A two-level consistency control method for piezoelectric sandwich performance

Abstract

The present invention discloses a two-level consistency control method for the performance of a piezoelectric interlayer, including: obtaining the impedance curve of piezoelectric elements produced in the same batch, extracting the impedance values ​​of the piezoelectric elements at multiple typical frequencies and constructing a multidimensional feature vector, eliminating piezoelectric elements with large impedance performance deviations according to the similarity distance of the multidimensional feature vectors of different piezoelectric elements, and realizing the first-level consistency control of the performance of the piezoelectric elements; preparing batch piezoelectric interlayers based on the piezoelectric elements screened after the first-level consistency control, obtaining the impedance curve of the piezoelectric elements on the piezoelectric interlayer, and combining the consistency threshold index to perform a second-level consistency control on the multidimensional feature vectors of the piezoelectric elements at the same position in the batch piezoelectric interlayer, and further screening batch piezoelectric interlayers with high consistency. The present invention greatly reduces the dispersion of the guided wave monitoring performance of batch piezoelectric interlayers in structural health monitoring applications, effectively improves the accuracy of quantitative diagnosis, and has a good prospect in the engineering application of single-machine structural health monitoring.

Description

Technical Field

[0001] The invention belongs to the technical field of aviation structure health monitoring, and in particular relates to a two-level consistency control method for piezoelectric interlayer performance. Background Art

[0002] Aviation structural health monitoring technology can monitor the health status of aviation structures online, and then predict and estimate structural damage and remaining life, so as to achieve the purpose of ensuring the safety of aircraft structures and reducing structural maintenance costs. Among the many structural health monitoring technologies, structural health monitoring technology based on guided wave has the advantages of long monitoring distance, large monitoring area, regional monitoring, and sensitivity to small structural damage. In the process of structural status monitoring, the quantitative damage diagnosis results can be used as an important basis for the evaluation and prediction of the remaining life of a single aircraft structure. Therefore, it is of great significance to improve the accuracy of quantitative damage diagnosis technology. Usually, the key step of quantitative damage diagnosis is to use batch structural parts as training parts, calibrate the quantitative relationship between signal characteristics and damage degree, and then apply the calibration model to the monitored structural parts. However, the dispersion of the performance of piezoelectric sensor networks will directly lead to the dispersion of the output guided wave signal characteristics, and ultimately reduce the accuracy of quantitative damage diagnosis based on guided wave monitoring technology. Therefore, the consistency technology of controlling the performance of piezoelectric sensor networks is crucial to improving the accuracy of quantitative damage diagnosis in the single-machine monitoring application of aviation structures.

[0003] However, current technology only targets the preliminary performance control of a single piezoelectric sensor, and there is no engineering application for quantitative damage diagnosis, and it cannot be used to improve the accuracy of quantitative damage diagnosis in single-machine monitoring applications. In recent years, piezoelectric interlayer technology has been gradually applied to aircraft structural health monitoring technology, and the effectiveness of piezoelectric interlayer technology in improving the consistency of sensor network performance has been verified. However, in current engineering applications, piezoelectric interlayer technology only ensures the consistency of piezoelectric sensor performance within a network composed of a single interlayer, and there is still a lack of batch piezoelectric interlayer consistency control technology for quantitative damage diagnosis. In summary, the development of batch piezoelectric interlayer performance consistency control technology is of great significance to improving the accuracy of quantitative damage diagnosis in single-machine monitoring applications. Summary of the invention

[0004] In view of the above-mentioned deficiencies of the prior art, the purpose of the present invention is to provide a two-level consistency control method for piezoelectric interlayer performance to solve the problem in the prior art that the piezoelectric interlayer performance has large dispersion, resulting in low accuracy in quantitative diagnosis of structural damage.

[0005] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0006] A two-level consistency control method for piezoelectric interlayer performance of the present invention comprises the following steps:

[0007] (1) Obtaining impedance curves of piezoelectric elements produced in the same batch, extracting impedance values ​​of the piezoelectric elements at multiple typical frequencies and constructing multidimensional feature vectors, eliminating piezoelectric elements with large impedance performance deviations based on the similarity distance of the multidimensional feature vectors of different piezoelectric elements, and achieving first-level consistency control of the performance of the piezoelectric elements;

[0008] (2) Based on the piezoelectric elements screened by the first-level consistency control, batch piezoelectric interlayers are prepared, and the impedance curves of the piezoelectric elements on the piezoelectric interlayers are obtained. The multi-dimensional feature vectors of the piezoelectric elements at the same position in the batch piezoelectric interlayers are subjected to second-level consistency control combined with the consistency threshold index, and batch piezoelectric interlayers with high consistency are further screened.

[0009] Furthermore, the step (1) specifically includes:

[0010] (11) Obtaining impedance curves of piezoelectric elements produced in the same batch, and deleting piezoelectric elements with distorted impedance curves;

[0011] (12) For the i-th piezoelectric element, extract K typical frequencies (f1, f2, …, f k ,…,f K ) under the impedance value, construct the multidimensional feature vector MFV i :

[0012]

[0013] In the formula, is the typical frequency f of the i-th piezoelectric element k The impedance value under

[0014] (13) Measuring the multidimensional feature vector MFV of the i-th piezoelectric element i Similarity distance D from the origin i ; Press D i In descending order, after sorting, the piezoelectric elements with similar distances are divided into one category, and the number of categories is determined according to the number of piezoelectric elements in the piezoelectric interlayer;

[0015] (14) Calculate the dispersion of the impedance performance of the piezoelectric element in each category, typical frequency f k The dispersion of the type C piezoelectric element is defined as follows:

[0016]

[0017] Among them, Z(c) max and Z(c) min Respectively represent the typical frequency f k The maximum and minimum impedance values ​​of the piezoelectric element of type c, k = 1, 2, 3, ..., K, mean(Z(c)) represents the typical frequency fk The average impedance of the piezoelectric elements of class c;

[0018] (15) Set the consistency enhancement threshold control index TH. If the dispersion E of the performance parameters of the piezoelectric elements of class c c > TH, then eliminate the piezoelectric element with the largest relative deviation in performance among the piezoelectric elements of class c, and repeat this step until E c < TH, so as to select piezoelectric elements with good consistency and finally achieve the consistency control of the performance of piezoelectric elements.

[0019] Furthermore, the step (2) specifically includes:

[0020] (21) Prepare a batch of piezoelectric laminates based on the piezoelectric elements screened by the first-level consistency control. Weld the piezoelectric elements of the same category after the first-level consistency control at the same position of the batch of piezoelectric laminates, and use the piezoelectric elements of the same category for welding at each position separately;

[0021] (22) Obtain the impedance curves of the piezoelectric elements at all positions on the piezoelectric laminate, and separately extract the impedance values of the piezoelectric elements at the same position of the batch of piezoelectric laminates at K typical frequencies (f1, f2,..., f k ,…, f K ) to construct a multi-dimensional feature vector;

[0022] (23) Calculate the dispersion of the impedance performance of the piezoelectric elements at position p of the batch of piezoelectric laminates, 0 < p ≤ P; the dispersion of the piezoelectric elements at position p at the typical frequency f k is defined as follows:

[0023]

[0024] where Z(p) max and Z(p) min respectively represent the maximum and minimum values of the impedance of the piezoelectric element at position p at the typical frequency f k , mean(Z(p)) represents the average impedance of the piezoelectric element at position p at the typical frequency f k , k = 1, 2, 3,..., K;

[0025] (24) If the dispersion E of the piezoelectric elements at position p of the batch of piezoelectric laminates p > TH, then eliminate the piezoelectric laminate with the largest relative deviation in impedance value at position p, and repeat this step until E p < TH, so as to select piezoelectric laminates with good consistency and finally achieve the second-level consistency control of the performance of piezoelectric laminates.

[0026] The beneficial effects of the present invention:

[0027] The present invention can significantly reduce the dispersion of batch piezoelectric interlayer waveguide monitoring performance in structural health monitoring applications, effectively improve the accuracy of quantitative diagnosis, and has good prospects in the engineering application of single-machine structural health monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 The figure is a flow chart of the method of the present invention.

[0029] Figure 2 Schematic diagram of the arrangement of the monitored structure and the piezoelectric sensor network in the embodiment.

[0030] Figure 3a Schematic diagram of impedance curve test results of all piezoelectric elements.

[0031] Figure 3b This is a schematic diagram of the impedance curve test results of batch piezoelectric components after eliminating the impedance curve distortion.

[0032] Figure 4 Schematic diagram of the multi-dimensional feature vector results constructed by extracting impedance values ​​at multiple typical frequencies from the same batch of piezoelectric components.

[0033] Figure 5 This is a schematic diagram of the piezoelectric element grouping results after the first level of control.

[0034] Figure 6 This is a schematic diagram of the impedance dispersion results of piezoelectric elements at different positions in the batch piezoelectric interlayer after the second-level control.

[0035] Figure 7 Schematic diagram of the establishment results of the calibration model for the relationship between damage factor and crack length.

[0036] Figure 8a It is a schematic diagram of the damage quantitative diagnosis result after implementing the method of the present invention.

[0037] Figure 8b It is a schematic diagram of the quantitative diagnosis result of damage after the method of the present invention is not implemented. DETAILED DESCRIPTION

[0038] In order to facilitate the understanding of those skilled in the art, the present invention is further described below in conjunction with embodiments and drawings. The contents mentioned in the implementation modes are not intended to limit the present invention.

[0039] Figure 2 The diagram is a schematic diagram of the monitored structure and the arrangement of the piezoelectric sensor network in the embodiment. The monitored structure is a typical aviation ear structure, made of LY12 aluminum alloy with a thickness of 5 mm. According to the characteristics of the structure and damage initiation and expansion, a sensor network consisting of 5 piezoelectric sensors is arranged on the structure to perform quantitative damage diagnosis.

[0040] Reference Figure 1 As shown, a two-level consistency control method for piezoelectric interlayer performance of the present invention comprises the following steps:

[0041] (1) Obtaining impedance curves of piezoelectric elements produced in the same batch, extracting impedance values ​​of the piezoelectric elements at multiple typical frequencies and constructing multidimensional feature vectors, eliminating piezoelectric elements with large impedance performance deviations based on the similarity distance of the multidimensional feature vectors of different piezoelectric elements, and achieving first-level consistency control of the performance of the piezoelectric elements;

[0042] (11) obtaining impedance curves of all piezoelectric elements manufactured from the same batch, and then deleting piezoelectric elements with distorted impedance curves;

[0043] In this embodiment, an impedance analyzer is used to measure the impedance curves of 100 piezoelectric elements produced in the same batch. Figure 3a Impedance curve test results for all piezoelectric elements. Figure 3b This is the impedance curve test result of batch piezoelectric components after removing the impedance curve distortion.

[0044] (12) For the i-th piezoelectric element, extract K typical frequencies (f1, f2, …, f k ,…,f K ) under the impedance value, construct the multidimensional feature vector MFV i :

[0045]

[0046] In the formula, is the typical frequency f of the i-th piezoelectric element k The impedance value under

[0047] In this embodiment, the impedance values ​​at the typical frequency of 70kHz and the typical frequency of 170kHz are extracted, that is, K=2; Figure 4 The impedance values ​​at multiple typical frequencies extracted from the same batch of piezoelectric elements are used to construct a multi-dimensional feature vector result; wherein the horizontal axis MFV(1) represents the impedance value at a typical frequency of 70kHz, and the vertical axis MFV(2) represents the impedance value at a typical frequency of 170kHz.

[0048] (13) Calculate the multidimensional feature vector MFV of the i-th piezoelectric element i Similarity distance D from the origin i In this embodiment, the Euclidean distance commonly used in engineering is used to measure similarity, and its calculation formula is as follows:

[0049]

[0050] Then, press D iSort in descending order. After sorting, group piezoelectric elements with similar distances into one category, and determine the number of piezoelectric elements in each category according to the number of batch piezoelectric interlayers. The number of categories is determined according to the number of piezoelectric elements in the piezoelectric interlayer.

[0051] (14) Calculate the dispersion of the impedance performance of piezoelectric elements in each category; for the typical frequency f k (k = 1, 2, 3,..., K), the dispersion of piezoelectric elements in category c is defined as follows:

[0052]

[0053] where Z(c) max and Z(c) min represent the maximum and minimum values of the impedance of piezoelectric elements in category c at the typical frequency f k respectively, and mean(Z(c)) represents the average impedance of piezoelectric elements in category c at the typical frequency f k .

[0054] (15) Set the consistency enhancement threshold control index TH = 5%. If the dispersion Ec of the performance parameters of the piezoelectric elements in category c > TH, then remove the piezoelectric element with the largest relative deviation in performance in category c, and repeat this step until Ec < TH is satisfied to select piezoelectric interlayers with good consistency, and finally achieve the consistency control of the performance of piezoelectric elements;

[0055] In this embodiment, finally, 6 batch piezoelectric interlayers need to be fabricated, and each piezoelectric interlayer has 5 sensor positions. To ensure the consistency of the performance of sensors at the same position on different piezoelectric interlayers, finally, according to the distance of the Euclidean distance, the piezoelectric elements are divided into 5 groups, with at least 10 in each group; Figure 5 is the grouping result of piezoelectric elements after the first - level control, where the scatter points of the same shape represent one group, and the impedance dispersion of piezoelectric elements in each group is controlled within 5%.

[0056] (2) Based on the piezoelectric elements screened by the first - level consistency control, fabricate batch piezoelectric interlayers, obtain the impedance curves of the piezoelectric elements on the piezoelectric interlayers, and perform a second - level consistency control on the multi - dimensional feature vectors of the piezoelectric elements at the same position in the batch piezoelectric interlayers in combination with the consistency threshold index to further screen batch piezoelectric interlayers with high consistency.

[0057] (21) Based on the piezoelectric elements screened by the first - level consistency control, fabricate batch piezoelectric interlayers, weld the piezoelectric elements of the same category after the first - level consistency control at the same position on the batch piezoelectric interlayers, and use piezoelectric elements of the same category for welding at each position separately;

[0058] (22) Obtain the impedance curves of piezoelectric elements at all positions on the piezoelectric sandwich layer, and respectively extract the impedance values of piezoelectric elements at the same position of a batch of piezoelectric sandwich layers at K typical frequencies (f1, f2, …, f k , …, f K ) to construct a multi-dimensional feature vector;

[0059] (23) Calculate the dispersion of the impedance performance of piezoelectric elements at position p of a batch of piezoelectric sandwich layers, where 0 < p ≤ P; the dispersion of piezoelectric elements at position p at the typical frequency f k is defined as follows:

[0060]

[0061] where Z(p) max and Z(p) min respectively represent the maximum and minimum values of the impedance of the piezoelectric element at position p at the typical frequency f k , mean(Z(p)) represents the mean impedance of the piezoelectric element at position p at the typical frequency f k , and k = 1, 2, 3, …, K;

[0062] (24) If the performance dispersion E p of the piezoelectric elements at position p of a batch of piezoelectric sandwich layers is greater than TH, then eliminate the piezoelectric sandwich layer with the largest relative deviation of impedance value at position p, and repeat this step until E p < TH, so as to select piezoelectric sandwich layers with good consistency, and finally achieve the second-level consistency control of the performance of piezoelectric sandwich layers. Figure 6 is a schematic diagram of the impedance dispersion results of piezoelectric elements at different positions of a batch of piezoelectric sandwich layers after the second-level control, and the impedance dispersion at the typical frequency is controlled within 5%.

[0063] Fix a batch of piezoelectric sandwich layers after implementing the two-level consistency control method proposed by the present invention on a batch of lug structural parts, carry out a fatigue test to cause the crack length of the structure to expand, use a microscope to record the real crack length data, and at the same time obtain the guided wave signals of typical excitation-sensing channels 1-3 and extract the corresponding damage factors. The obtained training data set is as Figure 7 shown; construct a calibration model between the crack length and the damage factor by using the polynomial shown in the following formula, and the parameters of the polynomial model are obtained by the least square method;

[0064] a = f(DI) = β0 + β1·DI + β2·(DI) 2 +... + β p ·(DI) p

[0065] Where a represents the crack length, f(.) represents the trained calibration model, DI represents the obtained damage factor, β0,β1,β2,...,β p is the polynomial fitting parameter, p is the order of the polynomial, and the establishment results of the calibration model of the relationship between damage factor and crack length are as follows: Figure 7 shown.

[0066] During the crack propagation process of the monitored test piece, obtain the current monitoring time T n The guided wave signal is extracted and the corresponding damage factor is recorded as GWF(T n ); when GWF(T n ) is input into the calibration model, the crack length can be quantitatively output:

[0067] d(T n )=f(GWF(T n ))

[0068] Among them, d(T n ) represents the current monitoring time T n The crack length of the monitoring structure; f(.) represents the trained calibration model; Figure 8a The damage quantitative diagnosis result after implementing the method of the present invention, where the horizontal axis is the number of load cycles and the vertical axis is the crack length. It can be seen that the crack quantitative diagnosis result of the monitored test piece is close to the actual crack length. The absolute error range of crack monitoring is ±1.1mm, and the root mean square error is 0.8mm. Figure 8b Compared with the quantitative diagnosis results without implementing the method of the present invention, the method of the present invention can effectively improve the accuracy of quantitative damage assessment, and the accuracy of quantitative damage assessment is improved by 38%.

[0069] The present invention has many specific application paths. The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements can be made without departing from the principle of the present invention. These improvements should also be regarded as the protection scope of the present invention.

Claims

1. A two-level consistency control method for piezoelectric interlayer performance, characterized in that: Here are the steps: (1) Obtaining impedance curves of piezoelectric elements produced in the same batch, extracting impedance values ​​of the piezoelectric elements at multiple typical frequencies and constructing multidimensional feature vectors, eliminating piezoelectric elements with large impedance performance deviations based on the similarity distance of the multidimensional feature vectors of different piezoelectric elements, and achieving first-level consistency control of the performance of the piezoelectric elements; (2) Preparing batch piezoelectric interlayers based on the piezoelectric elements screened by the first-level consistency control, obtaining the impedance curve of the piezoelectric elements on the piezoelectric interlayer, and performing second-level consistency control on the multi-dimensional feature vectors of the piezoelectric elements at the same position in the batch piezoelectric interlayer in combination with the consistency threshold index, and further screening batch piezoelectric interlayers with high consistency; The step (1) specifically comprises: (11) Obtaining impedance curves of piezoelectric elements produced in the same batch, and deleting piezoelectric elements with distorted impedance curves; (12) For the i-th piezoelectric element, extract the impedance values ​​at K typical frequencies and construct a multidimensional feature vector MFV i : In the formula, is the typical frequency f of the i-th piezoelectric element k The impedance value under (13) Measuring the multidimensional feature vector MFV of the i-th piezoelectric element i Similarity distance D from the origin i ; Press D i In descending order, after sorting, the piezoelectric elements with similar distances are divided into one category, and the number of categories is determined according to the number of piezoelectric elements in the piezoelectric interlayer; (14) Calculate the dispersion of the impedance performance of the piezoelectric element in each category, typical frequency f k The dispersion of the type C piezoelectric element is defined as follows: Among them, Z(c) max and Z(c) min Respectively represent the typical frequency f k The maximum and minimum impedance values ​​of the piezoelectric element of type c, k = 1, 2, 3, ..., K, mean(Z(c)) represents the typical frequency f k The impedance mean of the piezoelectric element of type c below; (15) Set the consistency enhancement threshold control index TH. If the dispersion E of the performance parameters of the piezoelectric elements of the c-th category c > TH, then eliminate the piezoelectric element with the largest relative deviation in performance among the piezoelectric elements of the c-th category, and repeat this step until E c < TH to select piezoelectric elements with good consistency, and finally achieve the consistency control of the performance of piezoelectric elements; The step (2) specifically comprises: (21) preparing a batch of piezoelectric interlayers based on the piezoelectric elements screened by the first-level consistency control, welding the piezoelectric elements of the same category after the first-level consistency control at the same position of the batch of piezoelectric interlayers, and welding each position with a piezoelectric element of the same category; (22) obtaining impedance curves of piezoelectric elements at all positions on the piezoelectric interlayer, respectively extracting impedance values ​​of piezoelectric elements at the same position of a batch of piezoelectric interlayers at K typical frequencies to construct a multidimensional feature vector; (23) Calculate the dispersion of the impedance performance of the piezoelectric elements at the piezoelectric sandwich position p, where 0 < p ≤ P; at the typical frequency f k The dispersion of the piezoelectric elements at position p is defined as follows: Among them, Z(p) max and Z(p) min Respectively represent the typical frequency f k The maximum and minimum values ​​of the impedance of the piezoelectric element at position p, mean(Z(p)) represents the typical frequency f k The average impedance of the piezoelectric element at the lower position p, k = 1, 2, 3, ..., K; (24) If the performance dispersion E of the piezoelectric elements at the batch piezoelectric sandwich position p p > TH, then eliminate the piezoelectric sandwich with the largest relative deviation of impedance value at position p, and repeat this step until E p < TH, so as to select piezoelectric sandwiches with good consistency and finally achieve the second-level consistency control of the performance of piezoelectric sandwiches.

Images