Bolt looseness detection method, system and device based on passive flow-induced random guided waves

Abstract

The invention provides a bolt looseness detection method, system and device based on passive flow-induced random guided waves, and the method comprises the steps: enabling a plurality of sensors to be pasted on two flat plates which are connected through bolts, and collecting random guided wave signals which are transmitted in the plates and are caused by environmental noise; processing the acquired random guided wave signal by using a cross-correlation and band-pass filtering method; constructing an imaging equation by using the obtained cross-correlation function under different filtering frequencies; a to-be-imaged area containing all the bolts is divided into discrete grids, each node of the grids is assumed to be a scattering point of guided wave cross-plate propagation, pixel values needed by imaging are calculated through the extracted bolt tightness features, and imaging analysis of the tightness of the multi-bolt structure is achieved. The guided wave propagation model in the multi-bolt structure is provided, the propagation path and energy information of passive random guided waves are extracted, the multi-bolt tightness imaging method based on delay superposition is obtained, and online detection of the tightness of the multi-bolt structure is achieved.

Description

Technical Field

[0001] The present invention relates to the field of bolt loosening detection, and in particular to a bolt loosening detection method, system and device based on passive flow-induced random guided waves. Background Art

[0002] Bolt connections are widely used in various industrial fields. Due to mechanical stress and corrosion in the use environment, bolted connection structures often become loose. Loose bolts can cause parts that should be tightly connected to come apart, posing serious safety hazards to the entire mechanical structure and even leading to safety accidents. In order to ensure the smooth operation of the mechanical structure, it is very important to monitor the tightness of the bolts in real time and locate the loose bolts in time.

[0003] Manual detection is still the most commonly used method for detecting bolt looseness. This method is time-consuming and labor-intensive, relies on the experience of the inspector, and is difficult to apply to online detection. Therefore, it is necessary to develop a structural health monitoring method to replace manual labor and realize online detection of the tightness of bolt structures. Summary of the invention

[0004] In view of the problems existing in the prior art, the present invention provides a bolt loosening detection method, system and device based on passive flow-induced random waveguide. First, a plurality of piezoelectric ceramic sensors are pasted on two plates connected by bolts to collect random waveguide signals propagating in the plates caused by environmental noise; the collected random waveguide signals are processed by using cross-correlation and bandpass filtering methods; the waveguide path is identified and features are extracted using the cross-correlation function obtained at different filtering frequencies; the imaging area containing all bolts is divided into discrete grids, and each node of the grid is assumed to be a scattering point of the waveguide propagating across the plates. The pixel values ​​required for imaging are calculated using the extracted bolt tightness features, so as to realize online detection of the tightness of a multi-bolt structure.

[0005] The present invention provides a bolt loosening detection method based on passive flow-induced random guided waves, which specifically comprises the following steps:

[0006] S1. Collecting random guided wave signals: using sensors to collect multiple groups of random guided wave signals propagating between bolted flat plate structures;

[0007] S2, random waveguide signal processing: performing cross-correlation calculation on the multiple groups of random waveguide signals collected in step S1 to obtain cross-correlation signals and performing bandpass filtering to obtain multiple groups of cross-correlation signals at different filtering frequencies after filtering;

[0008] S3, constructing bolt tightness imaging equation: using the cross-correlation signals at different filtering frequencies obtained in step S2 to identify the waveguide path and construct the bolt tightness imaging equation, specifically including the following sub-steps:

[0009] S31. Divide the entire detection area into discrete grids. Each node of the grid is defined as a scattering point for waveguide propagation across the board. Assume that each node r and a set of sensor pairs located at r1′ and r2′ respectively constitute a theoretical propagation path r1′→r→r2′. When the filtering frequency is f, calculate the theoretical length of each waveguide path. The calculation formula for the theoretical length of the waveguide path corresponding to the node r is as follows:

[0010] d(r)=||r1′-r||+||r-r2′||

[0011] Where r is a node, r1′ represents the position of the sensor on the first flat plate structure, and r2′ represents the position of the sensor on the second flat plate structure;

[0012] S32, calculate the theoretical flight time of the waveguide path: for each node r1 = (x, y), the positions of the two sensors of the i-th sensor pair are r i1 ′=(x i1 ,y i1 ) and r i2 ′=(x i2 ,y i2 ), i = (1, 2, ..., m), m represents the number of sensor pairs;

[0013] Theoretical flight time of waveguide path τ ij The calculation formula for (x,y) is as follows:

[0014]

[0015] Among them, c g (f) represents the group velocity of the guided wave at frequency f; f j represents the jth filter frequency, j = (1, 2, ..., n);

[0016] S33, constructing the imaging equation of the sensor pair through the theoretical flight time of the waveguide path calculated in step S32:

[0017]

[0018] in, It means that the cross-correlation signal of the i-th sensor pair is filtered with a center frequency of f j The cross-correlation envelope amplitude at the theoretical flight time after bandpass filtering;

[0019] S34, superimposing the calculation results of the m groups of sensor pairs and the n filtering frequencies to obtain the final bolt tightness imaging equation:

[0020]

[0021] S4. Evaluate the tightness of the bolts: bring all grid node coordinates into the final bolt tightness imaging equation P(x, y) for calculation, use the imaging equation in step S3 to calculate the imaging amplitude at each grid node, draw a two-dimensional image based on multiple amplitudes and evaluate the tightness of each bolt.

[0022] Preferably, step S1 specifically includes the following sub-steps:

[0023] S11. A plurality of sensors are mounted on the surface of a flat plate structure connected by bolts to measure random guided wave signals, and any two sensors distributed on both sides of the bolts constitute a sensor pair;

[0024] S12. Use a digital sampling device to receive random waveguide signals within 10 to 30 seconds.

[0025] Preferably, in step S11, the sensor arrangement and the bolt spacing satisfy the following relationship:

[0026] |d(r1)-d(r2)|≥5c g (f) / f

[0027] Where f represents the center frequency of the bandpass filter. In the cross-correlation function after the bandpass filter with the center frequency f, the width of the wave packet is 5 / f, which includes five peaks, c g (f) represents the waveguide group velocity at the center frequency f. For a pair of sensors located at r1′ and r2′ respectively and two bolts located at r1 and r2 respectively, two waveguide paths propagating through the two bolts are constructed, namely r1′→r1→r2′ and r1′→r2→r2′, and their lengths are calculated by d1(r1)=||r1′-r1||+||r1-r2′|| and d2(r2)=||r1′-r2||+||r2-r2′|| respectively.

[0028] Preferably, the center frequencies of the different filter frequencies used in the cross-correlation bandpass filtering in step S2 are denoted as f1, f2, ..., f n , and m groups of sensor pairs are arranged on both sides of the bolt joint.

[0029] Preferably, step S2 specifically includes the following sub-steps:

[0030] S21. Calculate the cross-correlation of different sensor pairs. Take two sensors distributed on different plane structures and a bolt connecting the two plane structures as a guide path. The time domain random waveguide signals collected by the two sensors are x1(t) and x2(t), respectively. The cross-correlation function expression is:

[0031]

[0032] Where x1(t) represents the incident wave signal received by the sensor on the first side of the bolt, x2(t) represents the transmitted wave signal received by the sensor on the second side of the bolt, τ represents the time delay, and τ∈(-t0,t0), t0 represents the length of the time domain signal, and the wave packet in the cross-correlation function represents the flight time of the random guided wave propagating through the path between the two sensors, that is, a waveguide path; the cross-correlation signal obtained after the cross-correlation function is bandpass filtered is denoted as C 12 (τ; f), where f represents the center frequency of the bandpass filter;

[0033] S22. Since waveguide propagation has a dispersion effect, a bandpass filter is used to filter the cross-correlation signal. The cross-correlation signal after the bandpass filter with the center frequency f is recorded as C 12 (τ; f);

[0034] S23, calculate the cross-correlation signal C 12 The upper envelope value of (τ; f) is denoted as

[0035] S24, normalizing the cross-correlation signal, normalizing the cross-correlation signal The calculation formula is as follows:

[0036]

[0037] Among them, C 11 (0) represents the zero point value of the autocorrelation of the incident wave signal, which represents the energy level of the incident wave signal. The specific calculation is as follows:

[0038]

[0039] Preferably, the sensor in step S11 is a piezoelectric ceramic sensor.

[0040] Preferably, the specific evaluation method in step S4 is: using multiple amplitudes as pixel values ​​of the imaging thermal map, thereby drawing a two-dimensional image in the imaging area, the pixel value size of the two-dimensional image represents the tightness of the bolt, and based on the two-dimensional image, the tightness of each bolt can be evaluated; when the bolt is tightened in the imaging image, the pixel value of the area where the bolt is located is greater than that of other areas, and when the bolt is completely loose, the pixel value of the area where the bolt is located is smaller than that of other areas.

[0041] A second aspect of the present invention provides a bolt loosening detection system based on flow-induced random guided waves, which includes a random guided wave signal acquisition unit, a random guided wave signal processing unit, a bolt tightness imaging equation construction unit, and a bolt tightness evaluation unit;

[0042] The random guided wave signal acquisition unit uses sensors to collect multiple groups of random guided wave signals propagating between the bolted flat plate structures;

[0043] The random waveguide signal processing unit performs cross-correlation calculation on the multiple groups of random waveguide signals collected in step S1 to obtain cross-correlation signals and performs bandpass filtering to obtain multiple groups of cross-correlation signals at different filtering frequencies after filtering;

[0044] The bolt tightness imaging equation construction unit uses the cross-correlation signals at different filtering frequencies obtained in step S2 to perform waveguide path identification and feature extraction and construct the bolt tightness imaging equation;

[0045] The bolt tightness assessment unit brings all grid node coordinates into the final bolt tightness imaging equation P(x, y) for calculation, and uses the amplitude of the imaging equation at each grid node to draw a two-dimensional image. The pixel value of the two-dimensional image represents the bolt tightness. Based on the two-dimensional image, the tightness of each bolt is assessed.

[0046] According to a third aspect of the present invention, a bolt loosening detection device based on flow-induced random guided waves is provided, which includes a digital sampling device, a piezoelectric ceramic sensor, an air compressor and a jet gun. The digital sampling device is connected to the piezoelectric ceramic sensor, and the jet gun is connected to the air compressor. The air compressor and the jet gun are used to excite and simulate passive random guided waves under real working conditions on a planar structure.

[0047] Compared with the prior art, the present invention has the following advantages:

[0048] (1) The present invention proposes a waveguide propagation model in a multi-bolt structure. In a bolted connection structure, the waveguide propagates between plates only through the tightened bolts, and the transmitted waveguide energy is proportional to the bolt torque. The propagation path and energy information of the random waveguide are further extracted by using the cross-correlation of random waveguide signals combined with bandpass filtering, and a multi-bolt tightness imaging method based on delayed superposition is derived, thereby realizing online detection of the tightness of a multi-bolt structure.

[0049] (2) The detection method based on passive random waveguide proposed in the present invention does not require complex and bulky waveguide excitation equipment. It only requires mounting a number of small, light, and energy-free piezoelectric ceramic sensors on the inner surface of the structure. It has almost no effect on the volume, weight, and aerodynamic shape of the structure, and can be applied to the online detection of loose bolts on aircraft wing skins and other various plate connections. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a schematic diagram of the main process of the bolt loosening detection method based on passive flow-induced random guided waves of the present invention;

[0051] Figure 2 It is a flow chart of bolt tightness imaging in the bolt looseness detection method based on passive flow-induced random guided waves of the present invention;

[0052] Figure 3 It is a schematic diagram of a double-bolt connection structure sample and a flow-induced random guided wave excitation device in the bolt loosening detection method based on passive flow-induced random guided waves of the present invention;

[0053] Figure 4 It is a structural schematic diagram of a bolt loosening detection device for a flat plate structure based on flow-induced random guided waves of the present invention;

[0054] Figure 5 It is a schematic diagram of the experimental samples, specific parameters of sensor arrangement and blowing excitation position under different experimental conditions of the present invention;

[0055] Figure 6a-6b are cross-correlation bandpass filtering images under different experimental conditions of the present invention;

[0056] Figure 7a-7d The two-dimensional imaging results under different tightness conditions of the present invention are shown;

[0057] Figure 8 It is a schematic block diagram of the bolt loosening detection system of the present invention. DETAILED DESCRIPTION

[0058] In order to fully describe the technical content, objectives and effects of the present invention, the following will be described in detail with reference to the accompanying drawings.

[0059] The present invention provides a bolt loosening detection method based on passive flow-induced random guided waves, such as Figure 1 and Figure 2 As shown, the specific steps include:

[0060] S1. Collect random guided wave signals: Use sensors to collect multiple groups of random guided wave signals propagating between bolted flat plate structures.

[0061] This step specifically includes the following sub-steps:

[0062] S11, mounting a plurality of sensors on the surface of the flat plate structure connected by bolts to measure random guided wave signals, and any two sensors distributed on both sides of the bolts constitute a sensor pair. In a specific embodiment, the sensor is a piezoelectric ceramic sensor.

[0063] The sensor arrangement and bolt spacing satisfy the following relationship:

[0064] |d(r1)-d(r2)|≥5c g (f) / f

[0065] Where f represents the center frequency of the bandpass filter. In the cross-correlation function after the bandpass filter with the center frequency f, the width of the wave packet is 5 / f, which includes five peaks, c g(f) represents the waveguide group velocity at the center frequency f. For a pair of sensors located at r1′ and r2′ respectively and two bolts located at r1 and r2 respectively, two waveguide paths propagating through the two bolts are constructed, namely r1′→r1→r2′ and r1′→r2→r2′, and their lengths are calculated by d1(r1)=||r1′-r1||+||r1-r2′|| and d2(r2)=||r1′-r2||+||r2-r2′|| respectively.

[0066] S12. Use a digital sampling device to receive random waveguide signals within 10 to 30 seconds.

[0067] S2. Processing random waveguide signals: performing cross-correlation calculation on the multiple groups of random waveguide signals collected in step S1 to obtain cross-correlation signals and performing bandpass filtering to obtain multiple groups of cross-correlation signals at different filtering frequencies after filtering.

[0068] Preferably, the center frequencies of the different filter frequencies used in the cross-correlation bandpass filtering in step S2 are denoted as f1, f2, ..., f n , and m groups of sensor pairs are arranged on both sides of the bolt joint.

[0069] Preferably, step S2 specifically includes the following sub-steps:

[0070] S21. Calculate the cross-correlation of different sensor pairs. Take two sensors distributed on different plane structures and a bolt connecting the two plane structures as a guide path. The time domain random waveguide signals collected by the two sensors are x1(t) and x2(t), respectively. The cross-correlation function expression is:

[0071]

[0072] Where x1(t) represents the incident wave signal received by the sensor on the first side of the bolt, x2(t) represents the transmitted wave signal received by the sensor on the second side of the bolt, τ represents the time delay, and τ∈(-t0,t0), t0 represents the length of the time domain signal, and the wave packet in the cross-correlation function represents the flight time of the random guided wave propagating through the path between the two sensors, that is, a waveguide path; the cross-correlation signal obtained after the cross-correlation function is bandpass filtered is denoted as C 12 (τ; f), where f represents the center frequency of the bandpass filter.

[0073] S22. Since waveguide propagation has a dispersion effect, a bandpass filter is used to filter the cross-correlation. The cross-correlation signal after the bandpass filter with the center frequency f is recorded as C 12 (τ; f).

[0074] S23, calculate the cross-correlation signal C 12 The upper envelope value of (τ; f) is denoted as

[0075] S24, normalizing the cross-correlation signal, normalizing the cross-correlation signal The calculation formula is as follows:

[0076]

[0077] Among them, C 11 (0) represents the zero point value of the autocorrelation of the incident wave signal, which represents the energy level of the incident wave signal. The specific calculation is as follows:

[0078]

[0079] S3, constructing bolt tightness imaging equation: using the cross-correlation signals at different filtering frequencies obtained in step S2 to identify the waveguide path and construct the bolt tightness imaging equation, specifically including the following sub-steps:

[0080] S31. Divide the entire detection area into discrete grids. Each node of the grid is defined as a scattering point for waveguide propagation across the board. Assume that each node r and a set of sensor pairs located at r1′ and r2′ respectively constitute a theoretical propagation path r1′→r→r2′. When the filtering frequency is f, calculate the theoretical length of each waveguide path. The calculation formula for the theoretical length of the waveguide path corresponding to the node r is as follows:

[0081] d(r)=||r1′-r||+||r-r2′||

[0082] Wherein, r is a node, r1′ represents the position of the sensor on the first flat plate structure, and r2′ represents the position of the sensor on the second flat plate structure.

[0083] S32, calculate the theoretical flight time of the waveguide path: for each node r1 = (x, y), the positions of the two sensors of the i-th sensor pair are r i1 ′=(x i1 ,y i1 ) and r i2 ′=(x i2 ,y i2 ), i = (1, 2, ..., m), m represents the number of sensor pairs.

[0084] Theoretical flight time of waveguide path τ ij The calculation formula for (x,y) is as follows:

[0085]

[0086] Among them, c g (f) represents the group velocity of the guided wave at frequency f; f jrepresents the jth filtering frequency, j = (1, 2, ..., n).

[0087] S33, constructing the imaging equation of the sensor pair through the theoretical flight time of the waveguide path calculated in step S32:

[0088]

[0089] in, It means that the cross-correlation signal of the i-th sensor pair is filtered with a center frequency of f j The cross-correlation envelope amplitude at the theoretical flight time after bandpass filtering.

[0090] S34, according to the imaging equation in step S33, respectively calculate the imaging results of m groups of sensors under n filtering frequencies, superimpose the calculation results of m groups of sensor pairs and n filtering frequencies, and obtain the final bolt tightness imaging equation:

[0091]

[0092] S4. Evaluate the tightness of the bolts: Substitute all grid node coordinates into the final bolt tightness imaging equation P(x,y) for calculation, use the amplitude of the imaging equation at each grid node to draw a two-dimensional image, the pixel value of the two-dimensional image represents the tightness of the bolt, and then evaluate the tightness of each bolt based on the two-dimensional image. The volume evaluation method is: use multiple amplitudes as the pixel values ​​of the imaging heat map to draw a two-dimensional image in the imaging area, the pixel value of the two-dimensional image represents the tightness of the bolt, and the tightness of each bolt can be evaluated based on the two-dimensional image; when the bolt is tightened in the imaging image, the pixel value of the area where the bolt is located is greater than that of other areas, and when the bolt is completely loose, the pixel value of the area where the bolt is located is smaller than that of other areas.

[0093] A second aspect of the present invention provides a bolt loosening detection system based on a bolt loosening detection method using passive flow-induced random guided waves, such as Figure 8 As shown, it includes a random waveguide signal acquisition unit 1, a random waveguide signal processing unit 2, a bolt tightness imaging equation construction unit 3 and a bolt tightness evaluation unit 4.

[0094] The random guided wave signal acquisition unit 1 uses sensors to acquire multiple groups of random guided wave signals propagating between the bolted flat plate structures.

[0095] The random waveguide signal processing unit 2 performs cross-correlation calculation on the multiple groups of random waveguide signals collected in step S1 to obtain cross-correlation signals and performs bandpass filtering to obtain multiple groups of cross-correlation signals at different filtering frequencies after filtering.

[0096] The bolt tightness imaging equation construction unit 3 uses the cross-correlation signals at different filtering frequencies obtained in step S2 to perform waveguide path identification and feature extraction and construct the bolt tightness imaging equation.

[0097] The bolt tightness evaluation unit 4 brings all grid node coordinates into the final bolt tightness imaging equation P(x, y) for calculation, and uses the amplitude of the imaging equation at each grid node to draw a two-dimensional image. The pixel value size of the two-dimensional image represents the bolt tightness. Based on the two-dimensional image, the tightness of each bolt is evaluated.

[0098] A third aspect of the present invention provides a bolt loosening detection device based on a bolt loosening detection method using passive flow-induced random guided waves, such as Figure 4 As shown, it includes a digital sampling device 101, a piezoelectric ceramic sensor 102, an air compressor and an air gun, and the detection device is used to provide passive random waveguide and collect signals. The digital sampling device 101 is connected to the piezoelectric ceramic sensor 102, and the air gun is connected to the air compressor. The air compressor and the air gun are used to excite and simulate passive random waveguide under real working conditions on the plane structure 103, and the two plane structures 103 are connected by bolts 104. Specific embodiments

[0100] The following is a further description of a bolt loosening detection method based on passive flow-induced random guided waves proposed by the present invention in conjunction with an embodiment:

[0101] In a specific embodiment of the present invention, two 6061 aluminum plates connected by two M8 bolts are used, the elastic modulus E of the aluminum plate is 71 GPa, the Poisson's ratio v is 0.33, and the density ρ is 2750 kg / m 3 The specification of a single aluminum plate is 400mm×400mm×1mm, and the distance between the two bolts is 100mm. The experiment uses an air compressor and a spray gun to generate compressed gas and excite flow-induced vibration noise on the aluminum plate. The gas flow rate is about 30m / s. A total of eight piezoelectric ceramic sensors are arranged on the two aluminum plates to receive random waveguide signals. The experimental device and the passive random waveguide excitation device are shown in Figure 1. Figure 3 and Figure 4 The specific parameters of the experimental specimen, sensor arrangement, and air blowing excitation position are shown in Figure 5 During the experiment, a torque wrench was used to change the bolt tightness conditions to (i) T1 = 0 N·m, T2 = 0 N·m; (ii) T1 = 16 N·m, T2 = 0 N·m; (iii) T1 = 0 N·m, T2 = 16 N·m; (iv) T1 = 16 N·m, T2 = 16 N·m, for a total of four experimental conditions. After each adjustment of the experimental conditions, air blowing excitation was performed on the left side of the sample.

[0102] First, the flight time of the guided wave paths through different bolts was verified. Figure 6a is the cross-correlation image of sensors 2 and 5 under experimental condition (ii), Figure 6b is the cross-correlation image of sensors 3 and 8 under experimental condition (iii), Figure 6a and Figure 6b The mid-band filter frequencies are 20kHz, 40kHz, 60kHz, and 80kHz, respectively. The red cross indicates the theoretical flight time corresponding to the waveguide path through the first bolt, and the red circle indicates the theoretical flight time corresponding to the waveguide path through the second bolt. When the bolt is tightened, a wave packet will appear at the theoretical flight time of the bolt on the cross-correlation image. Conversely, when the bolt is loose, there is no wave packet or the wave packet is very small at the theoretical position of the cross-correlation image, which is consistent with the theoretical prediction results, ensuring that the tightness of the bolt can be clearly distinguished from the imaging results.

[0103] Then, the 400 mm × 400 mm area containing the two bolts to be detected is discretized, and then the bolt tightness imaging of this area is performed using the imaging equation in step S3, and four pairs of sensors are used for imaging. Figure 5 The sensor pairs are 1&6, 2&5, 3&7, 4&8, and 10 bandpass filter center frequencies of 10kHz, 20kHz, ..., 100kHz are used. Figure 5 The midpoint of the two bolts is set as the coordinate origin, and the imaging results of the four tightness conditions are as follows: Figure 7a-7d As shown, the red solid hexagon represents the actual position of the fully tightened bolt, and the black hollow hexagon represents the actual position of the fully loosened bolt. From the imaging results of Figure 7, it can be seen that when the bolt is fully tightened, the corresponding area will be highlighted in the imaging image; conversely, when the bolt is completely loose, the corresponding area will not be highlighted around it, or the imaging intensity will be much smaller than when the bolt is fully tightened.

[0104] In this embodiment, by performing two-dimensional imaging analysis on the tightness of a double-bolt connection structure, it is proved that the proposed passive detection method can simultaneously identify the tightness of multiple bolts. In addition, in specific practical applications, the size of the imaging function increases with the number of tightened bolts. This is because the greater the torque on the bolt, the higher the guided wave energy transmitted through it, and the amplitude of the imaging equation is related to the guided wave energy transmitted through the bolt, which can also help determine the tightness of the bolt joint.

[0105] The embodiments described above are only descriptions of the preferred implementation modes of the present invention, and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A bolt loosening detection method based on passive flow-induced random guided waves, characterized in that: The specific steps include: S1. Collecting random guided wave signals: using sensors to collect multiple groups of random guided wave signals propagating between bolted flat plate structures; S2, random waveguide signal processing: performing cross-correlation calculation on the multiple groups of random waveguide signals collected in step S1 to obtain cross-correlation signals and performing bandpass filtering to obtain multiple groups of cross-correlation signals at different filtering frequencies after filtering; S3, constructing bolt tightness imaging equation: using the cross-correlation signals at different filtering frequencies obtained in step S2 to identify the waveguide path and construct the bolt tightness imaging equation, specifically including the following sub-steps: S31. Divide the entire detection area into discrete grids. Each node of the grid is defined as a scattering point for waveguide propagation across the board. Assume that each node r and a set of sensor pairs located at r1′ and r2′ respectively constitute a theoretical propagation path r1′→r→r2′. When the filtering frequency is f, calculate the theoretical length of each waveguide path. The calculation formula for the theoretical length of the waveguide path corresponding to the node r is as follows: d(r)＝||r1′-r||+||r-r2′|| Where r is a node, r1′ represents the position of the sensor on the first flat plate structure, and r2′ represents the position of the sensor on the second flat plate structure; S32, calculate the theoretical flight time of the waveguide path: for each node r1 = (x, y), the positions of the two sensors of the i-th sensor pair are r i1 ′=(x i1 ,y i1 ) and r i2 ′=(x i2 ,y i2 ), i = (1, 2, ..., m), m represents the number of sensor pairs; Theoretical flight time of waveguide path τ ij The calculation formula for (x,y) is as follows: Among them, c g (f) represents the group velocity of the guided wave at frequency f; f j represents the jth filter frequency, j = (1, 2, ..., n); S33, constructing the imaging equation of the sensor pair through the theoretical flight time of the waveguide path calculated in step S32: in, It means that the cross-correlation signal of the i-th sensor pair is filtered with a center frequency of f j The cross-correlation envelope amplitude at the theoretical flight time after bandpass filtering; S34, superimposing the calculation results of the m groups of sensor pairs and the n filtering frequencies to obtain the final bolt tightness imaging equation: S4. Evaluate the tightness of the bolts: bring all grid node coordinates into the final bolt tightness imaging equation P(x, y) for calculation, use the imaging equation in step S3 to calculate the imaging amplitude at each grid node, draw a two-dimensional image based on multiple amplitudes and evaluate the tightness of each bolt.

2. The bolt loosening detection method based on passive flow-induced random guided waves according to claim 1 is characterized in that: Step S1 specifically includes the following sub-steps: S11. A plurality of sensors are mounted on the surface of a flat plate structure connected by bolts to measure random guided wave signals, and any two sensors distributed on both sides of the bolts constitute a sensor pair; S12. Use a digital sampling device to receive random waveguide signals within 10 to 30 seconds.

3. The bolt loosening detection method based on passive flow-induced random guided waves according to claim 2 is characterized in that: In step S11, the sensor arrangement and the bolt spacing satisfy the following relationship: |d(r1)-d(r2)|≥5c g (f) / f Where f represents the center frequency of the bandpass filter. In the cross-correlation function after the bandpass filter with the center frequency f, the width of the wave packet is 5 / f, which includes five peaks, c g (f) represents the waveguide group velocity at the center frequency f. For a pair of sensors located at r1′ and r2′ respectively and two bolts located at r1 and r2 respectively, two waveguide paths propagating through the two bolts are constructed, namely r1′→r1→r2′ and r1′→r2→r2′, and their lengths are calculated by d1(r1)=||r1′-r1||+||r1-r2′|| and d2(r2)=||r1′-r2||+||r2-r2′|| respectively.

4. The bolt loosening detection method based on passive flow-induced random guided waves according to claim 1 is characterized in that: The center frequencies of the different filter frequencies used in the cross-correlation bandpass filtering in step S2 are denoted as f1, f2, ..., f n , and m groups of sensor pairs are arranged on both sides of the bolt joint.

5. The bolt loosening detection method based on passive flow-induced random guided waves according to claim 1 is characterized in that: Step S2 specifically includes the following sub-steps: S21. Calculate the cross-correlation of different sensor pairs. Take two sensors distributed on different plane structures and a bolt connecting the two plane structures as a guide path. The time domain random waveguide signals collected by the two sensors are x1(t) and x2(t), respectively. The cross-correlation function expression is: Where x1(t) represents the incident wave signal received by the sensor on the first side of the bolt, x2(t) represents the transmitted wave signal received by the sensor on the second side of the bolt, τ represents the time delay, and τ∈(-t0,t0), t0 represents the length of the time domain signal, and the wave packet in the cross-correlation function represents the flight time of the random guided wave propagating through the path between the two sensors, that is, a waveguide path; the cross-correlation signal obtained after the cross-correlation function is bandpass filtered is denoted as C 12 (τ; f), where f represents the center frequency of the bandpass filter; S22, use bandpass filtering to filter the cross-correlation signal, and the cross-correlation signal after bandpass filtering with a center frequency f is recorded as C 12 (τ; f); S23, calculate the cross-correlation signal C 12 The upper envelope value of (τ; f) is denoted as S24, normalizing the cross-correlation signal, normalizing the cross-correlation signal The calculation formula is as follows: Among them, C 11 (0) represents the zero point value of the autocorrelation of the incident wave signal, which represents the energy level of the incident wave signal. The specific calculation is as follows:

6. The bolt loosening detection method based on passive flow-induced random guided waves according to claim 1 is characterized in that: In step S11, the sensor is a piezoelectric ceramic sensor.

7. The bolt loosening detection method based on passive flow-induced random guided waves according to claim 1 is characterized in that: The specific evaluation method in step S4 is: using multiple amplitudes as pixel values ​​of the imaging thermal map, thereby drawing a two-dimensional image in the imaging area, the pixel value size of the two-dimensional image represents the tightness of the bolt, and based on the two-dimensional image, the tightness of each bolt can be evaluated; when the bolt is tightened in the imaging image, the pixel value of the area where the bolt is located is greater than that of other areas, and when the bolt is completely loose, the pixel value of the area where the bolt is located is smaller than that of other areas.

8. A bolt loosening detection system for the bolt loosening detection method based on passive flow-induced random guided waves as claimed in claim 1, characterized in that: It includes a random waveguide signal acquisition unit, a random waveguide signal processing unit, a bolt tightness imaging equation construction unit and a bolt tightness evaluation unit; The random guided wave signal acquisition unit uses sensors to collect multiple groups of random guided wave signals propagating between the bolted flat plate structures; The random waveguide signal processing unit performs cross-correlation calculation on the multiple groups of random waveguide signals collected in step S1 to obtain cross-correlation signals and performs bandpass filtering to obtain multiple groups of cross-correlation signals at different filtering frequencies after filtering; The bolt tightness imaging equation construction unit uses the cross-correlation signals at different filtering frequencies obtained in step S2 to identify the waveguide path and construct the bolt tightness imaging equation; The bolt tightness assessment unit brings all grid node coordinates into the final bolt tightness imaging equation P(x, y) for calculation, and uses the amplitude of the imaging equation at each grid node to draw a two-dimensional image. The pixel value of the two-dimensional image represents the bolt tightness. Based on the two-dimensional image, the tightness of each bolt is assessed.

9. A bolt loosening detection device used in the bolt loosening detection method based on passive flow-induced random guided waves as claimed in claim 1, characterized in that: The invention comprises a digital sampling device, a piezoelectric ceramic sensor, an air compressor and an air jet gun. The digital sampling device is connected to the piezoelectric ceramic sensor, the air jet gun is connected to the air compressor, and the air compressor and the air jet gun are used to excite and simulate passive random guided waves under real working conditions on a planar structure.

Images