A full-field localization method for far- and near-field wall panel leakage sources based on linear acoustic array

Abstract

The present invention proposes a full-field positioning method for far- and near-field wall panel leakage sources based on a linear acoustic array, which belongs to the field of high-precision positioning. It solves the problems that the positioning solution for the leakage source of the wall panel structure requires a large number of sensor arrays, which is inconvenient to use in practice, and the amount of calculation required to ensure the accuracy of the positioning algorithm is large, which is not conducive to real-time monitoring. The method comprises the following steps: step 1: determining a dispersion curve according to the structure and material of the wall panel to be detected, extracting the main frequency of the signal, and determining an angle threshold according to the array structure and two selected sub-arrays; step 2: calculating the difference in the positioning results of the two sub-arrays, and comparing the positioning result difference with the threshold; step 3: if the positioning result difference is not greater than the threshold, far-field positioning is performed, and the far-field positioning is completed using the full array and the far-field positioning result is output; step 4: if the positioning result difference is less than the threshold, near-field positioning is performed, and the full array signal is processed using an improved 2Dmusic algorithm to obtain a near-field positioning result.

Description

Technical Field

[0001] The invention relates to a full-field positioning method for far-field and near-field wallboard leakage sources based on a linear acoustic array, and belongs to the field of high-precision positioning. Background Art

[0002] In recent years, space technology has advanced rapidly, extending the service lives of large spacecraft such as manned spacecraft and space stations. At the same time, the amount of space debris in low-Earth orbit is also rapidly increasing, increasing the risk of impacts and leaks on in-orbit spacecraft. Leaks in a spacecraft's outer structure can seriously destabilize the internal environment, threatening the safety of astronauts and posing significant risks to the operation of large-scale facilities such as space stations. Timely and effective detection of leak sources facilitates rapid repairs to the leaking area, ensuring the safety of spacecraft and astronauts.

[0003] Therefore, research has focused on locating leak sources in wall panel structures. Qi Lei et al. proposed using beamforming energy focusing to achieve localization. This method employs a far-field narrowband model, delaying and superimposing signals from various channels at different angles. The superimposed image is then searched for spectral peaks to determine the azimuth of the leak source.

[0004] In the technical solution proposed by Qi Lei et al., the accuracy of impact positioning is affected by the location of the leak source. When the leak source is close to the sensor array, the far-field condition is not met, resulting in a large positioning error. In their invention patent CN113405736A, Xin Fengxian et al. disclosed a "leak source positioning method and system based on acoustic emission characteristics of gas leaks in plate-like structures." This method uses a cross-shaped sensor array, pre-processes the signal, and then uses a surround sensor music algorithm for orientation. The orientation results of multiple groups of ring sensors are then used to obtain numerous intersection points, which are then processed using a K-means clustering algorithm to achieve high-precision positioning. In the technical solution proposed by Xin Fengxian et al., the influence of dispersion on leak source positioning is well addressed, achieving high-precision positioning. However, this solution requires a large number of sensor arrays, making it inconvenient to use in practice. Furthermore, the computational effort required to ensure the accuracy of the positioning algorithm is high, making it unsuitable for real-time monitoring. Summary of the Invention

[0005] To address the problems that the solution for locating the leakage source of a wall panel structure requires a large number of sensor arrays, which is inconvenient to use in practice, and the large amount of calculation required to ensure the accuracy of the positioning algorithm, which is not conducive to real-time monitoring, the present invention proposes a full-field positioning method for the near- and far-field wall panel leakage source based on a linear acoustic array, which specifically includes:

[0006] Step 1: Determine the dispersion curve based on the structure and material of the panel being tested, extract the main frequency of the signal, and determine the angle threshold based on the array structure and the two selected sub-arrays;

[0007] Step 2: Calculate the difference in positioning results of the two sub-arrays and compare the positioning result difference with the threshold;

[0008] Step 3: If the positioning result difference is not greater than the threshold, perform far-field positioning, use the full array to complete far-field positioning and output the far-field positioning result;

[0009] Step 4: If the positioning result difference is less than the threshold, near-field positioning is performed, and the improved 2Dmusic algorithm is used to process the full array signal to obtain the near-field positioning result.

[0010] Preferably, step 1 specifically includes:

[0011] Step 1.1: After receiving the acoustic emission signal from the leakage source, the linear acoustic array performs a Fourier transform on the measurement signal of the reference array element to identify the energy of the acoustic emission signal from the leakage source in different frequency bands. The acoustic emission signal from the leakage source is a broadband signal of (50-400) kHz.

[0012] Step 1.2: Use bandpass filtering to extract the 20 kHz bandwidth signal with the highest energy as the measurement signal, where the center frequency of the measurement signal is w and the propagation speed is a constant v;

[0013] Step 1.3: Calculate the delay function of the i-th element of the linear array with an element spacing of d relative to the reference element under far-field conditions;

[0014] Step 1.4: Multiply the received signals of all channels of the linear acoustic array by the conjugate of the delay coefficient and sum them to obtain the signal strength received in all directions. The direction with the maximum received signal energy is used as the direction and angle threshold of the leakage source.

[0015] The expression of the delay function is:

[0016]

[0017] Preferably, the threshold calculation step in step 2 includes:

[0018] Step 2.1: Calculate the output of the subarray in the θ direction;

[0019] Step 2.2: Calculate the different energy and angle correspondences based on the output of the sub-array in the θ direction, and obtain the positioning results of the two sub-arrays;

[0020] Step 2.3: Calculate the positioning result difference of the two sub-arrays under the theoretical critical case when the far-field condition is met, and use the positioning result difference of the two sub-arrays under the theoretical critical case as the threshold. The far-field condition is that the distance between the leakage source and the reference array element is greater than

[0021]

[0022] The expression of the subarray output in the θ direction is:

[0023]

[0024] In formula (2), M is the number of sensor array elements;

[0025] The expression of the difference in positioning results between the two sub-arrays in the theoretical critical case is:

[0026]

[0027] In formula (3), x is the distance between the two sub-array reference elements, and D is the aperture of the entire sensor array.

[0028] Preferably, step 3 specifically includes:

[0029] When the far-field condition is met, the angle threshold of the full array is determined according to step 1, and the positioning result of the full array is obtained according to step 2, and the positioning result of the full array is output as the far-field positioning result.

[0030] Preferably, step 4 specifically includes:

[0031] Step 4.1: Use the improved 2Dmusic algorithm to convert the variables θ and r into γ and φ, calculate the delay function matrix under near-field conditions, and decompose the delay function matrix Ω(θ,r) to obtain the noise subspace Ω1(γ)Ω2(φ);

[0032] Step 4.2: Calculate the music power spectrum P(θ,r) based on the noise subspace;

[0033] Step 4.3: Use the power spectrum function to calculate the music power spectrum within the search range of the linear acoustic array. Introduce the matrix e1 = (0, 0…0, 1) to transform the maximum value problem of P(θ, r) into a quadratic optimization problem in standard form.

[0034] Step 4.4: Use the Lagrange multiplier method to construct the cost function to solve the quadratic optimization problem;

[0035] Step 4.5: Adjust the cost function Find the partial derivative equal to zero to obtain the minimum value of the cost function, and Substitution Eliminate the unknown parameter η in the cost function in the partial derivative formula to obtain Ω2(φ);

[0036] Step 4.6: Substitute Ω2(φ) into the quadratic optimization problem and transform the maximum value problem of P(θ,r) into a maximum value problem of the single variable γ;

[0037] Step 4.7: Use the calculated music power spectrum P(θ, r) as a known condition and use it to search for spectrum peaks for different γ values ​​to obtain the value of γ. Repeat steps 4.1 to 4.6 to obtain the value of φ. The values ​​of θ and r are used as the position estimation results.

[0038] Step 4.8: Use the power spectrum function to perform a two-dimensional spectrum peak search within a limited range of ±10° and ±0.05m from the estimated position, and use the angle and distance corresponding to the maximum value as the near-field positioning result;

[0039] The expression of music power spectrum is:

[0040]

[0041] In formula (4), x(t) is the measurement signal matrix, * represents the conjugate transpose of the matrix, K is the number of snapshots, and U is S is the signal subspace composed of the largest eigenvalue, U N is the noise subspace composed of the remaining eigenvalues;

[0042] The expression of the power spectrum function is:

[0043]

[0044] In formula (5),

[0045] The expression of the quadratic optimization problem is:

[0046]

[0047] The expression of the cost function is:

[0048] L(γ,φ)=Ω2 * (φ)Q(γ)Ω2(φ)-η(e1 T Ω2(φ)-1) (7);

[0049] In formula (7), η is an unknown parameter;

[0050] right The expression for finding the partial derivative equal to zero is:

[0051]

[0052] The calculation formula for Ω2(φ) is:

[0053]

[0054] The calculation formula of the noise subspace Ω1(γ)Ω2(φ) is:

[0055]

[0056] Preferably, the calculation step of the delay function matrix under near-field conditions in step 4.1 includes:

[0057] Step 4.1.1: Calculate the difference between the distance from the received signal to the reference array element and the distance from the signal to any sub-array element;

[0058] Step 4.1.2: Using the center array element as the reference element, calculate the delay function matrix under near-field conditions based on the distance difference;

[0059] The expression for the difference between the distance from the received signal to the reference array element and the distance from the signal to any array element is:

[0060]

[0061] In formula (11), n ​​is the number of intervals between the array element and the reference array element;

[0062] The expression of the delay function matrix under near-field conditions is:

[0063]

[0064] In formula (12), is the number of sensor elements on a single side, d is the interval between adjacent elements, is a small first-order quantity in Δ, is a small second-order quantity in Δ.

[0065] The beneficial effects of the present invention are:

[0066] (1) The array structure used in the present invention is simple, and a single linear sensor array can achieve the function, which is easy to arrange and use

[0067] (2) The present invention can expand the positioning range. By analyzing the sub-array positioning results, the range of the leakage source can be preliminarily determined, and a corresponding positioning method can be designed for the near-field situation, so that the system can maintain a high positioning accuracy under both far-field and near-field conditions.

[0068] (3) The algorithm used in the present invention has low complexity. An improved 2Dmusic algorithm is used in near-field positioning, which converts the two-dimensional search into two one-dimensional searches while ensuring accuracy. At the same time, there is no need to process multiple sensor array signals, which improves the real-time performance of the algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 A schematic flow chart of a method for full-field localization of near- and far-field wall panel leakage sources based on a linear acoustic array provided by the present invention;

[0070] Figure 2The velocity-frequency dispersion curve of the aluminum alloy wall panel with a thickness of 2.5 mm provided by the present invention;

[0071] Figure 3 A schematic diagram of the position arrangement of the linear sensor array provided by the present invention;

[0072] Figure 4 Schematic diagram of the estimated signal source and 2D search range provided by the present invention;

[0073] Figure 5 This is a schematic diagram of the angle and distance information of the near-field positioning results provided by the present invention. DETAILED DESCRIPTION

[0074] Combine Figure 1-5 This embodiment is described as follows. Figure 1 As shown, the steps of the full-field positioning method for far-field and near-field wall panel leakage sources based on a linear acoustic array described in this embodiment include:

[0075] S1: Determine the dispersion curve based on the structure and material of the wall panel being tested, extract the main frequency of the signal, and determine the angle threshold based on the array structure and the two selected sub-arrays;

[0076] S101: In practice, the acoustic emission signal generated by the leakage source is a broadband signal of (50-400) kHz. There is a dispersion effect when the signal propagates in the wall panel, which causes different frequency components contained in the broadband signal to have different propagation speeds. In specific implementation, a certain frequency band in the broadband should be selected for measurement. The velocity-frequency dispersion curve of the aluminum alloy wall panel with a thickness of 2.5 mm selected in this embodiment is as follows: Figure 2 As shown in the figure, a leakage source signal with two main frequencies of 100 kHz and 200 kHz was simulated in the (50-400) kHz range. The system selected the frequency band with higher energy (90 kHz, 110 kHz) for detection and determined the corresponding propagation velocity to be 1430 m / s based on the dispersion curve.

[0077] After the array receives the signal, a Fourier transform is performed on the signal received by the reference array element to identify the energy level of different frequency bands. Then, a bandpass filter is used to extract the 20kHz bandwidth signal with the largest energy as the measurement signal. The center frequency of the measurement signal is w. The propagation speed varies little within the 20kHz range and can be regarded as a constant value v.

[0078] S102: Under the scenario described in S101, calculate the delay function of the i-th array element of the linear array with an array element spacing of d relative to the reference array element under far-field conditions;

[0079] The expression of the delay function is:

[0080]

[0081] S103: The received signals of all channels of the linear acoustic array are multiplied by the conjugate of the delay coefficient and summed. This can eliminate the effects of delay and noise and obtain the signal strength received in all directions. The direction with the maximum received signal energy is used as the direction and angle threshold of the leakage source. The output of the array in the θ direction is:

[0082]

[0083] In formula (2), M is the number of sensor array elements.

[0084] S2: Calculate the difference in positioning results of the two sub-arrays and compare the positioning result difference with the threshold;

[0085] S201: Select two different sub-arrays and calculate different energy and angle correspondences through formula (2) to obtain two positioning results. When the distance between the leakage source and the reference array element is greater than When the far-field condition is satisfied, theoretically, the positioning result difference between the two sub-arrays in the critical far-field condition is:

[0086]

[0087] In formula (3), x is the distance between the two sub-array reference elements, and D is the aperture of the entire sensor array.

[0088] S202: Set the theoretical angle difference as a threshold, such as Figure 3 As shown, this embodiment selects the first five and last five sensors as two sub-arrays for positioning. The angle threshold of 2.3° can be calculated based on the array parameters and the selected sub-array. The leakage source signal is applied at a distance of 0.1m and 45° from the signal source using software. The positioning results of the two sub-arrays are shown in the figure. The positioning difference is 8.9°, which is higher than the angle threshold. It is considered that the leakage source position does not meet the far-field positioning conditions.

[0089] S3: If the positioning result difference is not greater than the threshold, far-field positioning is performed, and the far-field positioning is completed using the full array and the far-field positioning result is output;

[0090] When the far-field condition is met, the angle threshold of the full array is determined according to S1, and the positioning result of the full array is obtained according to S2, and the positioning result of the full array is output as the far-field positioning result.

[0091] S4: If the positioning result difference is less than the threshold, near-field positioning is performed, and the improved 2Dmusic algorithm is used to process the full array signal to obtain the near-field positioning result.

[0092] S401: In positioning, the arrival wave of the near-field signal source is regarded as a spherical wave. The difference between the distance of the signal reaching the reference array element and the distance of the signal reaching any array element is:

[0093]

[0094] In formula (4), n is the number of intervals between the array element and the reference array element;

[0095] S402: Taking the center array element as the reference array element, the delay function matrix under the near field condition is written as follows according to the distance difference:

[0096]

[0097] In formula (5), is the number of sensor elements on a single side, d is the interval between adjacent elements, is a small first-order quantity in Δ, is a small second-order quantity in Δ;

[0098] S403: Corresponding to the first-order and second-order small quantities in Δ.

[0099] S403: If the near-field beamforming method is used, it is necessary to traverse every angle and distance in the near-field range, which is very computationally intensive. To reduce computational complexity, this embodiment uses an improved 2D music algorithm to achieve near-field positioning. The core of the algorithm is to convert the variables θ and r into γ and φ. γ and φ are obtained through two one-dimensional searches based on the music power spectrum, thereby estimating θ and r. Ω(θ, r) is decomposed into:

[0100]

[0101] S404: Use the measured signal to construct a covariance matrix and decompose it, extract the noise subspace and calculate the music power spectrum:

[0102]

[0103] In formula (7), x(t) is the measurement signal matrix, * represents the conjugate transpose of the matrix, K is the number of snapshots, and U is S is the signal subspace composed of the largest eigenvalue, U N is the noise subspace composed of the remaining eigenvalues;

[0104] S405: Calculating the music power spectrum within the search range of the linear acoustic array using a power spectrum function;

[0105] The expression of the power spectrum function is:

[0106]

[0107] In formula (8), Using formula (6), P(θ,r) can be further written as:

[0108]

[0109] In formula (9),

[0110] S406: At this point, finding the maximum value of P(θ, r) is equivalent to finding the maximum value of P(γ, φ). Introducing the matrix e1 = (0, 0…0, 1) can further express the maximum value problem as a quadratic optimization problem in standard form:

[0111]

[0112] S407: For the quadratic optimization problem, this embodiment uses the Lagrange multiplier method to construct a cost function to solve:

[0113] The expression of the cost function is:

[0114]

[0115] In formula (11), η is an unknown parameter;

[0116] S408: In order to minimize the cost function, this embodiment performs Find the partial derivative equal to zero:

[0117]

[0118] S409: Substituting into the above formula, we can eliminate η and get:

[0119]

[0120] Substitute formula (13) into formula (10) to eliminate φ, and transform the maximum value problem of P(θ,r) into the maximum value problem of the single variable γ. Use the known conditions to search for the spectrum peak of different γ to obtain the value of γ. Repeat S401-S409 to obtain the value of φ. The values ​​of θ and r are used as the position estimation results, as shown in Figure 4 As shown, the covariance matrix of the decomposed signal is obtained to obtain the noise subspace corresponding matrix, and the improved 2Dmusic algorithm is used to estimate the position of the leakage source. Its distance and angle relative to the sensor array are 0.13m and 49°, as shown in Figure 5 As shown in FIG, a two-dimensional spectrum peak search is performed using a power spectrum function within a limited range of ±10° and ±0.05m of the estimated position, and the angle and distance corresponding to the maximum value are taken as the near-field positioning result.

[0121] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as a preferred embodiment as above, it is not intended to limit the present invention. Any technician familiar with the present profession can make some changes or modifications to equivalent embodiments of equivalent changes using the technical content disclosed above without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent replacement and improvement of the above embodiments made according to the technical essence of the present invention, within the spirit and principles of the present invention, without departing from the content of the technical solution of the present invention, shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A full-field localization method for near- and far-field wall panel leakage sources based on a linear acoustic array, characterized in that: The method for full-field positioning of far- and near-field wall panel leakage sources based on a linear acoustic array comprises the following steps: Step 1: Determine the dispersion curve based on the structure and material of the panel being tested, extract the main frequency of the signal, and determine the angle threshold based on the array structure and the two selected sub-arrays; Step 2: Calculate the difference in positioning results of the two sub-arrays and compare the positioning result difference with the threshold; Step 3: If the positioning result difference is not greater than the threshold, perform far-field positioning, use the full array to complete far-field positioning and output the far-field positioning result; Step 4: If the positioning result difference is less than the threshold, near-field positioning is performed, and the improved 2Dmusic algorithm is used to process the full array signal to obtain the near-field positioning result.

2. The method for full-field localization of far- and near-field wall panel leakage sources based on a linear acoustic array according to claim 1, characterized in that: Step 1 specifically includes: Step 1.1: After receiving the acoustic emission signal from the leakage source, the linear acoustic array performs a Fourier transform on the measurement signal of the reference array element to identify the energy of the acoustic emission signal from the leakage source in different frequency bands. The acoustic emission signal from the leakage source is a broadband signal of (50-400) kHz. Step 1.2: Use bandpass filtering to extract the 20 kHz bandwidth signal with the highest energy as the measurement signal, where the center frequency of the measurement signal is w and the propagation speed is a constant v; Step 1.3: Calculate the delay function of the i-th element of the linear array with an element spacing of d relative to the reference element under far-field conditions; Step 1.4: Multiply the received signals of all channels of the linear acoustic array by the conjugate of the delay coefficient and sum them to obtain the signal strength received in all directions. The direction with the maximum received signal energy is used as the direction and angle threshold of the leakage source. The expression of the delay function is:

3. The method for full-field localization of far- and near-field wall panel leakage sources based on a linear acoustic array according to claim 1, characterized in that: The threshold calculation steps in step 2 include: Step 2.1: Calculate the output of the subarray in the θ direction; Step 2.2: Calculate the different energy and angle correspondences based on the output of the sub-array in the θ direction, and obtain the positioning results of the two sub-arrays; Step 2.3: Calculate the positioning result difference of the two sub-arrays under the theoretical critical case when the far-field condition is met, and use the positioning result difference of the two sub-arrays under the theoretical critical case as the threshold. The far-field condition is that the distance between the leakage source and the reference array element is greater than The expression of the subarray output in the θ direction is: In formula (2), M is the number of sensor array elements; The expression of the difference in positioning results between the two sub-arrays in the theoretical critical case is: In formula (3), x is the distance between the two sub-array reference elements, and D is the aperture of the entire sensor array.

4. The method for full-field localization of far- and near-field wall panel leakage sources based on a linear acoustic array according to claim 1, characterized in that: Step 3 specifically includes: When the far-field condition is met, the angle threshold of the full array is determined according to step 1, and the positioning result of the full array is obtained according to step 2, and the positioning result of the full array is output as the far-field positioning result.

5. The method for full-field localization of far- and near-field wall panel leakage sources based on a linear acoustic array according to claim 1, characterized in that: Step 4 specifically includes: Step 4.1: Use the improved 2Dmusic algorithm to convert the variables θ and r into γ and φ, calculate the delay function matrix under near-field conditions, and decompose the delay function matrix Ω(θ,r) to obtain the noise subspace Ω1(γ)Ω2(φ); Step 4.2: Calculate the music power spectrum P(θ,r) based on the noise subspace; Step 4.3: Use the power spectrum function to calculate the music power spectrum within the search range of the linear acoustic array. Introduce the matrix e1 = (0, 0…0, 1) to transform the maximum value problem of P(θ, r) into a quadratic optimization problem in standard form. Step 4.4: Use the Lagrange multiplier method to construct the cost function to solve the quadratic optimization problem; Step 4.5: Ω in the cost function * 2(φ) finds the partial derivative equal to zero to obtain the minimum value of the cost function, and e1 T Ω2(φ)=1Substitute into Ω * 2(φ) In the partial derivative formula, eliminate the unknown parameter η in the cost function to obtain Ω2(φ); Step 4.6: Substitute Ω2(φ) into the quadratic optimization problem and transform the maximum value problem of P(θ,r) into a maximum value problem of the single variable γ; Step 4.7: Use the calculated music power spectrum P(θ, r) as a known condition and use it to search for spectrum peaks for different γ values ​​to obtain the value of γ. Repeat steps 4.1 to 4.6 to obtain the value of φ. The values ​​of θ and r are used as the position estimation results. Step 4.8: Use the power spectrum function to perform a two-dimensional spectrum peak search within a limited range of ±10° and ±0.05m from the estimated position, and use the angle and distance corresponding to the maximum value as the near-field positioning result; The expression of music power spectrum is: In formula (4), x(t) is the measurement signal matrix, * represents the conjugate transpose of the matrix, K is the number of snapshots, and U is S is the signal subspace composed of the largest eigenvalue, U N is the noise subspace composed of the remaining eigenvalues; The expression of the power spectrum function is: In formula (5), The expression of the quadratic optimization problem is: The expression of the cost function is: In formula (7), η is an unknown parameter; right The expression for finding the partial derivative equal to zero is: The calculation formula for Ω2(φ) is: The calculation formula of the noise subspace Ω1(γ)Ω2(φ) is:

6. The method for full-field localization of near- and far-field wall panel leakage sources based on a linear acoustic array according to claim 5, characterized in that: The calculation steps of the delay function matrix under near-field conditions in step 4.1 include: Step 4.1.1: Calculate the difference between the distance from the received signal to the reference array element and the distance from the signal to any sub-array element; Step 4.1.2: Using the center array element as the reference element, calculate the delay function matrix under near-field conditions based on the distance difference; The expression for the difference between the distance from the received signal to the reference array element and the distance from the signal to any array element is: In formula (11), n ​​is the number of intervals between the array element and the reference array element; The expression of the delay function matrix under near-field conditions is: In formula (12), is the number of sensor elements on a single side, d is the interval between adjacent elements, is a small first-order quantity in Δ, is a small second-order quantity in Δ.

Images