Multi-source data fusion bridge pier impact damage detection device and method
By embedding piezoelectric sensors and surfaces inside the pier for image acquisition and combining with the convolutional neural network model for data processing, the problem of difficulty in accurately detecting pier damage in the existing technology is solved, and a rapid and accurate evaluation of pier damage is achieved. Improves the safety and reliability of the bridge.
Patent Information
- Application Number
- CN202411864618.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-18
- Publication Date
- 2025-02-25
- Estimated Expiration
- 2044-12-18
AI Technical Summary
The existing pier damage detection technology is difficult to accurately detect the surface and internal damage of the bridge pier in a short time after impact, and lacks a comprehensive analysis of surface and internal damage, which makes it difficult to meet the needs of long-term safe operation of bridges.
The impact damage detection device for the bridge pier that adopts multi-source data fusion includes a piezoelectric sensor array and a surface crack detection device that is embedded inside the bridge pier. The internal stress wave data and the surface crack image acquisition device are obtained through piezoelectric sensors for image acquisition, combined with the convolutional neural network model for crack identification, and finally the overall comprehensive damage index of the bridge pier is obtained through data processing.
It realizes rapid and accurate detection of surface and internal damage after impact of the bridge pier, improves the accuracy and reliability of damage assessment, and can obtain the overall damage of the bridge pier in a short time, meeting the needs of long-term safe operation of the bridge.
Smart Images

Figure CN119309633B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of bridge pier damage detection, and in particular to a bridge pier impact damage detection device and method using multi-source data fusion. Background Art
[0002] With the rapid development of the transportation industry, the scale of bridges is getting larger and larger. In some mountainous areas, rivers or urban areas, bridges may be subjected to impact loads such as falling rocks, ships, and vehicles during operation. As an important part of the transportation network, the safety and stability of bridges are increasingly valued by people.
[0003] As the main supporting structure of the bridge, the pier may cause changes in the internal force of the structure after being hit. What is more serious is that the pier may suffer varying degrees of structural damage. The cracks that appear after the structural damage will lead to a decrease in structural stiffness and corrosion of the steel bars, resulting in a reduction in bearing capacity, affecting the service life of the bridge and the safety of operation. This requires regular or timely damage detection of the impacted piers to determine the status of the piers and ensure the safety and reliability of the piers and bridges.
[0004] At present, after a bridge pier is hit, surface crack damage detection often relies on manual inspection or a single image data source, such as using handheld detectors and cameras for data collection, or using drones for multi-point collection. Not only is the degree of automation low and requires human assistance, most of them cannot be detected within a short period of time after the impact, and the collected images are discontinuous, which cannot guarantee comprehensive crack collection and identification. In subsequent processing, it is also impossible to ensure the continuous splicing of the collected images, and the specific positioning accuracy of the cracks is poor. These will lead to large deviations in the results of damage detection, resulting in extremely inaccurate detection results, affecting subsequent crack ratings or damage ratings, and the implementation of repair strategies.
[0005] At the same time, when conducting bridge pier inspections, most of the focus is on detecting damage to the surface of the piers, while paying no attention to damage inside the piers. Internal damage to the piers can cause structural stability to be compromised and bearing capacity to be reduced. Currently, surface inspections are mostly used to predict internal damage to the piers. Although this can reflect internal damage to the piers to a certain extent, the results are not accurate enough and the effect is poor. In addition, in the current pier impact inspection methods, surface inspections or internal damage inspections are used as independent data to evaluate the overall damage of the piers, without comprehensively considering the damage changes caused by both the surface and internal damage of the piers. This makes it difficult for current pier inspection technology to meet the needs of long-term safe operation of bridges.
[0006] Based on this, how to timely obtain and comprehensively analyze and evaluate the surface and internal damage of a bridge when detecting damage after a collision is an urgent problem that needs to be solved. Summary of the invention
[0007] The purpose of the present invention is to provide a surface crack detection device and method for the impact area of a pier. The device can detect not only cracks on the surface of the pier, but also internal damage to the pier, and integrate various detection data to comprehensively determine the damage state of the pier after the impact, thereby improving the accuracy and reliability of the pier damage assessment.
[0008] In the first aspect, the purpose of the present invention is mainly achieved through the following technical solutions: providing a bridge pier impact damage detection device with multi-source data fusion, comprising:
[0009] A plurality of piezoelectric sensors with a positive piezoelectric effect are pre-buried in the bridge pier, and the plurality of piezoelectric sensors are distributed in an array inside the bridge pier to form a sensor array;
[0010] A surface crack detection device is mounted outside the bridge pier and is used to collect images of cracks on the surface of the bridge pier;
[0011] in,
[0012] The surface crack detection device comprises a vertical walking support and an annular walking mechanism;
[0013] The vertical walking support is used to enclose the surface of the pier and walk along the height direction of the pier;
[0014] The annular walking mechanism is arranged on the vertical walking support and can walk along the pier height direction with the vertical walking support; the annular walking mechanism includes an annular track fixedly connected to the vertical walking support, a rotator that can rotate in a circle in the annular track is arranged in the annular track, and a circumferential driving system is also arranged on the annular track to drive the rotator to rotate in a circle; the rotator is also provided with an image acquisition device that rotates in a circle with the rotator;
[0015] as well as,
[0016] A signal processing system is connected to the piezoelectric sensor, the vertical walking support, the circumferential driving system and the image acquisition device through signals.
[0017] Based on the above technical solution, the annular track is composed of an annular plate body and an enclosure plate enclosed on the outside of the annular plate body, the upper end of the annular plate body inside the enclosure plate forms an installation area, and the rotator is rotatably installed in the installation area.
[0018] Based on the above technical solution, the rotator is a ring-shaped rotating wheel structure; the outer circle and the bottom of the rotator are connected to the pulley through the first spring.
[0019] Based on the above technical solution, the circumferential drive system includes a circumferential drive and a driving gear connected to the driving end of the circumferential drive; a gear structure is provided on the inner side of the rotator, and the driving gear meshes with the gear structure to form a transmission connection.
[0020] Based on the above technical solution, the vertical walking bracket is assembled from multiple bracket components; the bracket component includes a vertical plate and a horizontal plate connected to both ends of the vertical plate, and multiple bracket components are overlapped in sequence through the horizontal plates at both ends to enclose a complete frame structure, and the overlap of any two horizontal plates is fastened by fasteners; the inner side of each of the vertical plates is connected to a driven wheel through a second spring; one of the driven wheels is coaxially connected to a driven gear; a vertical drive is also fixed to one of the horizontal plates corresponding to the vertical plate, and the driving end of the vertical drive is connected to a driving gear, and the driving gear meshes with the driven gear to form a transmission connection.
[0021] Based on the above technical solution, the cross plate is provided with a plurality of vertical through holes evenly spaced along its length direction; the fastener is a bolt, and when the two cross plates are overlapped in sequence, the fastener is fastened by penetrating the coaxial vertical through holes at the overlap of the two cross plates.
[0022] Based on the above technical solution, the image acquisition device includes a bracket connected to the rotator, an industrial camera movably connected to the bracket, and a protective cover mounted on the outside of the industrial camera.
[0023] Based on the above technical solution, an equipment bin with a lower opening is also provided on the top of the pier, and the vertical walking bracket can be moved into the equipment bin to place the vertical walking bracket, the annular walking mechanism and the image acquisition device inside the equipment bin.
[0024] In a second aspect, the present invention also provides a method for detecting pier impact damage by multi-source data fusion based on the above pier impact damage detection device, comprising the following steps:
[0025] A plurality of piezoelectric sensors with a positive piezoelectric effect are embedded in the bridge pier, and the plurality of piezoelectric sensors are distributed in an array inside the bridge pier to form a sensor array; when the bridge pier is hit, coordinate information data of all piezoelectric sensors that generate electrical signals are obtained, and the hit area of the bridge pier is determined based on the coordinate information data;
[0026] The sensors in the sensor array in the entire pier are set as the excitation sensors of the next layer of sensors in the row where the sensors are located, and the sensors in the sensor array are the receiving sensors of the previous layer of sensors in the row where the sensors are located; the excitation sensors are used to emit vibration excitation, and the receiving sensors are used to receive stress wave signals formed inside the pier based on the vibration excitation; when the pier is hit, the excitation sensors and receiving sensors in the corresponding sensor array generate vibration excitation and receive corresponding signals, so as to obtain stress wave data;
[0027] After data processing, the acquired stress wave data is compared with the stress wave data of the intact bridge pier to obtain the internal damage index of the bridge pier after impact;
[0028] At the same time, the crack images of the bridge pier surface in the area are collected, and the collected results are processed to obtain the surface damage index of the bridge pier after the impact;
[0029] Based on the internal damage index and surface damage index, the data of the two are calculated and processed to obtain the overall comprehensive damage index of the bridge pier.
[0030] Based on the numerical range of the overall comprehensive damage index of the bridge pier, the degree of damage to the bridge pier after the impact was rated.
[0031] Based on the above detection method, when the bridge pier is hit, the coordinate information data of all piezoelectric sensors that generate electrical signals are obtained, and the impact area of the bridge pier is determined based on the coordinate information data, specifically:
[0032] Taking any position of the bridge pier as the coordinate origin, numbering the piezoelectric sensors, obtaining the coordinate information of each piezoelectric sensor, and associating the number and coordinate information of each piezoelectric sensor;
[0033] When the bridge pier is hit, the number of the piezoelectric sensor that generates the electrical signal is obtained, and the coordinate information of the corresponding area of the piezoelectric sensor corresponding to all the numbers is obtained, and the bridge pier area involved in the coordinate information is determined as the impacted area of the bridge pier.
[0034] Based on the above detection method, when the bridge pier is hit, the number of the piezoelectric sensor that generates the electrical signal is obtained, and the coordinate information of the corresponding area of the piezoelectric sensor corresponding to all the numbers is obtained. The bridge pier area involved in the coordinate information is determined as the bridge pier hit area, specifically:
[0035] A corresponding electric signal threshold is set, and when the bridge pier is hit, the number of the piezoelectric sensor whose electric signal peak value is greater than the electric signal threshold is obtained;
[0036] Extract the height coordinate information of all piezoelectric sensors whose electrical signal peaks exceed the electrical signal threshold, traverse the height coordinate information of all piezoelectric sensors, and obtain the maximum height and the minimum height in the height coordinate information;
[0037] The pier surface area intercepted between the maximum height and the minimum height is determined as the pier impact area.
[0038] Based on the above detection method, the crack image of the bridge pier surface in the area is collected synchronously, and the collected results are processed to obtain the surface damage index of the bridge pier after the impact; specifically:
[0039] Based on the determined impact area of the bridge pier, determine the initial acquisition position and the acquisition end position of the image acquisition device in the height direction of the bridge pier;
[0040] The vertical traveling support is controlled to move, the image acquisition device is moved to the initial acquisition position, and the image acquisition device moves to the designed spacing. The annular drive system drives the rotator to rotate at a constant speed. The image acquisition device synchronously acquires images of the annular area. After completion, it moves with the same designed spacing and acquires images of the next annular area, until the image acquisition device moves to the acquisition end position and completes the crack image acquisition;
[0041] Constructing a convolutional neural network model, establishing a sample set of bridge pier images with cracks, training the convolutional neural network model, and obtaining a crack recognition prediction model;
[0042] Inputting the collected crack images into the crack recognition prediction model, and obtaining the number of cracks, crack width and crack length in all images based on the crack recognition prediction model;
[0043] Based on the obtained crack number, crack width and crack length, the crack width damage coefficient WDIC is defined as:
[0044]
[0045] Where WC i is the width of the i-th crack; n1 is the number of cracks;
[0046] The length damage coefficient LDIC of the crack is defined simultaneously as:
[0047]
[0048] Where WL i is the length of the i-th crack; n1 is the number of cracks;
[0049] The surface damage index DIC of the pier surface is defined as:
[0050]
[0051] Where W W and W L is the weight coefficient of the crack width and crack length damage index, and the sum of the two is 1.
[0052] Based on the above detection method, based on the determined impact area of the bridge pier, the initial acquisition position and the acquisition end position of the image acquisition device in the height direction of the bridge pier are determined, specifically:
[0053] The height coordinate position corresponding to the maximum height in the impact area of the bridge pier is used as the initial acquisition position;
[0054] The height coordinate position corresponding to the minimum height value in the impact area of the pier is taken as the end position of the acquisition.
[0055] Based on the above detection method, a convolutional neural network model is constructed, a sample set of bridge pier images with cracks is established, the convolutional neural network model is trained, and a crack recognition prediction model is obtained; specifically:
[0056] Select the convolutional neural network model as the base model;
[0057] Obtain crack images of bridge piers or other concrete structures, and process and annotate crack information in the crack images, wherein the crack information includes at least the number of cracks, crack length, and crack width;
[0058] The annotated crack image is processed and then input into the convolutional neural network model for training to obtain the convolutional neural network model.
[0059] Based on the above detection method, the acquired stress wave data is processed and compared with the stress wave data of the intact pier to obtain the internal damage index of the pier after impact; specifically:
[0060] When the bridge pier is hit, n2 stress wave time domain data are obtained, and the obtained n2 stress wave data are sorted from small to large according to the stress wave peak value. After sorting, the stress wave data are processed to obtain the total energy of the stress wave time domain data, and the n2 stress wave data when the bridge pier is intact are obtained and processed in the same way;
[0061] Taking two adjacent piezoelectric sensors in the same column as a pair, let the damage index DII of the j-th pair of piezoelectric sensors be j for:
[0062] ①
[0063] In formula ①, since the time-domain peak value and total energy of the stress wave after the pier is damaged will decay, the larger its value is, the more serious the damage to the pier is; n2 is the number of stress waves generated; SWT is j,i is the time domain peak value of the stress wave ranked i generated by the jth pair of piezoelectric sensors when the pier is intact; CSWT j,i is the time domain peak value of the stress wave ranked i generated by the jth pair of piezoelectric sensors after the bridge pier is impacted; E j,i is the total energy of stress waves generated by the jth pair of piezoelectric sensors ranked i when the pier is intact; CE j,i is the total energy of stress waves ranked i generated by the jth pair of piezoelectric sensors after the bridge pier is impacted;
[0064] Get the crack depth CD corresponding to the jth pair of piezoelectric sensors in the pier j ; When 0<DII j ≤1, obtain the crack depth CD corresponding to the jth pair of piezoelectric sensors in the pier j ; When DII j When ≤0, it is considered that the damage to the bridge pier is very small, and almost no cracks occur, then CD j =0, then there is no need to obtain the crack depth CD of the corresponding piezoelectric sensor j ;
[0065] Based on the obtained crack depth CD j , and according to the DII obtained for each pair of piezoelectric sensors j The weighted internal damage index DII of the bridge pier after impact is obtained:
[0066] ②
[0067] In formula ②, CD max is the maximum crack depth determined by n3 pairs of piezoelectric sensors, n3 is the number of all piezoelectric sensor pairs in the entire pier, k is 0<DII j Number of piezoelectric sensor pairs ≤ 1.
[0068] Based on the above detection method, the internal damage index and surface damage index are obtained, and the data of the two are calculated and processed to obtain the overall comprehensive damage index of the pier, which is:
[0069] Based on the obtained internal damage index and surface damage index, the following formula is established:
[0070] ③
[0071] In formula ③, W DII and W DICare weight coefficients, and their sum is 1; D is the overall comprehensive damage index of the pier, DII is the internal damage index, and DIC is the surface damage index;
[0072] The overall comprehensive damage index of the pier can be obtained based on formula ③.
[0073] Compared with the prior art, the present invention has the following beneficial effects:
[0074] The piezoelectric sensors pre-buried in the present invention are arranged in an array. Based on their positive piezoelectric effect pressure, the coordinates of the piezoelectric sensors can be used after the pier hits or generates an electrical signal, thereby quickly and accurately obtaining the impact position of the pier and confirming the accuracy of the area. The detection device is arranged on the pier as a whole, and image acquisition can be performed at any time to perform pier surface crack detection. The detection is convenient, and based on the vertical walking bracket and the annular walking mechanism, the image acquisition device can realize comprehensive detection along the height and annularity of the pier, with a high degree of automation. At the same time, the detection speed and the detection area are uniform, and the images obtained by the detection are continuous, which is convenient for the subsequent crack positioning on the pier. The vertical walking bracket is composed of a plurality of bracket components assembled and can be flexibly adjusted based on the shape and size of the pier, thereby adapting to the use of diverse piers, and has broad application prospects.
[0075] At the same time, the pre-buried array-arranged piezoelectric sensors can also be synchronously formed into multiple pairs in columns, and each pair can emit vibration excitation and receive stress wave signals formed by corresponding vibration excitation, so as to detect damage inside the pier, and obtain internal damage indicators in combination with the acquired stress wave signal data, so as to evaluate the internal damage, and obtain surface damage indicators in combination with the collected pier surface crack image data, and finally perform comprehensive processing based on the internal damage indicators and surface damage indicators, so as to comprehensively evaluate the overall damage of the pier by combining the multi-source data of internal and external damage indicator data, and perform damage assessment in this way. Compared with judging the damage of the pier by single internal damage or surface damage, the assessment result is more accurate and suitable for large-scale promotion and application. BRIEF DESCRIPTION OF THE DRAWINGS
[0076] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of this application, and do not constitute a limitation of the embodiments of the present invention. In the drawings:
[0077] Figure 1 This is a schematic diagram of the application of the device for detecting damage caused by impact to a bridge pier according to the present invention;
[0078] Figure 2 It is a structural schematic diagram of a device for detecting damage caused by impact to a bridge pier according to the present invention;
[0079] Figure 3 It is a structural schematic diagram of the rotating device of the present invention;
[0080] Figure 4 It is a top view of the structure of the device for detecting damage caused by impact to a bridge pier according to the present invention;
[0081] Figure 5 It is a schematic structural diagram of a vertical traveling support in the device for detecting damage to a bridge pier caused by impact according to the present invention;
[0082] Figure 6 It is a schematic diagram of the layout of the piezoelectric sensor of the present invention when it is used, wherein the bridge pier is a square bridge pier;
[0083] Figure 7 It is a flow chart of the method for detecting bridge pier impact damage by multi-source data fusion of the present invention;
[0084] Figure 8 It is a working principle diagram of a piezoelectric sensor for internal damage detection in the bridge pier impact damage detection method based on multi-source data fusion of the present invention;
[0085] Fig. 9 It is a schematic diagram of the piezoelectric sensor signal processing for internal damage detection in the bridge pier impact damage detection method with multi-source data fusion of the present invention.
[0086] The numbers in the figure represent respectively: 1. piezoelectric sensor; 2. vertical walking bracket; 3. annular walking mechanism; 4. annular track; 5. rotator; 6. annular drive system; 7. image acquisition device; 8. annular plate; 9. enclosure plate; 10. first spring; 11. pulley; 12. annular drive; 13. drive gear; 14. vertical plate; 15. horizontal plate; 16. second spring; 17. driven wheel; 18. driven gear; 19. vertical drive; 20. driving gear; 21. equipment compartment. DETAILED DESCRIPTION
[0087] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with embodiments and drawings. The exemplary embodiments of the present invention and their description are only used to explain the present invention and are not intended to limit the present invention.
[0088] like Figure 1-Figure 6As shown, the first embodiment of the present invention discloses a multi-source data fusion bridge pier impact damage detection device, comprising: a plurality of piezoelectric sensors 1 with positive piezoelectric effect, all of which are pre-buried in the bridge pier, and a plurality of the piezoelectric sensors 1 are arrayed inside the bridge pier to form a sensor array; a surface crack detection device, which is sleeved outside the bridge pier and is used for collecting images of cracks on the surface of the bridge pier; wherein the surface crack detection device comprises a vertical walking bracket 2 and an annular walking mechanism 3; the vertical walking bracket 2 is used to enclose the bridge pier surface and walk along the height direction of the bridge pier; the annular walking mechanism 3 is arranged on the vertical The annular walking mechanism 3 includes an annular track 4 fixedly connected to the vertical walking bracket 2, a rotator 5 which can rotate in a circle in the annular track 4 is arranged in the annular track 4, and a circumferential driving system 6 is also arranged on the annular track 4 to drive the rotator 5 to rotate in a circle; the rotator 5 is also provided with an image acquisition device 7 which rotates in a circle with the rotator 5; and a signal processing system, wherein the signal processing system is connected with the piezoelectric sensor 1, the vertical walking bracket 2, the circumferential driving system 6 and the image acquisition device 7 through signals.
[0089] During application, if the pier is a circular pier, the piezoelectric sensor 1 can be arranged at a certain distance in the circumferential, vertical and height directions of the pier, such as arranging multiple circles in a circular shape from the outside to the inside along the cross-section of the pier, with several piezoelectric sensors in each circle, and arranging multiple groups at intervals along the height direction of the pier, thereby forming the array arrangement; similarly, if the pier is a square pier, the piezoelectric sensor 1 can be arranged along the side length direction, vertical and height directions, such as evenly distributing multiple piezoelectric sensors 1 in a matrix form in each cross-section of the pier, and arranging multiple groups at intervals along the height direction of the pier, thereby forming the array arrangement.
[0090] After the piezoelectric sensor array is completed, the surface crack detection device is integrally mounted on the outside of the pier. Specifically, the entire device is enclosed on the surface of the pier by a vertical walking bracket 2, and then directly assembled on the pier in a sleeve manner. When it is necessary to move along the height direction of the pier to cooperate with multi-region image acquisition, it can also move along the height direction of the pier, thereby driving the annular walking mechanism 3 and the image acquisition device 7 to follow up, thereby ensuring the image acquisition requirements in the height direction. At the same time, the annular walking mechanism 3 is arranged on the vertical walking bracket 2 to ensure that it can be moved in the height direction. It can also rotate synchronously in a circle, that is, the rotator 5 is driven by the annular drive system 6 to rotate in a circle on the annular track 4, thereby driving the image acquisition device 7 to perform annular scanning along the circumference of the pier. Then, through the coordinated action of each structure, it is ensured that the image acquisition device 7 can move and stay at any position along the height and circumference of the pier, thereby realizing comprehensive image acquisition.
[0091] Based on this, the detection device is based on the pre-buried piezoelectric sensor 1. When the pier is hit, based on the piezoelectric effect, the piezoelectric sensor in the impact area can generate an electrical signal, which can be obtained and processed by the signal processing system, so as to determine whether the pier is hit and obtain the impact area of the pier. At the same time, based on the piezoelectric sensor 1, internal damage detection can also be performed based on it, and internal damage assessment can be performed in combination with stress wave data. After the impact area of the pier is obtained, the detection device is directly installed on the pier, and the signal processing system can directly control the detection device to quickly perform surface detection of the impact area of the pier, and accurately control the walking distance of the vertical walking bracket 2 and the rotation angle of the rotator 5, so as to realize the entire pier or The whole process of full-surface crack detection in the impacted area is completed automatically without manual assistance, with high acquisition accuracy. The vertical walking support 2 and the rotator 5 are relatively stable during operation, and the overall walking path is controllable, thereby ensuring the detection quality and the uniformity and continuity of the detection area. The image acquisition results are more accurate, and the continuous acquisition results are also convenient for subsequent image processing, such as continuous stitching of images, locating the coordinate position of cracks on the piers, etc., so that not only comprehensive and stable image acquisition can be achieved, but also the timeliness, accuracy and continuity of the acquisition can be ensured. After the acquisition is completed, the signal processing system can make a comprehensive judgment on the degree of damage to the pier based on the acquisition results and the internal damage of the pier, thereby realizing the fusion of multi-source data.
[0092] like Figure 2 As shown, in specific applications, the annular track 4 is composed of an annular plate body 8 and an enclosing plate 9 enclosed on the outside of the annular plate body 8, and the upper end of the annular plate body 8 on the inner side of the enclosing plate 9 forms an installation area, and the rotator 5 is rotatably installed in the installation area.
[0093] The annular plate body 8 is horizontally arranged as a whole to form a supporting structure, and the enclosure plate 9 can limit the rotation of the rotator 5. The two as a whole form a semi-enclosed structure with an outer side convex upward, and the inner area can form a larger space for installing supporting equipment and facilities. In this embodiment, the upper end of the annular plate body 8 on the inner side of the enclosure plate 9 is used as the installation area to install the rotator 5, so that the annular notch in the middle of the annular plate body 8 is convenient for being spaced apart from the bridge pier, ensuring that the annular walking mechanism 3 will not contact the bridge pier when it rotates as a whole, thereby ensuring the circumferential movement space of the annular walking mechanism 3.
[0094] As a further embodiment, in combination Figure 2 , Figure 3As shown, the rotator 5 is a circular rotating wheel structure; the outer circle and the bottom of the rotator 5 are connected to the pulley 11 through the first spring 10. In order to ensure the stable rotation of the rotator 5 and make up for some installation defects, the bottom and the outer circumference of the rotator 5 of this embodiment are connected to the pulley 11 through the first spring 10, so that the pulley 11 can reduce the friction between the rotator 5 and the annular track 4, ensuring its smooth rotation. Specifically, the pulley 11 on the outer circle of the rotator 5 abuts against the inner side of the enclosure plate 9, and the pulley 11 at the bottom of the rotator 5 is supported on the upper end surface of the annular plate body 8, thereby not only ensuring the support of the rotator 5, but also ensuring the smoothness of its circumferential rotation. Furthermore, the pulley 11 is connected by the first spring 10. For the pulley 11 abutting against the enclosure plate 9, the first spring 10 can apply force to squeeze it toward the enclosure plate 9, so that the enclosure plate 9 provides a directional force to support the circumferential rotation of the rotator 5 without shaking or deflecting, thereby ensuring that the rotation path is unique and reliable, and the first spring 10 located at the bottom can support the rotator 5, reduce the pressure of the gravity of the rotator 5 on the pulley 11, and ensure smooth rotation of the pulley 11.
[0095] As a further embodiment, multiple groups of pulleys 11 are evenly spaced on the outer circle and the bottom of the rotator 5, and each pulley 11 is connected by a first spring 10. The first springs 10 are all in a compressed state, and then, the first springs 10 located on the outside of the rotator 5 interact with each other, which can stabilize the rotator 5 in the required rotation area, and when the overall device shakes or is slightly installed and tilted, it can also ensure that the rotator remains in a reasonable installation position, improve the fault tolerance rate, and at the same time, the design of multiple first springs 10 and pulleys 11 at the bottom of the rotator 5 can also evenly support the rotator 5, and reduce the friction between the pulley 11 and the annular plate 8 when the pulley 11 rotates, further ensuring the smooth rotation of the rotator 5.
[0096] As a further embodiment, the circumferential drive system 6 includes a circumferential drive 12 and a driving gear 13 connected to the driving end of the circumferential drive 12; a gear structure is provided on the inner side of the rotator 5, and the driving gear 13 meshes with the gear structure to form a transmission connection.
[0097] In a specific implementation, the annular driver 12 is fixed on the annular track 4, such as the annular plate 8 in some embodiments. As a specific selection, the annular driver 12 is a servo motor, and the servo motor can lock the free rotation state of the rotator 5, thereby preventing the rotator 5 from rotating freely when not driven and affecting the position acquisition of image acquisition.
[0098] In the prior art, the diameters of bridge piers vary greatly based on bridge design standards. There are also different types of pier structures, such as square piers and circular piers. Therefore, when the vertical walking bracket 2 of this embodiment is matched with the piers, it is also extremely critical to ensure that it is compatible with piers of different sizes and types.
[0099] Based on this, Figure 4 , Figure 5 As shown, the vertical walking support 2 is assembled from a plurality of support components; the support component includes a vertical plate 14 and a horizontal plate 15 connected to both ends of the vertical plate 14, and a plurality of the support components are overlapped in sequence through the horizontal plates 15 at both ends to enclose a complete frame structure, and the overlap of any two horizontal plates 15 are fastened by fasteners; the inner side of each of the vertical plates 14 is connected to a driven wheel 17 through a second spring 16; one of the driven wheels 17 is coaxially connected to a driven gear 18; a vertical driver 19 is also fixed to one of the horizontal plates 15 corresponding to the vertical plate 14, and a driving gear 20 is connected to the driving end of the vertical driver 19, and the driving gear 20 is meshed with the driven gear 18 to form a transmission connection.
[0100] During implementation, a suitable number of bracket assemblies are selected according to the size and shape of the pier to ensure that the frame structure composed of all bracket assemblies can be completely enclosed on the surface of the pier. When enclosed, the bracket assemblies are overlapped by the cross plates 15 at the upper and lower ends. After completion, a complete frame is formed, and the inner side of the frame is in contact with the surface of the pier through multiple driven wheels 17 and is tightened by the second spring 16, so that the entire frame can be firmly maintained at any position on the surface of the pier. When walking is required, the vertical drive 19 is turned on to drive one of the driven wheels 17 to rotate, thereby driving the entire frame to walk.
[0101] As a specific implementation, in any bracket assembly: the two horizontal plates 15 are provided with a through-opening groove along the length direction thereof, or the two horizontal plates 15 are provided with a plurality of vertical through holes evenly distributed along the length direction thereof. When overlapped, the fastener passes through the through-opening groove or the vertical through holes corresponding to the upper and lower horizontal plates 15, thereby stabilizing the overlapped joint and realizing the sequential connection of multiple bracket assemblies.
[0102] As a further implementation method, the size of the pier can be measured in advance, so that corresponding slots can be set at corresponding positions of the cross plates 15, and the cross plates 15 overlapped therewith can be directly inserted therein for overlap. At the same time, threaded holes are set in the slots, and through holes are set at corresponding positions of the cross plates 15 overlapped therewith. After overlap, bolts can be used to penetrate the through holes and be threadedly connected to the threaded holes to tighten the overlapped parts of the two, so that the cross plates 15 can be firmly overlapped with each other.
[0103] In a specific implementation, when both of the two transverse plates 15 are provided with through-opening grooves along their length direction, the through-opening grooves are only provided at the two ends of the transverse plate 15, so as to avoid the strength reduction of the transverse plate 15 due to the overall grooves.
[0104] In a specific implementation, the two transverse plates 15 are each provided with a plurality of vertical through holes evenly distributed along their length direction, specifically: the transverse plates 15 are provided with a plurality of vertical through holes evenly spaced along their length direction; the fasteners are bolts, and when the two transverse plates are overlapped in sequence, the fasteners are fastened by penetrating the coaxial vertical through holes at the overlap of the two plates.
[0105] When the two transverse plates 15 are evenly distributed with a number of vertical through holes along their length direction, after the two transverse plates 15 are overlapped, the vertical through holes corresponding to the two overlapping transverse plates 15 may not be able to remain coaxial, thereby affecting subsequent fastening. Based on this, the vertical through holes can be set as waist-shaped holes, so that after the overlap, the two corresponding vertical through holes have more space for pairing and installing fasteners, thereby improving the fault tolerance during assembly.
[0106] As a specific implementation, the vertical driver 19 is a servo motor, and the servo motor can lock the free rotation state of the driven gear 18, further ensuring the stability of the vertical walking support 2 in the vertical direction of the pier.
[0107] It should be noted that after installation, all the second springs 16 are in a compressed state, so based on their interaction force on the pier and the frame, the frame structure as a whole will not automatically shift under the influence of gravity by squeezing the driven wheel 17, that is, the driven wheel 17 is acted upon by the elastic force of the second spring 16, and its force on the surface of the pier increases, and the sum of the friction forces received by all the driven wheels 17 is greater than or equal to the overall gravity of the vertical walking bracket 2, and thus the vertical walking bracket 2 can be stably stayed at any position of the pier as a whole. When the vertical drive 19 starts to drive the corresponding driven wheel 17 to rotate, this state is broken, and then the driving action of the corresponding driven wheel can drive the vertical walking bracket 2 to move as a whole, and due to the above reasons, it can also provide a certain reverse force when moving, and can also slow down its walking speed, and thus the walking state of the vertical walking bracket 2 is easier to control and more stable.
[0108] As a specific implementation, the image acquisition device 7 includes a bracket connected to the rotator 5, an industrial camera movably connected to the bracket, and a protective cover mounted on the outside of the industrial camera. In specific applications, the industrial camera can be selected to have a built-in memory card so that it can store the acquired images. Furthermore, the industrial camera has a built-in wireless function, so that the acquired images can be transmitted to the signal processing system in a timely manner through the wireless function, realizing convenient image data transmission, and also facilitating the reception of remote signals to open and close the device itself and retrieve the required image data under the signal command.
[0109] As a specific implementation, the driven wheel 17 and the pulley 11 are both made of rubber material, which can increase the friction between the two and the bridge pier to ensure the overall stability of the device.
[0110] As a specific implementation, any of the driven wheels 17 is also provided with a distance sensor. When the driven wheel 17 starts to rotate, its travel distance can be detected at any time, so that the travel distance of the entire device can be detected based on the distance sensor, ensuring that the device can comprehensively and completely scan the required area on the pier surface, and can accurately control the overall travel distance of the device and the coordinate position that needs to be positioned and stopped, further improving the overall accuracy and reliability of the device. Specifically, the distance sensor can be connected to the signal processing system signal. When walking, the distance sensor can timely transmit the distance data to the signal processing system, so that the signal processing system can timely control the overall walking state and walking distance of the device.
[0111] like Figure 1 As shown, the top of the pier is also provided with an equipment bin 21 with a lower opening, and the vertical traveling support 2 can be moved into the equipment bin 21 to place the vertical traveling support 2, the annular traveling mechanism 3 and the image acquisition device 7 inside the equipment bin 21.
[0112] When in use, this embodiment is arranged outside the bridge pier and exposed to the outdoor environment, and is extremely susceptible to damage and erosion. In order to ensure the safety of the equipment and extend its service life, the equipment bin 21 of this embodiment is an overall cover structure with an opening at the bottom. When the detection device is not in use, it can walk into the equipment bin 21, so that it can be shielded and protected by the equipment bin 21, thereby protecting the corresponding components and electrical devices on the device.
[0113] As a further embodiment, a socket sleeve may be provided in the middle of the equipment bin 21 for socketing on the bridge pier, and a through hole is provided on the socket sleeve, and fasteners such as bolts are provided through the through hole to fix the equipment bin 21 on the bridge pier.
[0114] As a further implementation, a power supply system is also provided in the equipment compartment 21, and the power supply system is connected to the electrical equipment of the detection device through a line. Specifically, the circumferential drive 12, the vertical drive 19, and the image acquisition device 7 in the detection device can all be connected to the power supply system to realize power supply.
[0115] As another implementation, when the signal processing system is arranged at the near end, it can also be arranged in the equipment compartment 21. Thus, it can not only play a good protective role for it, but also solve the power supply problem.
[0116] As another embodiment, the circumferential drive 12, the vertical drive 19 and the image acquisition device 7 are all connected to a battery, and the battery is arranged on the vertical walking support 2 or the circular track 4. A power supply system is also arranged in the equipment compartment 21, and the power supply system is connected to a power supply plug. When the detection device is located in the equipment compartment 21, the power supply plug is plugged into the charging port of the battery to supply power to the battery. Furthermore, by using this method, the setting of the line can be omitted, reducing the influence of the line on the movement of the device.
[0117] In specific applications, the power supply system can be a power supply cable system or a battery. When it is a battery, the battery can be connected to the power supply cable system for power replenishment, or it can be converted through external renewable energy. For example, a photovoltaic system can be set inside or outside the equipment warehouse 21, and the photovoltaic system can be used to obtain electrical energy and store it in the battery. It can also be a wind power generation system set outside the equipment warehouse 21, and wind energy can be used for power replenishment. The specific structures and implementation methods of the photovoltaic system, wind power generation system, etc. belong to the prior art, and this embodiment will not be redundantly described.
[0118] As a specific implementation, the signal processing system of this embodiment is used to connect the piezoelectric sensor 1, the circumferential driver 12, the vertical driver 19 and the image acquisition device 7 by signal. When used, the signal processing system is connected to the piezoelectric sensor 1 by signal, which can not only obtain the piezoelectric signal generated by it in time, so as to detect whether the bridge pier is hit, but also send out a signal to control the piezoelectric sensor 1 to be powered on and vibrate to send out vibration excitation and receive the stress wave signal generated by the vibration excitation, so as to detect the internal damage of the bridge pier. Under the control of the signal processing system, the circumferential driver 12 and the vertical driver 19 can not only control the opening and closing time of the two, but also control the working frequency of the two to change the output power of the two, adjust the running speed of the vertical walking support 2 and the rotator 5, and can also accurately control the opening and closing time of the two, so as to control the walking path length or the number of rotations of the two, and the image acquisition device 7 can realize the opening and closing of the device under the control of the signal processing system, and obtain the collected image data.
[0119] As a further implementation method, the signal processing system can be arranged in the vertical traveling support 2, the circular track 4 or the equipment bin 21, so that it can be controlled by remote signals; it can also be a remote control system arranged at the far end, which is connected to the corresponding electrical device by wireless or wired means; it can also be composed of a proximal control system and a remote control system, wherein the proximal control system can be arranged in the vertical traveling support 2, the circular track 4 or the equipment bin 21, and the remote control system can be arranged in the corresponding remote control area, and the two can realize data intercommunication by wireless or wired means, so that the remote control system can receive the corresponding data of the proximal control system, and the proximal control system can operate autonomously or under the instruction of the remote control system, and send the corresponding operation data to the remote control system for processing. Based on the prior art, there are many types of control systems that can realize the above functions, such as PLC control systems. Therefore, the specific architecture, control method, principle, etc. of the control system will not be redundantly described in this embodiment. Similarly, for the above signal processing and electrical device control, there are many principles, architectures and implementation methods of signal processing systems in the prior art. This embodiment can refer to the prior art to achieve its functions and effects, and thus this embodiment will not be further described.
[0120] The above is the full content of the multi-source data fusion bridge pier impact damage detection device. For better understanding and implementation, please refer to Figure 7 Based on the above-mentioned multi-source data fusion bridge pier impact damage detection device, the present invention also provides a multi-source data fusion bridge pier impact damage detection method, which includes the following steps:
[0121] S1: A plurality of piezoelectric sensors 1 with a positive piezoelectric effect are embedded in the pier, and the piezoelectric sensors 1 are distributed in an array inside the pier to form a sensor array; when the pier is hit, the coordinate information data of all the piezoelectric sensors 1 that generate electrical signals are obtained, and the hit area of the pier is determined based on the coordinate information data.
[0122] After the bridge pier is hit, the impact area is the most severely damaged. Therefore, before conducting damage detection, we need to determine the distribution of the impact area of the bridge pier in order to provide coordinate guidance for the detection and achieve more accurate detection.
[0123] Based on this, when the bridge pier is hit, the following method is used in this step to determine the impact area of the bridge pier. The specific determination method is:
[0124] S11: pre-embed a plurality of piezoelectric sensors 1 having a positive piezoelectric effect in the bridge pier, wherein the plurality of piezoelectric sensors are distributed in an array inside the bridge pier to form a sensor array.
[0125] In application, if the pier is a circular pier, the piezoelectric sensor can be arranged at a certain distance in the circumferential, vertical and height directions of the pier, such as arranging multiple circles in a ring from the outside to the inside along the cross section of the pier, with several piezoelectric sensors arranged in each circle, and arranging multiple groups at intervals along the height direction of the pier to form the array arrangement; similarly, if the pier is a square pier, the piezoelectric sensor can be arranged along the side length direction, vertical and height directions, such as evenly distributing multiple piezoelectric sensors in a matrix form in each cross section of the pier, and arranging multiple groups at intervals along the height direction of the pier to form the array arrangement, for example Figure 7 shown.
[0126] S12 takes any position of the bridge pier as the coordinate origin, numbers the piezoelectric sensors 1, obtains the coordinate information of each piezoelectric sensor 1, and associates the number and coordinate information of each piezoelectric sensor 1.
[0127] When applied, in order to better determine the coordinate position of the area and confirm the subsequent crack position, in this step, the bridge pier can be used as a reference object, and any position on it can be used as the coordinate origin, so that after each piezoelectric sensor is set, it corresponds to unique three-dimensional coordinate information (such as longitudinal x, lateral y and vertical z). Furthermore, after each piezoelectric sensor is numbered, the coordinate information corresponding to each numbered piezoelectric sensor is also unique. The corresponding number is associated with the corresponding coordinate information, so that the position of all piezoelectric sensors 1 on the bridge pier and their positional relationship can be determined based on the number and coordinate information, which is convenient for subsequent application.
[0128] Specifically, a three-dimensional coordinate system can be established with the center position of the top of the pier as the coordinate origin, the longitudinal direction of the bridge as the X-axis, the transverse direction of the bridge as the Y-axis, and the vertical direction of the pier as the Z-axis. Therefore, according to the position where the piezoelectric sensor 1 is arranged, each piezoelectric sensor 1 can obtain unique coordinate information. Then, when numbering, each piezoelectric sensor 1 in the piezoelectric sensors 1 forming the array is numbered according to the priority order of the XYZ coordinate axes and the principle of coordinates from small to large. Then, each piezoelectric sensor 1 corresponds to a unique number and displacement coordinates. For example, 8 piezoelectric sensors 1 are arranged in the pier, and their coordinates are (0,1,0), (0,2,0), (1,1,0), (1, 2,0), (0,1,1), (0,2,1), (1,1,1), (1,2,1), then after the numbering is associated with the coordinates, the relationship information can be PZT-111 (0,1,0), PZT-121 (0,2,0), PZT-211 (1,1,0), PZT-221 (1,2,0), PZT-112 (0,1,1), PZT-122 (0,2,1), PZT-212 (1,1,1), PZT-222 (1,2,1), where PZT-111 to PZT-222 are the corresponding numbers of piezoelectric sensor 1, followed by their corresponding coordinates.
[0129] In this step, the serial number and coordinate association information of the piezoelectric sensor 1 can be stored in the signal processing system so that the signal processing system can retrieve and use it at any time.
[0130] S13 When the bridge pier is hit, the number of the piezoelectric sensor 1 that generates the electrical signal is obtained, and the coordinate information of the corresponding area of the piezoelectric sensor 1 corresponding to all the numbers is obtained. The bridge pier area involved in the coordinate information is determined as the bridge pier hit area.
[0131] During application, since the piezoelectric sensor 1 has a positive piezoelectric effect, the piezoelectric sensor 1 will generate an electrical signal after being subjected to impact pressure. The electrical signal of the piezoelectric sensor 1 will not change in the area that is not impacted. Therefore, the number of the piezoelectric sensor 1 that generates the electrical signal can be obtained, and the coordinate information of the corresponding piezoelectric sensor 1 can be obtained through the corresponding number, so that the area enclosed by the coordinate information of the corresponding piezoelectric sensor 1 can be determined as the impact area of the bridge pier.
[0132] In this step, the number and coordinate information of the piezoelectric sensor 1 can be obtained through the signal processing system and the impacted area of the bridge pier can be finally confirmed. Specifically, the signal processing system can store the number and coordinate information of all piezoelectric sensors 1. When the number of the piezoelectric sensor 1 that generates the electrical signal is obtained, the corresponding coordinate information can be obtained by calling the corresponding data, and the bridge pier area involved in all the coordinates can be determined as the impacted area of the bridge pier through data processing.
[0133] In actual application, since the purpose of identifying the impacted area is to subsequently detect the degree of damage on the surface of the pier, and the image acquisition device 7 can acquire circumferential images of the pier, it is only necessary to obtain the coordinate information of the surface piezoelectric sensor 1, so that the coordinates corresponding to the surface piezoelectric sensor 1 and all areas within its transverse cross section can be identified as the impacted area. Therefore, when acquiring the coordinate information, it is only necessary to determine the height coordinate of the impacted area.
[0134] Furthermore, step S13 can be further performed in the following manner:
[0135] S131 sets a corresponding electric signal threshold, and when the bridge pier is hit, obtains the serial number of the piezoelectric sensor 1 whose electric signal peak value is greater than the electric signal threshold.
[0136] In this step, the electrical signal threshold may be a smaller value greater than 0, such as 0.1V.
[0137] In practical applications, since only the height coordinates of the impacted area need to be determined, when the bridge pier is impacted, only the number of the piezoelectric sensor 1 on the outermost layer of the bridge pier and whose electrical signal peak value is greater than the electrical signal threshold is obtained, thereby reducing the amount of data processing.
[0138] This step can also be implemented through a signal processing system, that is, the signal processing system has a built-in electrical signal threshold. After obtaining the corresponding data, it can directly compare the data with the electrical signal threshold to obtain the piezoelectric sensor 1 number that meets the conditions.
[0139] S132 extracts the height coordinate information of all piezoelectric sensors 1 whose electrical signal peaks exceed the electrical signal threshold, traverses the height coordinate information of all piezoelectric sensors 1, and obtains the maximum height and the minimum height in the height coordinate information;
[0140] In this step, after the height coordinate information of the corresponding piezoelectric sensor 1 is obtained and extracted, only the corresponding height coordinate information, that is, the height direction coordinate of the pier in the three-dimensional coordinate is extracted, and the values of all height coordinates are compared to obtain the maximum height and minimum height in the height coordinate information.
[0141] S133 determines the bridge pier surface area intercepted between the maximum height and the minimum height as the bridge pier impact area.
[0142] In this step, after the maximum height and the minimum height are determined, it means that the bridge piers between these two coordinates are all impacted. Therefore, the surface area of the bridge pier intercepted between the maximum height and the minimum height is determined as the impact area of the bridge pier. The distribution position of the impact area of the bridge pier can be accurately found, and the highest point (corresponding to the maximum height coordinate position) and the lowest point (corresponding to the minimum height coordinate position) of the corresponding bridge pier impact area can be obtained through the height coordinate information of the piezoelectric sensor 1.
[0143] S2 sets the sensors in the sensor array inside the entire pier as the excitation sensors of the next layer of sensors in the column where the sensor is located, and the sensors in the sensor array are the receiving sensors of the previous layer of sensors in the column where the sensor is located; the excitation sensors are used to emit vibration excitation, and the receiving sensors are used to receive stress wave signals formed inside the pier based on the vibration excitation; when the pier is hit, the excitation sensors and receiving sensors in the corresponding sensor array generate vibration excitation and receive corresponding signals, so as to obtain stress wave data.
[0144] When applied, such as Figure 8 , Fig. 9 As shown, the piezoelectric sensors are layered along the Z coordinate of the pier and also along the vertical or height direction of the pier, and then each layer corresponds to the piezoelectric sensor corresponding to the next layer, so that each column of sensors formed in this way can correspond to each other up and down to form a pair of special sensors, one of which serves as a signal receiving device to receive stress wave signals, and at the same time it also serves as an excitation source for the next layer of sensors (formed by the vibration of the piezoelectric sensor when it is powered on) to emit vibration excitation to form a corresponding stress wave signal inside the pier, and so on to configure into a signal receiving and sending group, and then when the pier is hit, the stress wave data inside the pier can be collected in this way. The reason why the present invention performs pairing according to this principle is that under the action of the impact load, the pier produces cracks that are mainly perpendicular to the axis of the pier, and the piezoelectric sensors of the upper and lower layers can easily determine whether there are cracks and detect the depth of the cracks.
[0145] For example, when the piezoelectric sensor 1 is arranged in the manner of this step, when numbering and coordinates are used, PZT-111 (0,1,0) corresponds to PZT-112 (0,1,1), PZT-121 (0,2,0) corresponds to PZT-122 (0,2,1), PZT-211 (1,1,0) corresponds to PZT-212 (1,1,1), and PZT-221 (1,2,0) corresponds to PZT-222 (1,2,1), where PZT-XXX is the sensor number, and the following is its three-dimensional coordinate information. In this way, the sensors on the same column can achieve the above pairing, and it is also more convenient in subsequent data processing. For example, in this step, the corresponding sensor number and coordinate information data can be directly stored in the signal processing system, and the signal processing system can directly retrieve the sensor coordinate information. Through its coordinate information, all sensors can be arranged in columns, so that they can be directly used.
[0146] It should be noted that the sensor pairing method of this step can also be extended to more sensor pairings.
[0147] At the same time, after the signal processing system completes the pairing of the piezoelectric sensors, when the pier is intact, the piezoelectric sensor with a smaller (or larger) Z coordinate emits excitation, and the piezoelectric sensor with a larger (or smaller) Z coordinate receives the stress wave. By collecting the stress wave several times, sufficient stress wave data is obtained and stored in the signal processing system. Through deep learning algorithms, such as long short-term memory networks (LSTM), recursive neural networks (RNN), convolutional neural networks (CNN), etc., the stress wave data of the pier in the intact state is reconstructed, and a stress wave data prediction model is established and stored in the signal processing system, which can be conveniently used for comparison of stress wave data collected when the pier is damaged in the future.
[0148] S3 processes the acquired stress wave data and compares it with the stress wave data of the intact pier to obtain the internal damage index of the pier after impact.
[0149] In specific applications, this step can be performed as follows:
[0150] S31 When the bridge pier is hit, n2 stress wave time domain data are obtained, and the n2 stress wave data are sorted from small to large according to the stress wave peak value. After sorting, the stress wave data are processed to obtain the total energy of the stress wave time domain data, and the n2 stress wave data when the bridge pier is intact are obtained and processed in the same way.
[0151] In this step, after the bridge pier is hit, the stress wave time domain signal is cyclically collected through the cooperation of the two pairs of piezoelectric sensors 1 in the upper and lower layers, and the total energy of the collected stress wave signal is calculated (through the signal processing system). The total energy of the stress wave time domain signal can be realized by Laplace transform or wavelet packet decomposition. For example, wavelet packet decomposition is specifically used in this step, which can provide higher resolution and better localization capability, and has good applicability to both high-frequency and low-frequency signals. Finally, the transformed total energy data can be sorted according to the order of the stress wave time domain peak value.
[0152] For better understanding, the following example is given:
[0153] Taking the collection of 10 stress waves as an example, the stress wave time domain data reconstructed when the pier is intact is first processed. In order to minimize the impact of the randomness of the stress wave, multiple stress wave time histories can be obtained as samples, 10 stress waves are obtained and the stress wave time domain peak values are extracted. The peak values of the stress wave data are sorted from small to large, in the order of SWT j,1 , SWT j,2 , ..., SWT j,10 ; By performing wavelet packet decomposition on the 10 stress wave time domain data reconstructed when the pier is intact, the total energy data is obtained, and the total energy is sorted from small to large according to the order corresponding to the stress wave time domain peak value, which is E j,1 , E j,2 ,……,E j,10 When the bridge pier is detected to be hit, the piezoelectric sensor is used to collect stress wave data again to obtain 10 stress waves, which are sorted from small to large according to the time domain peak value of the stress wave, in order of CSWT j,1 ,CSWT j,2 , ..., CSWT j,10 ; The stress wave time domain data is decomposed by wavelet packet to obtain its total energy data, and the total energy is sorted from small to large according to the order corresponding to the stress wave time domain peak value, which is CE j,1 , C.E. j,2 ,……,CE j,10 .
[0154] It should be noted that the total energy acquisition and sorting of stress wave data in this step can be completed by the signal processing system.
[0155] S32 When the bridge pier is damaged, the cracks will cause the time domain amplitude and total energy of the stress wave to decrease. Therefore, in this step, two adjacent piezoelectric sensors in the same column are taken as a pair, and the damage index DII of the j-th pair of piezoelectric sensors is set as j for:
[0156] ①
[0157] In formula ①, since the time domain amplitude and total energy of the stress wave after the pier is damaged will decay, the larger its value is, the more serious the damage to the pier is; n2 is the number of stress waves generated; SWT is j,i is the time domain peak value of the stress wave with order i generated by the jth pair of piezoelectric sensors when the pier is intact, CSWT j,i is the time domain peak value of the stress wave of the jth pair of piezoelectric sensors ranked i after the pier is impacted, E j,i is the total energy of stress waves generated by the jth pair of piezoelectric sensors ranked i when the pier is intact; CE j,i is the total energy of the stress wave with order i generated by the jth pair of piezoelectric sensors after the bridge pier is impacted.
[0158] S33 Obtain the crack depth CD corresponding to the jth pair of piezoelectric sensors in the pier j .
[0159] In this step, for the crack depth CD j The acquisition of can also be achieved through the signal processing system:
[0160] In this step, when 0<DII j ≤1, obtain the crack depth CD corresponding to the jth pair of piezoelectric sensors in the pier j ; When DII j When ≤0, it is considered that the damage to the bridge pier is very small, and it is almost like no cracks occur, then CD j =0. The reason why 0<DII is obtained j The coordinates of the piezoelectric sensor are ≤1 because: when DII j When ≤0, it is considered that the time domain peak value and total energy of stress waves of the intact pier and the pier after impact are basically consistent, and it is considered that the damage to the pier is very small and almost no cracks are generated.
[0161] In this step, the vertical or longitudinal position of the crack in the pier is determined by the value of the piezoelectric sensor at the Z coordinate of the pier, and the depth CD of the crack in the pier can be determined by the X or Y coordinate. j Specifically, since the generated cracks will cause the amplitude of the stress wave to decrease, when 0<DII corresponding to a pair of piezoelectric sensors is obtained j When ≤1, it means that cracks have appeared in the area detected by the pair of piezoelectric sensors, and then the DII of another pair of piezoelectric sensors located further inside the pier is obtained based on this. j , if it is still 0<DII j ≤1, it means that cracks also appear in the area detected by the pair of piezoelectric sensors. If the DII of another pair of piezoelectric sensors located further inside the pier is obtained, j=0, it means that there is no crack in the area detected by the pair of piezoelectric sensors. Therefore, the depth of the crack in the pier can be obtained based on the coordinates of the above piezoelectric sensors, so as to determine the vertical or longitudinal position of the crack in the pier and its specific depth.
[0162] S34 is based on the crack depth CD obtained j , and according to the DII obtained for each pair of piezoelectric sensors j The weighted internal damage index DII of the bridge pier after impact is obtained:
[0163] ②
[0164] In formula ②, CD max is the maximum crack depth determined by n3 pairs of piezoelectric sensors, n3 is the number of all piezoelectric sensor pairs in the entire pier, k is 0<DII j Number of piezoelectric sensor pairs ≤ 1.
[0165] S4 simultaneously collects crack images on the surface of the bridge piers in the area, and processes the collected results to obtain surface damage indicators of the bridge piers after impact.
[0166] When this step is applied, based on the above-mentioned bridge pier impact damage detection device, rapid data processing and collection can be achieved. Specifically, the surface crack detection device is installed on the bridge pier, and its action can be controlled by the signal processing system to achieve crack image collection, and the collected crack image data can be processed to obtain the surface damage index of the bridge pier after impact. The specific method steps are as follows:
[0167] S41, based on the determined impact area of the bridge pier, determining the initial acquisition position and the acquisition end position of the image acquisition device 7 in the height direction of the bridge pier;
[0168] In this step, when the impact position of the bridge pier is determined, the coordinates of the highest point and the lowest point of the impact area of the bridge pier are also obtained at the same time. Therefore, when surface crack damage detection is required, it is only necessary to collect the bridge pier surface between the highest point and the lowest point of the impact area of the bridge pier. Furthermore, in this step, the maximum height and the minimum height in step S133 can be used as the collection limit positions. Specifically:
[0169] The height coordinate position corresponding to the maximum height in the impact area of the bridge pier is used as the initial acquisition position;
[0170] The height coordinate position corresponding to the minimum height value in the impact area of the pier is taken as the end position of the acquisition.
[0171] Therefore, the image acquisition device 7 only needs to start acquiring images from the initial acquisition position and end the acquisition operation after reaching the acquisition end position.
[0172] S42 controls the vertical walking support 2 to move, moves the image acquisition device 7 to the initial acquisition position, and moves along the designed spacing. The annular drive system 6 drives the rotator 5 to rotate at a constant speed. The image acquisition device 7 synchronously performs annular area image acquisition. After completion, it moves with the same designed spacing and performs the next annular area image acquisition, until the image acquisition device 7 moves to the acquisition end position and completes the crack image acquisition.
[0173] In this step, since the image acquisition device 7 has a certain acquisition range, the maximum height difference of the acquisition area of the image acquisition device 7 in the height direction can be set to X, and the design spacing can be any value less than or equal to X, thereby ensuring the image acquisition of the entire pier impact area.
[0174] As one of the specific collection methods, the maximum height h corresponding to the initial collection position and the collection end position can be obtained. max and the minimum height h min , assuming that the initial speed position of the image acquisition device 7 is h1, then the total distance that the image acquisition device 7 needs to move can be calculated as h min -h1, when collection is required, the vertical driver 19 is turned on to drive the vertical walking support 2 to move to the initial collection position, that is, the maximum height h max The height coordinate position is at which the bridge pier reaches the top of the impact area. After that, the vertical driver 19 continues to be turned on until it reaches the first acquisition area after walking the designed spacing. At this time, the vertical driver 19 stops, and the vertical walking bracket 2 and the image acquisition device 7 stop synchronously at this height position. At this time, the annular driver 12 is turned on to drive the rotator 5 to rotate, and the image acquisition device 7 is turned on synchronously. Then, the image acquisition device 7 moves with the rotator 5 to perform annular image acquisition on the first acquisition area. When the image acquisition device 7 rotates back to the initial position, it means that the image acquisition of the first acquisition area is completed. At this time, the annular driver 12 is turned off, and the vertical driver 19 is turned on again to continue walking the designed spacing and reach the second acquisition area. At this time, the vertical driver 19 is turned off, and the annular driver 12 is turned on again. The image acquisition device 7 performs annular image acquisition on the second acquisition area, and this is repeated until the total distance moved by the image acquisition device 7 is h. min -h1, the image acquisition device 7 is synchronously located at the minimum height h min position, it means that the vertical walking device has reached the specified position, and the image collection of the entire bridge pier impact area is completed. At this time, the vertical walking support 2 can be moved back to the initial position.
[0175] It should be noted that in the above acquisition method, the design spacing can be arbitrarily determined according to the shooting range of the image acquisition device 7, or it can be based on the maximum height hmax and the minimum height h min To obtain the value, for example, the following formula can be used:
[0176] H = (h max -h min ) / n
[0177] Where H is the design spacing, and n is the number of acquisitions by the image acquisition device.
[0178] Furthermore, based on the above formula, the design spacing can be set as needed, but it should be noted that the final H value also needs to be less than or equal to X (X is the maximum height difference of the acquisition area of the above-mentioned image acquisition device 7 in the height direction) to better ensure the comprehensive acquisition of the impact area of the bridge pier.
[0179] S43 constructing a convolutional neural network model, establishing a sample set of bridge pier images with cracks, training the convolutional neural network model, and obtaining a crack recognition prediction model;
[0180] When this step is applied, the specific steps are:
[0181] S4311 Select the convolutional neural network model as the base model;
[0182] S4312 obtaining crack images of bridge piers or other concrete structures, and processing and marking crack information in the crack images, wherein the crack information at least includes the number of cracks, crack length, and crack width;
[0183] S4313 performs image processing on the annotated crack image and inputs it into the convolutional neural network model to train it to obtain the convolutional neural network model.
[0184] When implementing steps S4311 to S4313, as an optional method, a convolutional neural network is used to identify cracks in the image. First, a large number of image samples are required as an image data set. The target in the sample can be a bridge pier or other concrete structure, and the number, length, width and other information of different types of cracks in the bridge pier or other concrete structure are collected. The relevant information can be obtained through open source data on the Internet, laboratory data or engineering practice data, and the data is processed and labeled, especially the number of cracks, crack width and length information, and then the acquired image is cropped, scaled, normalized and other image preprocessing operations are performed to meet the input requirements of the deep neural network. For the crack width and length recognition task in the present invention, as an optional solution, a convolutional neural network (CNN) can be selected as the basic architecture, depending on the excellent performance of CNN in the field of image recognition; specifically, the CNN model can be selected according to the complexity of the image recognition task and the size of the data set, such as AlexNet, ResNet, VGG, SegNet, etc. As an optional approach, the architecture of the CNN model can be adjusted according to the task requirements, such as increasing or decreasing the number of convolutional layers, pooling layers, and fully connected layers; after selecting the convolutional neural network model, the model is trained in combination with the labeled image data set, and the mean square error loss function is used to evaluate the performance of the model, and the weights and biases of the model are updated through the forward propagation and back propagation algorithms, and the hyperparameters such as the learning rate, batch size, and number of iterations are adjusted to optimize the training effect of the model. After the crack recognition convolutional neural network model training is completed, the bridge pier damage pictures collected by the image acquisition device can be identified to obtain the width and length of the cracks.
[0185] S44: inputting the collected crack images into the crack recognition prediction model, and obtaining the number of cracks, crack widths, and crack lengths in all images based on the crack recognition prediction model;
[0186] S45 Based on the obtained crack number, crack width and crack length, the crack width damage coefficient WDIC is defined as:
[0187]
[0188] Where WC i is the width of the ith crack; n1 is the number of cracks; I-VII represent different environmental condition categories, among which Class I is general environment, Class II is freeze-thaw environment, Class III is offshore or marine chloride environment, Class IV is other chloride environment such as deicing salt, Class V is salt crystallization environment, Class VI is chemical corrosion environment, and Class VII is corrosion environment.
[0189] It should be noted that for specific environmental condition category standards, please refer to the standard: "Highway Reinforced Concrete and Prestressed Concrete Bridge and Culvert Design Specifications" (JTG3362-2018), which will not be further described in this embodiment.
[0190] The length damage coefficient LDIC of the crack is defined simultaneously as:
[0191]
[0192] Where WL i is the length of the i-th crack; n1 is the number of cracks;
[0193] The comprehensive damage coefficient DIC of cracks on the pier surface is defined as:
[0194]
[0195] Where W W and W L is the weight coefficient of the damage index of crack width and crack length;
[0196] In this step, it should be noted that W W and W L is the weight coefficient of the crack width and crack length damage indicators, and the weight coefficient can be determined by expert scoring method or hierarchical analysis method.
[0197] As a solution for determining the weight coefficient, the present invention adopts an expert scoring method to determine the weight coefficient. Several experts with rich experience in the industry can be selected to determine the weight coefficient according to W. W and W L Scoring, the scoring restriction rule is W W and W L If the sum is 1, the average of the scores given by several experts is taken as the weight coefficient.
[0198] S5 calculates and processes the internal damage index and surface damage index to obtain the overall comprehensive damage index of the pier.
[0199] In this step, based on the internal damage index and surface damage index obtained, the following formula is established:
[0200] ③
[0201] In formula ③, W DII and W DIC are weight coefficients, and their sum is 1; D is the overall comprehensive damage index of the pier, DII is the internal damage index, and DIC is the surface damage index;
[0202] Based on formula ③, the overall comprehensive damage index of the pier can be obtained.
[0203] In this step, the weight coefficient W DII and W DIC It can be selected according to the situation. It reflects the importance of internal damage and surface damage. If the sum of the two is 1, the value of the overall comprehensive damage index of the pier is between 0 and 1. The weight coefficient can be determined by expert scoring method or hierarchical analysis method.
[0204] As a way to determine the weight coefficient W DII and W DIC The present invention adopts the expert scoring method to determine the weight coefficient W DII and W DIC , you can select several experienced experts in the industry to DII and W DIC Scoring, finally select several experts to DII and W DIC The average of the scores is used as the weight coefficient.
[0205] S6 Rates the degree of damage to the bridge piers after the impact based on the numerical range of the overall comprehensive damage index of the bridge piers.
[0206] After rating the overall comprehensive damage index of the bridge pier in this step, the damage condition of the corresponding bridge pier can be determined based on its grade, and corresponding remedial measures can be implemented according to the rating results.
[0207] In actual application, for the obtained overall comprehensive damage index value of the bridge pier, the rating range and rating level can be determined according to the actual situation.
[0208] Specifically, how to rate the pier based on the overall comprehensive damage index can be referred to the following table:
[0209] Table 1: Bridge pier damage rating table
[0210]
[0211] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.
Claims
1. A bridge pier impact damage detection method based on multi-source data fusion, characterized in that: The following steps are involved: A plurality of piezoelectric sensors with a positive piezoelectric effect are embedded in the bridge pier, and the plurality of piezoelectric sensors are distributed in an array inside the bridge pier to form a sensor array; When the bridge pier is hit, the coordinate information data of all piezoelectric sensors that generate electrical signals are obtained, and the hit area of the bridge pier is determined based on the coordinate information data; The sensors in the sensor array in the entire pier are set as the excitation sensors of the next layer of sensors in the row where the sensors are located, and the sensors in the sensor array are the receiving sensors of the previous layer of sensors in the row where the sensors are located; the excitation sensors are used to emit vibration excitation, and the receiving sensors are used to receive stress wave signals formed inside the pier based on the vibration excitation; when the pier is hit, the excitation sensors and receiving sensors in the corresponding sensor array generate vibration excitation and receive corresponding signals, so as to obtain stress wave data; After data processing, the acquired stress wave data is compared with the stress wave data of the intact bridge pier to obtain the internal damage index of the bridge pier after impact; At the same time, the crack images of the bridge pier surface in the area are collected, and the collected results are processed to obtain the surface damage index of the bridge pier after the impact; Based on the internal damage index and surface damage index, the data of the two are calculated and processed to obtain the overall comprehensive damage index of the bridge pier. Based on the numerical range of the overall comprehensive damage index of the bridge pier, the damage degree of the bridge pier after the collision was rated; Among them, the acquired stress wave data is processed and compared with the stress wave data of the intact pier to obtain the internal damage index of the pier after impact; specifically: When the bridge pier is hit, n2 stress wave time domain data are obtained, and the obtained n2 stress wave data are sorted from small to large according to the stress wave peak value. After sorting, the stress wave data are processed to obtain the total energy of the stress wave time domain data, and the n2 stress wave data when the bridge pier is intact are obtained and processed in the same way; Taking two adjacent piezoelectric sensors in the same column as a pair, let the damage index DII of the j-th pair of piezoelectric sensors be j for: ① In formula ①, n2 is the number of stress waves generated; SWT j,i is the time domain peak value of the stress wave ranked i generated by the jth pair of piezoelectric sensors when the pier is intact; CSWT j,i is the time domain peak value of the stress wave ranked i generated by the jth pair of piezoelectric sensors after the bridge pier is impacted; E j,i is the total energy of stress waves generated by the jth pair of piezoelectric sensors ranked i when the pier is intact; CE j,i is the total energy of stress waves ranked i generated by the jth pair of piezoelectric sensors after the bridge pier is impacted; Get the crack depth CD corresponding to the jth pair of piezoelectric sensors in the pier j :When 0<DII j ≤1, obtain the crack depth CD corresponding to the jth pair of piezoelectric sensors in the pier j ; Among them, when the damage index DIIj of the j-th pair of piezoelectric sensors is ≤0, the crack depth CDj=0, and at this time, it is not necessary to obtain the crack depth CDj of the corresponding piezoelectric sensor; Based on the obtained crack depth CD j , and according to the DII obtained for each pair of piezoelectric sensors j The weighted internal damage index DII of the bridge pier after impact is obtained: ② In formula ②, CD max is the maximum crack depth determined by n3 pairs of piezoelectric sensors, n3 is the number of all piezoelectric sensor pairs in the entire pier, k is 0<DII j Number of piezoelectric sensor pairs ≤ 1.
2. The method for detecting bridge pier impact damage according to claim 1, characterized in that: When the bridge pier is hit, the coordinate information data of all piezoelectric sensors that generate electrical signals are obtained, and the impact area of the bridge pier is determined based on the coordinate information data, specifically: Taking any position of the bridge pier as the coordinate origin, numbering the piezoelectric sensors, obtaining the coordinate information of each piezoelectric sensor, and associating the number and coordinate information of each piezoelectric sensor; When the bridge pier is hit, the number of the piezoelectric sensor that generates the electrical signal is obtained, and the coordinate information of the corresponding area of the piezoelectric sensor corresponding to all the numbers is obtained, and the bridge pier area involved in the coordinate information is determined as the impacted area of the bridge pier.
3. The method for detecting bridge pier impact damage according to claim 2, characterized in that: When the bridge pier is hit, the number of the piezoelectric sensor that generates the electrical signal is obtained, and the coordinate information of the corresponding area of the piezoelectric sensor corresponding to all the numbers is obtained. The bridge pier area involved in the coordinate information is determined as the bridge pier hit area, specifically: A corresponding electric signal threshold is set, and when the bridge pier is hit, the number of the piezoelectric sensor whose electric signal peak value is greater than the electric signal threshold is obtained; Extract the height coordinate information of all piezoelectric sensors whose electrical signal peaks exceed the electrical signal threshold, traverse the height coordinate information of all piezoelectric sensors, and obtain the maximum height and the minimum height in the height coordinate information; The pier surface area intercepted between the maximum height and the minimum height is determined as the pier impact area.
4. The method for detecting bridge pier impact damage according to claim 3, characterized in that: At the same time, the crack images of the bridge pier surface in the area are collected, and the collected results are processed to obtain the surface damage index of the bridge pier after the impact; specifically: Based on the determined impact area of the bridge pier, determine the initial acquisition position and the acquisition end position of the image acquisition device in the height direction of the bridge pier; The vertical traveling support is controlled to move, the image acquisition device is moved to the initial acquisition position, and the image acquisition device moves to the designed spacing. The annular drive system drives the rotator to rotate at a constant speed. The image acquisition device synchronously acquires images of the annular area. After completion, it moves with the same designed spacing and acquires images of the next annular area, until the image acquisition device moves to the acquisition end position and completes the crack image acquisition; Constructing a convolutional neural network model, establishing a sample set of bridge pier images with cracks, training the convolutional neural network model, and obtaining a crack recognition prediction model; Inputting the collected crack images into the crack recognition prediction model, and obtaining the number of cracks, crack width and crack length in all images based on the crack recognition prediction model; Based on the obtained crack number, crack width and crack length, the crack width damage coefficient WDIC is defined as: Where WC i is the width of the i-th crack; n1 is the number of cracks; I-VII represent different environmental condition categories, among which Class I is general environment, Class II is freeze-thaw environment, Class III is offshore or marine chloride environment, Class IV is other chloride environment such as deicing salt, Class V is salt crystallization environment, Class VI is chemical corrosion environment, and Class VII is corrosive environment; The length damage coefficient LDIC of the crack is defined simultaneously as: Where WL i is the length of the i-th crack; n1 is the number of cracks; The surface damage index DIC of the pier surface is defined as: Where W W and W L is the weight coefficient of the crack width and crack length damage indicators, the sum of the two is 1, and the weight coefficient is determined by expert scoring method or hierarchical analysis method.
5. The method for detecting impact damage to bridge piers according to claim 1, characterized in that: Based on the internal damage index and surface damage index, the data of the two are calculated and processed to obtain the overall comprehensive damage index of the pier, which is: Based on the obtained internal damage index and surface damage index, the following formula is established: ③ In formula ③, W DII and W DIC are weight coefficients, the sum of which is 1. The weight coefficients are determined by expert scoring method or hierarchical analysis method; D is the overall comprehensive damage index of the pier, DII is the internal damage index, and DIC is the surface damage index; The overall comprehensive damage index of the pier can be obtained based on formula ③.
Citation Information
Patent Citations
Underwater building quality detection device
CN116626054A
Bridge deformation monitoring method and monitoring system
CN116910423A
Concrete structure defect monitoring system and method based on piezoelectric intelligent aggregate
CN118549632A