Wearable sensor-based device and method for detecting internal voids in wooden structures
By using wearable sensors and deep neural network detection technology on wooden structure buildings, the losslessness and accuracy problems of internal hollow detection of wooden structures in the prior art are solved, and efficient and accurate detection results are achieved.
Patent Information
- Application Number
- CN202110312438.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-03-24
- Publication Date
- 2024-11-05
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In the prior art, wooden structure buildings are difficult to effectively detect internal hollows and corrosion during service, resulting in structural safety hazards. The detection methods are mostly semi-destructive, making efficient and accurate non-destructive testing impossible.
Using a wearable sensor-based detection device, including a flexible PZT sensing unit and a signal detection and analysis device, stress waves are generated through the inverse piezoelectric effect and short-time Fourier transform is performed, and detection is achieved by combining a deep neural network to achieve the detection of the wood. Non-destructive testing of hollows inside the structure.
It improves detection efficiency and accuracy of results, is suitable for wooden structures with different cross-sectional shapes, enhances the reliability and flexibility of the sensor, and can be flexibly applied to wooden components of different cross-sectional sizes.
Smart Images

Figure CN112986386B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of wooden structure buildings, and in particular to a device and method for detecting internal cavities of wooden structures based on wearable sensors. Background Art
[0002] Wood structures use biomass building materials. During normal service, they are susceptible to termite infestation and changes in environmental humidity, causing internal voids and corrosion in the components. Internal voids and corrosion weaken the bearing capacity of the component cross-section, posing a huge safety hazard to the structure.
[0003] At present, the detection of cavities and corrosion inside wooden components in service is still at the semi-nondestructive testing stage. The commonly used methods in engineering are the micro-drill resistance meter method and the Piro nail method. These two methods use a special drill bit or probe to drill into the wooden component, and judge whether there is a cavity or corrosion inside the wooden component by the resistance during the drilling process. Therefore, it is particularly necessary to provide a device and method for detecting cavities inside wooden structures based on wearable sensors to perform nondestructive detection of cavities inside wooden structures. Summary of the invention
[0004] The purpose of the present invention is to provide a device and method for detecting internal voids in wooden structures based on wearable sensors to solve the technical problems existing in the prior art. The device and method are easy to disassemble and assemble, have strong practicality and reliability, realize non-destructive detection of internal voids in wooden structures to be tested, and effectively improve the detection efficiency and the accuracy of the detection results.
[0005] To achieve the above object, the present invention provides the following solution: The present invention provides a device for detecting internal cavities of a wooden structure based on a wearable sensor, comprising: a wearable sensor, a signal detection and analysis device, wherein the wearable sensor is electrically connected to the signal detection and analysis device;
[0006] The wearable sensor includes a connecting belt, which can be detachably wrapped around the outside of the wooden structure to be tested; a plurality of flexible PZT sensor units are provided on the connecting belt, and the number of the flexible PZT sensor units on each connecting belt is greater than or equal to two, wherein any one of the flexible PZT sensor units serves as a driver, and any other one of the flexible PZT sensor units serves as a receiver; the connecting belt includes a hard protection zone and a non-hard protection zone, the position of the hard protection zone corresponds to the position of the flexible PZT sensor unit, and the size of the hard protection zone is larger than the size of the flexible PZT sensor unit; the flexible PZT sensor unit is electrically connected with a lead, and the lead is electrically connected to the signal detection and analysis device.
[0007] Preferably, the connecting belt is a flexible fabric belt, and both ends of the flexible fabric belt are connected by buckles.
[0008] Preferably, the flexible fabric belt comprises an upper fabric layer and a lower fabric layer which are stacked in sequence, the upper fabric layer and the lower fabric layer are continuous along the length direction of the wearable sensor, and the lower fabric layer is in contact with the wooden structure to be measured; the flexible PZT sensing unit is arranged between the upper fabric layer and the lower fabric layer, and the hard protection zone is arranged in the upper fabric layer and / or the lower fabric layer.
[0009] Preferably, the flexible PZT sensing unit comprises an upper bonding layer, a flexible PZT unit, a lower bonding layer and a hard bonding layer which are arranged in sequence from top to bottom, the upper bonding layer is connected to the upper fabric layer, and the hard bonding layer is connected to the lower fabric layer.
[0010] Preferably, the hardness of the hard bonding layer is greater than the hardness of the upper bonding layer and / or the lower bonding layer.
[0011] Preferably, the flexible PZT unit is made of a PvDF film prepared by melt casting.
[0012] Preferably, a first connector and a second connector are respectively provided at both ends of the connecting belt, the first connector is a convex structure, the second connector is a concave structure, the size of the convex structure matches the size of the concave structure, and the first connector and the second connector are connected by buckle.
[0013] Preferably, the hardness of the hard protection area is greater than the hardness of the non-hard protection area.
[0014] The present invention also provides a method for detecting internal cavities of a wooden structure based on a wearable sensor, comprising the following steps:
[0015] Wrapping the wearable sensor around the to-be-tested position of the to-be-tested wooden structure, and electrically connecting the lead wire to the signal detection and analysis device;
[0016] Applying an excitation electrical signal to the driver to generate a stress wave in the wood structure to be tested through the inverse piezoelectric effect of the driver; receiving the stress wave passing through the cross section of the wood structure to be tested through the receiver, and transmitting the received stress wave to the signal detection and analysis device;
[0017] Based on the received stress wave, the signal detection and analysis device is used to detect the internal cavities of the wooden structure to be tested; wherein, the signal detection and analysis device uses a deep neural network based on short-time Fourier transform to detect the internal cavities of the wooden structure to be tested.
[0018] Preferably, the method for the signal detection and analysis device to detect the internal cavity of the wooden structure to be tested specifically includes:
[0019] First, the received stress wave is subjected to short-time Fourier transform to obtain the spectrum of the stress wave;
[0020] Secondly, the spectrum diagram obtained by short-time Fourier transform is input into the trained deep neural network to complete the detection of internal cavities in the wooden structure to be tested.
[0021] The present invention discloses the following technical effects:
[0022] (1) The wearable sensor of the present invention adopts a flexible fabric belt and a flexible PZT sensing unit, which can be applied to the wooden structures to be tested with different cross-sectional shapes, and has the advantage of strong cross-sectional applicability;
[0023] (2) The flexible PZT sensing unit of the wearable sensor of the present invention is provided with a hard adhesive layer, and a hard protection zone is provided at a position on the connecting belt corresponding to the flexible PZT sensing unit, which can effectively prevent the flexible PZT sensing unit from being damaged due to excessive curvature, thereby improving the reliability of the wearable sensor;
[0024] (3) The wearable sensor of the present invention adopts a buckle connection, has the advantages of being detachable and easy to connect, and can be flexibly applied to wooden structures to be tested with different cross-sectional sizes;
[0025] (4) The wearable sensor of the present invention obtains the stress wave generated in the wooden structure to be tested through the positive and negative piezoelectric effects of the flexible PZT sensing unit, performs short-time Fourier transform on the stress wave passing through the cross section of the wooden structure to be tested, obtains the spectrum of the stress wave, and extracts features of the spectrum of the stress wave through a deep neural network, thereby realizing non-destructive detection of the internal voids of the wooden structure to be tested and effectively improving the detection efficiency and the accuracy of the detection results. BRIEF DESCRIPTION OF THE DRAWINGS
[0026] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0027] Figure 1 It is a structural schematic diagram of the wearable sensor of the present invention installed on the wooden structure to be tested;
[0028] Figure 2 This is a schematic diagram of the structure of a wearable sensor of the present invention;
[0029] Figure 3 It is a cross-sectional view of the structure of the flexible PZT sensing unit of the present invention;
[0030] Figure 4This is a schematic diagram of the DNN structure of the present invention;
[0031] In the figure, 1 is a wooden structure to be tested, 2 is a wearable sensor, 11 is a hard protection area, 12 is a first connector, 13 is a flexible PZT sensing unit, 14 is a lead, 15 is a second connector, 21 is an upper fabric layer, 22 is an upper bonding layer, 23 is a flexible PZT unit, 24 is a lower bonding layer, 25 is a hard bonding layer, and 26 is a lower fabric layer. DETAILED DESCRIPTION
[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0033] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.
[0034] Reference Figure 1-2 As shown, this embodiment provides a wood structure internal cavity detection device based on a wearable sensor, which is used to perform internal cavity detection on a wood structure 1 to be detected, comprising: a wearable sensor 2, a signal detection and analysis device, wherein the wearable sensor 2 is electrically connected to the signal detection and analysis device;
[0035] The wearable sensor 2 includes a connecting belt, which can be detachably wrapped around the outside of the wooden structure 1 to be tested; a plurality of flexible PZT sensor units 13 are provided on the connecting belt, and the number of the flexible PZT (Piezoelectric Transducer) sensor units 13 on each connecting belt is greater than or equal to two, wherein any one of the flexible PZT sensor units 13 serves as a driver, and any other one of the flexible PZT sensor units 13 serves as a receiver; the connecting belt includes a hard protection zone 11 and a non-hard protection zone, the position of the hard protection zone 11 corresponds to the position of the flexible PZT sensor unit 13, and the size of the hard protection zone 11 is larger than the size of the flexible PZT sensor unit 13; the flexible PZT sensor unit 13 is electrically connected to a lead 14, and the lead 14 is electrically connected to the signal detection and analysis device.
[0036] According to a further optimization scheme, the connecting belt is a flexible fabric belt, both ends of which are connected by buckles, so the belt is easy to wear, and the flexible fabric belt can be applied to wooden components to be inspected with different cross-sectional shapes.
[0037] According to a further optimization scheme, the flexible fabric belt includes an upper fabric layer 21 and a lower fabric layer 26 which are stacked in sequence, wherein the upper fabric layer 21 and the lower fabric layer 26 are continuous along the length direction of the wearable sensor 2, and the lower fabric layer 26 is in contact with the wooden structure 1 to be measured; the flexible PZT sensing unit 13 is arranged between the upper fabric layer 21 and the lower fabric layer 26, and the hard protection zone 11 is arranged on the upper fabric layer 21 and / or the lower fabric layer 26.
[0038] Further optimizing the scheme, the flexible PZT sensing unit 13 includes an upper bonding layer 22, a flexible PZT unit 23, a lower bonding layer 24, and a hard bonding layer 25 arranged in sequence from top to bottom, the upper bonding layer 22 is connected to the upper fabric layer 21, and the hard bonding layer 25 is connected to the lower fabric layer 26, as shown in FIG. Figure 3 As shown; wherein the flexible PZT unit 23 is the main sensing element; the upper bonding layer 22 and the lower bonding layer 24 play the role of bonding and insulation; the hardness of the hard bonding layer 25 is slightly greater than the hardness of the upper bonding layer 22 and / or the lower bonding layer 24, to prevent the flexible PZT unit 23 from being damaged due to excessive curvature. In this embodiment, the hardness of the hard bonding layer 25 is expressed by bending stiffness, and the bending stiffness of the hard bonding layer 25 is 8 to 10 units (N·m) higher than that of the upper bonding layer 22 and / or the lower bonding layer 24. 2 ).
[0039] According to a further optimization scheme, the flexible PZT unit 23 is made of a PvDF (Polyvinylidene Fluoride) film prepared by a melt casting method, which has positive and negative piezoelectric effects.
[0040] According to a further optimization scheme, a first connector 12 and a second connector 15 are respectively provided at both ends of the connecting belt, wherein the first connector 12 is a convex structure, and the second connector 15 is a concave structure, and the size of the convex structure matches that of the concave structure, and the connecting belt is connected by snaps through the first connector 12 and the second connector 15 to achieve detachable wrapping around the outside of the wooden structure 1 to be tested; in addition, by adopting the snap connection, the appropriate number of wearable sensors 2 can be selected according to the cross-sectional size of the wooden structure 1 to be tested, and multiple wearable sensors 2 can be connected together by the snap connection to be suitable for the detection of wooden components of more sizes.
[0041] Further optimization scheme, the hardness of the hard protection zone 11 is slightly greater than the hardness of the non-hard protection zone; in this embodiment, the weaving density of the hard protection zone 11 is greater than the weaving density of the non-hard protection zone, which is used to protect the flexible PZT sensor unit 13 from damage due to excessive curvature when the wearable sensor 2 is used for cylindrical wooden structure detection with a smaller diameter. In this embodiment, the hardness of the flexible fabric belt is expressed by bending stiffness, and the bending stiffness of the hard protection zone 11 is 8 to 10 units (N·m) higher than that of the non-hard protection zone. 2 ).
[0042] To further optimize the solution, the signal detection and analysis device adopts a multifunctional piezoelectric signal detection and analysis system of model SC-HY-PZT-2.0.
[0043] This embodiment also provides a method for detecting internal cavities in a wooden structure based on a wearable sensor, comprising the following steps:
[0044] Wrapping the wearable sensor around the to-be-tested position of the to-be-tested wooden structure, and electrically connecting the lead wire to the signal detection and analysis device;
[0045] Applying an excitation electrical signal to the driver to generate a stress wave in the wood structure to be tested through the inverse piezoelectric effect of the driver; receiving the stress wave passing through the cross section of the wood structure to be tested through the receiver, and transmitting the received stress wave to the signal detection and analysis device;
[0046] Based on the received stress wave, the signal detection and analysis device detects the internal cavity of the wood structure to be tested; wherein the signal detection and analysis device uses a deep neural network (DNN) based on short-time Fourier transform (STFT) to detect the internal cavity of the wood structure to be tested, specifically including:
[0047] First, the received stress wave is subjected to short-time Fourier transform to obtain the spectrum of the stress wave. The calculation formula of STFT is as follows:
[0048] F(mΔt,f)=∑x(n)g(n-mΔt)e -jωn
[0049] Wherein, j is a complex unit, g(n) is a window function, ω is an angular frequency, f is a frequency, n is a discrete time point, m is time, x(n) is a stress wave signal received by the monitoring device, and Δt is the distance that the window function moves each time.
[0050] Secondly, the spectrum obtained by STFT transformation is input into the trained DNN to detect the internal cavities of the wooden structure to be tested.
[0051] DNN has the advantages of fast recognition efficiency and high recognition accuracy for classification and recognition based on image features. In this embodiment, DNN is designed to be 8 layers, and the specific structure is as follows Figure 4 As shown in FIG, it includes 5 convolutional layers and 3 fully connected layers connected in sequence. The DNN is connected with the first convolutional layer, the second convolutional layer, the third convolutional layer, the fourth convolutional layer, the fifth convolutional layer, the first fully connected layer, the second fully connected layer, and the third fully connected layer in sequence from the input layer to the output layer; the feature image size of the first convolutional layer input is 227×227×3, there are 96 convolution kernels, the convolution kernel size is 11×11, and the step size is 4. The first convolutional layer is connected with a ReLU activation layer and a maximum pooling layer, and the feature image size of the first convolutional layer output is 27×27×96; the feature image size of the second convolutional layer input is 27×27×96, there are 256 convolution kernels, the convolution kernel size is 5×5, and the step size is 1. The second convolutional layer is connected with a ReLU activation layer and a maximum pooling layer, and the feature image size of the second convolutional layer output is 13×13×256; The feature image size of the input of the convolution layer and the fourth convolution layer is 13×13×256, with 384 convolution kernels, the convolution kernel size is 3×3, and the step size is 1. The third and fourth convolution layers are connected with ReLU activation layers, and the feature image size of the output of the fourth convolution layer is 13×13×384; the feature image size of the input of the fifth convolution layer is 13×13×384, with 256 convolution kernels, the convolution kernel size is 3×3, and the step size is 1. The fifth convolution layer is connected with ReLU activation layer and maximum pooling layer, and the feature image size of the output of the fifth convolution layer is 6×6×256; the first fully connected layer and the second fully connected layer each include 4096 neurons, and the first and second fully connected layers are connected with Dropout random inactivation layers; the number of neurons in the third fully connected layer is equal to the number of damage degree categories of the cavity (i.e., the output layer), and p in the output layer is 6×6×256. i is the probability of identifying each damage degree, n is the total number of damage degree categories, the sum of the probabilities of different damage degrees is 1, and the damage degree category with the maximum probability value is taken as the cavity damage degree of the wood structure to be tested. Among them, DNN is trained through historical cavity detection data of different types of wood structures.
[0052] The present invention has the following technical effects:
[0053] (1) The wearable sensor of the present invention adopts a flexible fabric belt and a flexible PZT sensing unit, which can be applied to the wooden structures to be tested with different cross-sectional shapes, and has the advantage of strong cross-sectional applicability;
[0054] (2) The flexible PZT sensing unit of the wearable sensor of the present invention is provided with a hard adhesive layer, and a hard protection zone is provided at a position on the connecting belt corresponding to the flexible PZT sensing unit, which can effectively prevent the flexible PZT sensing unit from being damaged due to excessive curvature, thereby improving the reliability of the wearable sensor;
[0055] (3) The wearable sensor of the present invention adopts a buckle connection, has the advantages of being detachable and easy to connect, and can be flexibly applied to wooden structures to be tested with different cross-sectional sizes;
[0056] (4) The wearable sensor of the present invention obtains the stress wave generated in the wooden structure to be tested through the positive and negative piezoelectric effects of the flexible PZT sensing unit, performs short-time Fourier transform on the stress wave passing through the cross section of the wooden structure to be tested, obtains the spectrum of the stress wave, and extracts features of the spectrum of the stress wave through a deep neural network, thereby realizing non-destructive detection of the internal voids of the wooden structure to be tested and effectively improving the detection efficiency and the accuracy of the detection results.
[0057] The embodiments described above are only descriptions of the preferred 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 device for detecting internal cavities in a wooden structure based on a wearable sensor, used for detecting internal cavities in a wooden structure (1) to be detected, characterized in that: include: A wearable sensor (2) and a signal detection and analysis device, wherein the wearable sensor (2) is electrically connected to the signal detection and analysis device; The wearable sensor (2) comprises a connecting belt, which is detachably wrapped around the outside of the wooden structure (1) to be tested; a plurality of flexible PZT sensor units (13) are arranged on the connecting belt, and the number of the flexible PZT sensor units (13) on each connecting belt is greater than or equal to two, wherein any one of the flexible PZT sensor units (13) serves as a driver, and any other one of the flexible PZT sensor units (13) serves as a receiver; the connecting belt comprises a hard protection zone (11) and a non-hard protection zone, the position of the hard protection zone (11) corresponds to the position of the flexible PZT sensor unit (13), and the size of the hard protection zone (11) is larger than the size of the flexible PZT sensor unit (13); a lead wire (14) is electrically connected to the flexible PZT sensor unit (13), and the lead wire (14) is electrically connected to the signal detection and analysis device; The connecting belt is a flexible fabric belt, and the two ends of the flexible fabric belt are connected by buckles; The detection method of the internal cavity detection device of the wooden structure based on the wearable sensor comprises the following steps: Wrapping the wearable sensor around the to-be-tested position of the to-be-tested wooden structure, and electrically connecting the lead wire to the signal detection and analysis device; Applying an excitation electrical signal to the driver to generate a stress wave in the wood structure to be tested through the inverse piezoelectric effect of the driver; receiving the stress wave passing through the cross section of the wood structure to be tested through the receiver, and transmitting the received stress wave to the signal detection and analysis device; Based on the received stress wave, the signal detection and analysis device detects the internal cavity of the wood structure to be tested; wherein the signal detection and analysis device uses a deep neural network based on short-time Fourier transform to detect the internal cavity of the wood structure to be tested; The flexible fabric belt comprises an upper fabric layer (21) and a lower fabric layer (26) which are stacked in sequence, the upper fabric layer (21) and the lower fabric layer (26) extending along the length direction of the wearable sensor (2), and the lower fabric layer (26) is in contact with the wooden structure (1) to be measured; the flexible PZT sensing unit (13) is arranged between the upper fabric layer (21) and the lower fabric layer (26), and the hard protection area (11) is arranged on the upper fabric layer (21) and / or the lower fabric layer (26); The flexible PZT sensing unit (13) comprises an upper bonding layer (22), a flexible PZT unit (23), a lower bonding layer (24), and a hard bonding layer (25) which are arranged in sequence from top to bottom, the upper bonding layer (22) being connected to the upper fabric layer (21), and the hard bonding layer (25) being connected to the lower fabric layer (26); A first connector (12) and a second connector (15) are respectively provided at two ends of the connecting belt. The first connector (12) is a convex structure, and the second connector (15) is a concave structure. The sizes of the convex structure and the concave structure match each other. The first connector (12) and the second connector (15) are connected by buckle.
2. The wearable sensor-based wood structure internal cavity detection device according to claim 1, characterized in that: The hardness of the hard bonding layer (25) is greater than the hardness of the upper bonding layer (22) and / or the lower bonding layer (24).
3. The wearable sensor-based wood structure internal cavity detection device according to claim 1, characterized in that: The flexible PZT unit (23) is a PvDF film prepared by a melt casting method.
4. The wearable sensor-based wood structure internal cavity detection device according to claim 1, characterized in that: The hardness of the hard protection area (11) is greater than the hardness of the non-hard protection area.
5. The wearable sensor-based wood structure internal cavity detection device according to claim 1, characterized in that: The method for the signal detection and analysis device to detect the internal cavity of the wooden structure to be tested specifically includes: First, the received stress wave is subjected to short-time Fourier transform to obtain the spectrum of the stress wave; Secondly, the spectrum diagram obtained by short-time Fourier transform is input into the trained deep neural network to complete the detection of internal cavities in the wooden structure to be tested.
