laplace wavelet family signal to replace Fourier dictionary signal, making it sparse under the projection of the basis, and reconstructing it through compressed sensing algorithm. The specific implementation plan is as follows:

[0053] 1. Tip timing measurement principle

[0054] Assume that there are K blades on the impeller, and evenly install I fiber optic BTT sensors around the casing to measure the arrival time of each blade passing the sensor. Install a synchronous speed sensor L0 on the shaft, and set a reflective position on the shaft. Get a speed reference signal every time the impeller rotates one circle, which is used as a reference. Let the angular position of each fiber optic BTT sensor relative to the reference sensor L0 along the circumference be expressed as α i , the relative angle of the blade along the rotation direction is expressed as θ k (1≤k≤K), it can be calculated that under ideal conditions, when the kth blade is free of vibration, the theoretical arrival time of the timing sensor at the i-th blade tip in the nth rotation cycle is:

[0055]

[0056] Since the blades vibrate during rotation, the actual arrival time may be ahead of or behind the theoretical arrival time. Assuming the blade vibration displacement is y and the blade rotation radius is r, the actual arrival time of the kth blade passing the nth sensor in the first circle is:

[0057]

[0058] The time difference is compared with the measured arrival time at the sensor and is used to calculate the vibration displacement of the blade:

[0059]

[0060] Among them, f r is the impeller rotation frequency. The principle of blade tip timing measurement technology is as follows Figure 5 shown.

[0061] 2. Equivalent sampling model for blade tip timing measurement

[0062] Any signal can usually be represented as a linear combination of a series of basis functions. In the sparse framework, the set of these basis functions is called a dictionary, and each basis function is called an atom in the dictionary. If a signal's characteristics can be fully represented using only a few atoms in a certain dictionary, then the representation of the signal in this dictionary is called sparse. The fewer atoms used, the more similar the structure of the atoms in the dictionary is to the signal.

[0063] For the kth blade, assuming that L blade tip timing sensors are evenly arranged along the circumference of the casing, I arrival times of the kth blade can be obtained for each rotation of the impeller. Therefore, the equivalent sampling frequency of the BTT measurement method is Lf r , assuming the maximum frequency of the blade vibration signal is f max, then the blade tip timing system can satisfy the Nyquist sampling theorem and can process the BTT sampling signal to obtain blade vibration information. However, due to the limitations of installation conditions, in reality, only I (I≤L) BTT sensors are installed in L selectable installation positions, and the equivalent sampling rate If r ≤Lf r , resulting in If r ≤f max , so the BTT measurement signal is a typical under-sampling signal.

[0064] N points represent the non-undersampled blade impulse response signal X measured by L sensors over a period of time. N×1 According to the theory of harmonic analysis, any signal can usually be expressed as a series of basis functions A linear combination of

[0065]

[0066] Among them, θ N×1 For X N×1 In the dictionary N×N The projection vector on has only a few non-zero elements and is a sparse vector. The actual installation position of the blade tip timing system sensor can be regarded as extracting I (I < L) positions from L virtual sensors. Therefore, the measurement result of the actual blade tip timing system can be regarded as the ideal non-undersampled signal X N×1 M data points are extracted from M×1 , where Y M×1 is the column vector composed of the calculated blade vibration displacement signal. This process can be expressed as:

[0067] Y M×1 =Φ M×N X N×1

[0068] Among them, Φ M×N is the measurement matrix, which is composed of a specific row element extracted from the N×N unit matrix. The position of the extracted row element is determined by the position of the actual blade tip timing sensor installed in the virtual blade tip timing sensor. M×1 We can get a sparse representation under a certain dictionary:

[0069] Y M×1 =Φ M×N Ψ N×N θ N×1

[0070] Therefore, the blade impulse response signal can be reconstructed by solving a convex optimization problem:

[0071]

[0072] Y M×1 =Φ M×N Ψ N×N θ N×1

[0073] Find the sparse solution θ N×1 After that, substitute The non-undersampled blade impulse response signal can be obtained.

[0074] 3. Dictionary Acquisition

[0075] In the reconstruction of the blade impulse response signal, it is necessary to construct a dictionary Ψ. Different from the traditional compressed sensing algorithm, this scheme chooses to use a set of real part signals of Laplace wavelets as the basis matrix, in order to make the projection of the blade impulse response signal under the basis matrix sparse.

[0076] The coupling of general multi-order structural modes uses the following mathematical model to simulate the dynamic response signal of the structure under pulse excitation:

[0077]

[0078] The Laplace wavelet is a complex, analytical, unilaterally decaying exponential function, which is generated by taking into account the response characteristics of the impulse excitation system and is defined as follows:

[0079]

[0080] Among them, ω∈R + is the angular frequency, ζ∈[0,1], the time parameter τ∈R, and the coefficient A is an arbitrary normalization factor that normalizes the wavelet. The real part projection function of the Laplace wavelet is:

[0081]

[0082] like Figure 6 As shown in Figure 2, the real part projection of the Laplace wavelet shows the similarity between the real part projection of the Laplace wavelet and the single-mode impulse response. Further, based on the similarity between the real part projection of the Laplace wavelet and the single-mode impulse response signal, a set of real part projections of the Laplace wavelet are selected as atoms in the dictionary. For m modes, we set m sub-dictionaries Ψ i , with fixed intervals Δω and Δξ, setting the time parameter τ = 0 in all atoms. The sub-dictionary is as follows:

[0083]

[0084] The complete dictionary consisting of m sub-dictionaries is as follows:

[0085] Ψ=[Ψ1 ... Ψ m ]

[0086] 4. Implementation steps:

[0087] ①First, this method requires certain prior knowledge, namely the possible modes of blade vibration and the approximate range of each vibration mode parameter. These parameters can be obtained from the finite element calculation of the blade.

[0088] ②For the i-th mode, design the sub-dictionary Ψ i ,Ψ i The atoms in enclose the modal parameters of the mode. For example, the blade vibration mode is h(t) = e -2πωζt cos(2πωt),ω=130,ζ=0.004;

[0089] Set n = 16, k = 32, ω ψ1 =122×2π, Δω=1×2π, Δζ=0.001, then we get a sub-dictionary Ψ i , the atomic frequency range in the sub-dictionary is 122~137Hz, and the damping range is 0.001~0.032. The m sub-dictionaries are merged into the dictionary Ψ.

[0090] ③ Set the number of virtual sensors L and the number of real sensors I, and determine the measurement matrix Φ according to the positions of the real sensors.

[0091] ④ Assume that the signal sampled by I real sensors is Y, and substitute Y, Φ, Ψ into After solving, we get the optimal solution θ N×1 .

[0092] ⑤ Substitute θ into The signal X without undersampling can be obtained.

[0093] ⑥ The vibration modal parameters of the rotating blade can be obtained by using the matrix bundle method on the signal X.

[0094] It should be noted that when designing a dictionary, prior knowledge may not be available in some cases. When designing a dictionary, the frequency range and damping range covered by the atoms in the dictionary can be expanded, the number of atoms can be increased, and accurate estimation of the modal parameters can also be achieved.

[0095] According to another aspect of an embodiment of the present invention, a device for acquiring vibration modal parameters of a blade is also provided. Figure 7 is a schematic diagram of a device for acquiring blade vibration modal parameters according to an embodiment of the present invention, such as Figure 7As shown, the device for acquiring blade vibration modal parameters includes: an acquisition module 702, a construction module 704, a first determination module 706, a second determination module 708 and a processing module 710. The device for acquiring blade vibration modal parameters is described in detail below.

[0096] An acquisition module 702 is used to acquire an undersampled impulse response signal of a blade; a construction module 704 is connected to the acquisition module 702 and is used to construct a compressed sensing dictionary using a real part projection of a Laplace wavelet, wherein a similarity relationship exists between the real part projection and the undersampled impulse response signal; a first determination module 706 is connected to the construction module 704 and is used to determine a measurement matrix according to an installation position of a target sensor; a second determination module 708 is connected to the first determination module 706 and is used to determine a non-undersampled impulse response signal according to the dictionary, the measurement matrix and the undersampled impulse response signal; a processing module 710 is connected to the second determination module 708 and is used to process the non-undersampled impulse response signal using a matrix bundle method to obtain vibration modal parameters of the blade.

[0097] It should be noted that the above-mentioned modules can be implemented by software or hardware. For example, for the latter, it can be implemented in the following ways: the above-mentioned modules can be located in the same processor; and / or the above-mentioned modules are located in different processors in any combination.

[0098] In the above embodiment, the device for acquiring the vibration modal parameters of the blade can acquire the undersampled impulse response signal of the blade; construct a compressed sensing dictionary using the real part projection of the Laplace wavelet, wherein there is a similarity relationship between the real part projection and the undersampled impulse response signal; determine the measurement matrix according to the installation position of the target sensor; determine the non-undersampled impulse response signal according to the dictionary, the measurement matrix and the undersampled impulse response signal; and process the non-undersampled impulse response signal using the matrix beam method to obtain the vibration modal parameters of the blade. That is to say, the embodiment of the present invention uses the compressed sensing theory and the blade tip timing technology to obtain the impulse response signal of the blade, realizes the comprehensive estimation of the vibration modal parameters of the blade under the working condition of the impeller, and constructs the compressed sensing dictionary using the real part projection of the Laplace wavelet, improves the sparsity of the impulse response signal projected on the dictionary, thereby solving the problem of poor processing effect caused by the compressed sensing algorithm using the Fourier dictionary in the related technology to effectively improve the probability of complete signal recovery, reduce the number of sensors required, and thus improve the technical effect of accurate estimation of modal parameters.

[0099] It should be noted here that the above-mentioned acquisition module 702, construction module 704, first determination module 706, second determination module 708 and processing module 710 correspond to steps S402 to S410 in the method embodiment, and the examples and application scenarios implemented by the above-mentioned modules are the same as the corresponding steps, but are not limited to the contents disclosed in the above-mentioned method embodiment.

[0100] Optionally, the construction module 704 includes: an acquisition unit, used to acquire the number of vibration modes and initial parameters of the vibration modes of the blade; and a determination unit, used to determine a dictionary according to the number of vibration modes and the vibration mode parameters.

[0101] In the above implementation, the dictionary is set using the number of vibration modes of the blade and the initial parameters of the vibration modes, that is, the dictionary of compressed sensing theory is constructed using the real part function of the Laplace wavelet to increase the sparsity of the projection of the impulse response signal on the dictionary, thereby increasing the probability of completely recovering the signal and reducing the number of required sensors.

[0102] Optionally, the above-mentioned determination unit includes: a sub-determination unit, used to determine the same number of sub-dictionaries according to the number of vibration modes, wherein the real part projection of Laplace wavelet is used as the atom of the sub-dictionary, the frequency in the vibration mode parameter is within the frequency range of the atom, and the damping in the vibration mode parameter is within the damping range of the atom; a merging unit, used to merge the same number of sub-dictionaries into a dictionary.

[0103] Optionally, the same number of sub-dictionaries is set according to the number of vibration modes, and the real part projection of Laplace wavelet is used as the atom of the sub-dictionary to ensure that the frequency in the vibration mode parameter is within the frequency range of the atom, and the damping in the vibration mode parameter is within the damping range of the atom; and then the above sub-dictionaries are merged into a dictionary. In the above implementation, the real part projection of Laplace wavelet is used to construct the basis matrix in the compression sense to ensure the sparsity of the impulse response signal of the blade under the basis matrix.

[0104] It should be noted that when designing the dictionary, it is necessary to determine the number of blade vibration modes and the approximate range of each vibration mode parameter, and to divide the frequency domain into multiple frequency bands to add atoms, which can reduce the number of atoms and improve the calculation time.

[0105] Optionally, the second determination module 708 includes: a first calculation unit, configured to calculate an optimal sparse solution according to a dictionary, a measurement matrix and an undersampled impulse response signal; and a second calculation unit, configured to calculate a non-undersampled impulse response signal according to the optimal sparse solution.

[0106] Optionally, the dictionary, the measurement matrix, and the undersampled impulse response signal can be used as input values ​​of the optimal sparse solution solution, and the optimal sparse solution can be calculated, and the optimal sparse solution can be used as the input value of the non-undersampled impulse response signal solution to calculate the non-undersampled impulse response signal. The non-undersampled impulse response signal can be accurately calculated in the above manner.

[0107] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is further provided. The computer-readable storage medium includes a stored program, wherein the program executes any of the above methods when it is run.

[0108] In this embodiment, the computer-readable storage medium may be located in any computer terminal in a computer terminal group in a computer network, and / or located in any mobile terminal in a mobile terminal group, and the computer-readable storage medium includes a stored program.

[0109] Optionally, when the program is running, the device where the computer-readable storage medium is located is controlled to perform the following functions: obtaining an undersampled impulse response signal of the blade; constructing a compressed sensing dictionary using the real part projection of the Laplace wavelet, wherein there is a similarity relationship between the real part projection and the undersampled impulse response signal; determining a measurement matrix according to the installation position of the target sensor; determining a non-undersampled impulse response signal according to the dictionary, the measurement matrix and the undersampled impulse response signal; and processing the non-undersampled impulse response signal using a matrix bundle method to obtain vibration modal parameters of the blade.

[0110] Optionally, constructing a compressed sensing dictionary includes: obtaining the number of vibration modes and initial vibration mode parameters of the blade; and determining a dictionary according to the number of vibration modes and the vibration mode parameters.

[0111] Optionally, a dictionary is determined based on the number of vibration modes and vibration modal parameters, including: determining the same number of sub-dictionaries based on the number of vibration modes, wherein real part projection of Laplace wavelet is used as atoms of the sub-dictionary, the frequency in the vibration modal parameters is within the frequency range of the atoms, and the damping in the vibration modal parameters is within the damping range of the atoms; and merging the same number of sub-dictionaries into a dictionary.

[0112] Optionally, determining a non-undersampled impulse response signal according to a dictionary, a measurement matrix and an undersampled impulse response signal includes: calculating an optimal sparse solution according to the dictionary, the measurement matrix and the undersampled impulse response signal; and calculating a non-undersampled impulse response signal according to the optimal sparse solution.

[0113] According to another aspect of an embodiment of the present invention, there is provided an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to execute any one of the above methods through the computer program.

[0114] Optionally, an embodiment of the present invention provides a device, which includes a processor, a memory, and a program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the program: obtaining an undersampled pulse response signal of a blade; constructing a compressed sensing dictionary using the real part projection of a Laplace wavelet, wherein a similarity relationship exists between the real part projection and the undersampled pulse response signal; determining a measurement matrix according to the installation position of the target sensor; determining a non-undersampled pulse response signal according to the dictionary, the measurement matrix, and the undersampled pulse response signal; and processing the non-undersampled pulse response signal using a matrix bundle method to obtain vibration modal parameters of the blade.

[0115] Optionally, constructing a compressed sensing dictionary includes: obtaining the number of vibration modes and initial vibration mode parameters of the blade; and determining a dictionary according to the number of vibration modes and the vibration mode parameters.

[0116] Optionally, a dictionary is determined based on the number of vibration modes and vibration modal parameters, including: determining the same number of sub-dictionaries based on the number of vibration modes, wherein real part projection of Laplace wavelet is used as atoms of the sub-dictionary, the frequency in the vibration modal parameters is within the frequency range of the atoms, and the damping in the vibration modal parameters is within the damping range of the atoms; and merging the same number of sub-dictionaries into a dictionary.

[0117] Optionally, determining a non-undersampled impulse response signal according to a dictionary, a measurement matrix and an undersampled impulse response signal includes: calculating an optimal sparse solution according to the dictionary, the measurement matrix and the undersampled impulse response signal; and calculating a non-undersampled impulse response signal according to the optimal sparse solution.

[0118] The above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention.

Claims

1. A method for obtaining blade vibration modal parameters, characterized in that: The impulse response signal of the blade is obtained by using compressed sensing theory and blade tip timing counting, and the vibration modal parameters of the blade are estimated under the working conditions of the impeller, including: Acquire an under-sampled impulse response signal of the blade; A compressed sensing dictionary is constructed by using the real part projection of the Laplace wavelet, wherein there is a similarity relationship between the real part projection and the under-sampled impulse response signal; the compressed sensing dictionary is constructed, comprising: Obtaining the number of vibration modes and initial parameters of the vibration modes of the blade; Determining the dictionary according to the number of vibration modes and the vibration mode parameters; The determining of the dictionary comprises: Determine the same number of sub-dictionaries according to the number of vibration modes, wherein the real part projection of Laplace wavelet is used as the atom of the sub-dictionary, the frequency in the vibration mode parameter is within the frequency range of the atom, and the damping in the vibration mode parameter is within the damping range of the atom; Merging the same number of sub-dictionaries into the dictionary; Determine the measurement matrix according to the installation position of the target sensor; Determining a non-undersampled impulse response signal according to the dictionary, the measurement matrix and the undersampled impulse response signal; comprising: Calculating an optimal sparse solution according to the dictionary, the measurement matrix and the under-sampled impulse response signal; Calculating the non-undersampled impulse response signal according to the optimal sparse solution; The non-undersampled impulse response signal is processed using a matrix bundle method to obtain vibration modal parameters of the blade.

2. A device for acquiring blade vibration modal parameters, characterized in that: The impulse response signal of the blade is obtained by using compressed sensing theory and blade tip timing counting, and the vibration modal parameters of the blade are estimated under the working conditions of the impeller, including: An acquisition module, used for acquiring an under-sampled impulse response signal of a blade; A construction module is used to construct a compressed sensing dictionary using a real part projection of a Laplace wavelet, wherein there is a similarity relationship between the real part projection and the under-sampled impulse response signal; the construction module comprises: An acquisition unit, used for acquiring the number of vibration modes and initial parameters of the vibration modes of the blade; A determination unit, configured to determine the dictionary according to the number of vibration modes and the vibration mode parameters; the determination unit comprises: A determination subunit is used to determine the same number of sub-dictionaries according to the number of vibration modes, wherein Laplace wavelet real part projection is used as the atom of the sub-dictionary, the frequency in the vibration mode parameter is within the frequency range of the atom, and the damping in the vibration mode parameter is within the damping range of the atom; a merging unit, used for merging the sub-dictionaries of the same number into the dictionary; A first determination module, used to determine a measurement matrix according to an installation position of a target sensor; A second determination module is configured to determine a non-undersampled impulse response signal according to the dictionary, the measurement matrix and the undersampled impulse response signal; the second determination module includes: A first calculation unit, configured to calculate an optimal sparse solution according to the dictionary, the measurement matrix and the under-sampled impulse response signal; A second calculation unit, configured to calculate the non-undersampled impulse response signal according to the optimal sparse solution; The processing module is used to process the non-undersampled impulse response signal using a matrix bundle method to obtain the vibration modal parameters of the blade.

3. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein the program executes the method of claim 1 when executed.

4. An electronic device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to execute the method according to claim 1 through the computer program.

Images