Bridge crack automatic detection method

By setting up an acoustic conduction feature detection module and a load acquisition device on the bridge, using the discrete difference of acoustic conduction feature to divide abnormal categories and adjust the detection parameters, the computing power consumption problem caused by massive data in real-time detection of large-area bridges is solved, and the calculation power consumption caused by massive data is achieved. Efficient and accurate crack detection.

CN118916810APending Publication Date: 2024-11-08河南易跃建设工程有限公司 +2
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202410970154.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-19
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

When the existing technology uses the Internet of Things technology to detect gaps in large-area bridges in real time, massive data collection leads to high computing power consumption, making it difficult to achieve efficient detection.

Method used

Determine the detection area along the equal spacing of the bridge, set up a sound conduction feature detection module and a load acquisition device, divide the crack abnormal tendency categories through the discrete difference of the sound conduction feature, adjust the detection module parameters, cluster and fit the acoustic conduction feature curve to determine Exception, reducing computing power consumption.

Benefits of technology

It improves the accuracy and data reliability of bridge crack detection, saves computing resources, and is suitable for real-time detection of large-area bridges.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118916810A_ABST
    Figure CN118916810A_ABST
Patent Text Reader

Abstract

The invention relates to the field of bridge crack detection, in particular to an automatic bridge crack detection method, which comprises the following steps of: determining a plurality of detection areas at equal intervals along a bridge, and arranging a detection module for acquiring sound conduction characteristics and a load acquisition device for acquiring borne loads in each detection area; based on the current load of the detection area, calling a plurality of corresponding sound conduction characteristics under the same load from a historical database, and comparing the current sound conduction characteristics with the called sound conduction characteristics to obtain the discrete difference degree of the sound conduction characteristics so as to divide the crack anomaly tendency category of the detection area; the detection parameters of the detection module are adaptively adjusted and the abnormal condition of the detection area is analyzed, so that when the gap of the large-area bridge is detected in real time, the detection accuracy is improved and the computing power resource is saved on the premise of ensuring the reliability of the detection data.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of bridge crack detection, and in particular to an automatic bridge crack detection method. Background Art

[0002] As an important part of the highway transportation system, bridges are the foundation of social operation and the support of economic development. With the increase of bridge service life and the influence of external environment, the problem of bridge cracks follows. In order to ensure the safety of bridge structure and extend the service life of bridge, it is particularly important to conduct regular inspection and maintenance of bridge cracks.

[0003] Chinese patent publication number: CN111402227A, discloses a bridge crack detection method, including the following steps: crack segmentation of a group of collected bridge images; detection and classification of bridge cracks using a pre-built bridge crack classification model based on the crack segmentation results; segmentation of bridge cracks using an improved GAC algorithm model to segment visible cracks in the bridge bottom image captured by a high-definition camera of a drone; construction of the bridge crack classification model using a deep learning method, designing a deep convolutional neural network model for bridge identification; three-dimensional reconstruction of bridge cracks and crack information detection using a moving cube algorithm to determine the number of cracks, average width, geometric properties and their spatial relationship with the whole, so that professionals can perform qualitative or quantitative analysis of the cracks.

[0004] However, the existing technology still has the following problems: if the Internet of Things technology is used to detect gaps in large-area bridges in real time, the relevant feature data required to be collected is massive. Under high data collection density, the real-time analysis method will be accompanied by a large amount of data analysis and processing, and the computing power consumption is high. Summary of the invention

[0005] To this end, the present invention provides an automatic detection method for bridge cracks to overcome the problem in the prior art that, when the Internet of Things technology is used to detect cracks in a large area of ​​bridges in real time, the relevant characteristic data required to be collected is massive. Under a high data collection density, the real-time analysis method is used, which will be accompanied by a large amount of analysis and processing of the data and high computing power consumption.

[0006] To achieve the above object, the present invention provides a method for automatically detecting bridge cracks, which comprises:

[0007] Step S1, determining a number of detection areas at equal intervals along the bridge, and setting a detection module for collecting acoustic conduction characteristics and a load collection device for collecting the load in each of the detection areas, wherein the acoustic conduction characteristics include vibration frequency and vibration amplitude;

[0008] Step S2, based on the current load on the detection area, calling a number of acoustic conduction features corresponding to the same load from the historical database, comparing the current acoustic conduction feature with the called acoustic conduction feature, and obtaining the discrete difference of the acoustic conduction feature to classify the abnormal crack tendency category of the detection area;

[0009] Step S3, adjusting the detection parameters of the detection module according to the abnormal crack tendency category corresponding to the detection area and analyzing the abnormal situation of the detection area, including:

[0010] Adjusting the frequency response range of the detection module in the detection area, recording the fine acoustic conduction characteristics of the detection area and the corresponding load within a predetermined period after the adjustment is completed, and obtaining a fine feature set, so as to cluster the fine acoustic conduction feature set and screen a specific clustering group;

[0011] Extracting the fine acoustic conduction features in the specific cluster group to construct a corresponding fine acoustic conduction feature curve, and fitting it with the sample curve in the standard database, and determining whether there is an abnormality in the detection area based on the degree of fit;

[0012] Or, maintain the frequency response range of the detection module within the detection area.

[0013] Furthermore, in step S2, the discrete difference degree of the acoustic conduction characteristic is calculated according to formula (1):

[0014]

[0015] In formula (1), S represents the discrete difference of the acoustic conduction feature, ΔV represents the vibration frequency difference between the current vibration frequency and the average value of the called vibration frequency, V0 represents the standard deviation of the vibration frequency, Vei represents the difference between the current vibration frequency and the vibration frequency corresponding to the called ith acoustic conduction feature, ΔF represents the vibration amplitude difference between the current vibration amplitude and the average value of the called vibration amplitude, F0 represents the standard deviation of the vibration amplitude, Fei represents the difference between the current vibration amplitude and the vibration amplitude corresponding to the called ith acoustic conduction feature, and n represents the number of called acoustic conduction features.

[0016] Furthermore, in step S2, the abnormal crack tendency categories of the detection area are divided into:

[0017] If the acoustic conduction characteristic discrete difference is greater than or equal to a preset acoustic conduction characteristic discrete difference threshold, the crack abnormal tendency category of the detection area is a high abnormal tendency category;

[0018] If the acoustic conduction characteristic discrete difference is less than a preset acoustic conduction characteristic discrete difference threshold, the crack abnormal tendency category of the detection area is a low abnormal tendency category.

[0019] Furthermore, in step S3, the detection parameters of the detection module are adjusted and the abnormal conditions of the detection area are analyzed, including:

[0020] If the abnormal crack tendency category of the detection area is a high abnormal tendency category, the frequency response range of the detection module in the detection area is adjusted, and the fine acoustic conduction characteristics and the corresponding load of the detection area within a predetermined period after the adjustment are recorded to obtain a fine feature set, so as to cluster the fine acoustic conduction feature set and screen a specific clustering group;

[0021] Extracting the fine acoustic conduction features in the specific cluster group to construct a corresponding fine acoustic conduction feature curve, and fitting it with the sample curve in the standard database, and determining whether there is an abnormality in the detection area based on the degree of fit;

[0022] If the crack abnormal tendency category of the detection area is a low abnormal tendency category, the frequency response range of the detection module in the detection area is maintained.

[0023] Furthermore, in step S3, adjusting the frequency response range of the detection module in the detection area includes:

[0024] The frequency response range of the detection module in the detection area is increased, and the increase in the frequency response range is positively correlated with the discrete difference of the acoustic conduction characteristics.

[0025] Furthermore, in step S3, the process of clustering the fine acoustic conduction feature set and screening the specific clustering group includes:

[0026] Determine the load corresponding to each fine acoustic conduction feature;

[0027] Clustering each fine acoustic conduction feature according to the load to obtain cluster groups within different load ranges;

[0028] The cluster group corresponding to the load range with the largest upper limit value is selected as the specific cluster group;

[0029] The load difference between any two fine acoustic conduction features in the obtained cluster group must be less than a predetermined load difference threshold. Further, in step S3, the process of extracting the fine acoustic conduction features in the specific cluster group and constructing the corresponding fine acoustic conduction feature curve includes:

[0030] extracting the vibration amplitude of the specific cluster group;

[0031] A rectangular coordinate system is constructed with time as the horizontal axis and vibration amplitude as the vertical axis;

[0032] Marking the coordinate points of the vibration amplitude at each moment in each rectangular coordinate system;

[0033] The coordinate points are connected with a smooth curve to obtain a fine sound conduction characteristic curve.

[0034] Furthermore, in step S3, a number of sample curves corresponding to different loads are stored in the standard database.

[0035] Furthermore, in step S3, the process of fitting the sample curve in the standard database includes:

[0036] The fine sound conduction characteristic curve is fitted with several sample curves corresponding to the same load in the standard database, and several fitting degrees between the fine sound conduction characteristic curve and the sample curve are obtained.

[0037] Furthermore, in step S3, judging whether there is an abnormality in the detection area based on the degree of fit includes:

[0038] Calculating the mean of the degrees of fit of each of the degrees of fit;

[0039] If the fitting degree mean fitting degree is less than or equal to a preset fitting degree threshold, it is determined that an abnormality exists in the detection area.

[0040] Compared with the prior art, the present invention determines a number of detection areas at equal intervals along the bridge, and sets a detection module for collecting sound conduction characteristics and a load collection device for collecting the load in each detection area. Based on the current load in the detection area, a number of sound conduction characteristics corresponding to the same load are called from the historical database, and the current sound conduction characteristics are compared with the called sound conduction characteristics to obtain the discrete difference of the sound conduction characteristics, so as to classify the abnormal crack tendency category in the detection area, adaptively adjust the detection parameters of the detection module and analyze the abnormal situation in the detection area. Furthermore, when detecting gaps in large-area bridges in real time, the accuracy of detection is improved while ensuring the reliability of detection data, and computing resources are saved.

[0041] In particular, the present invention characterizes the abnormal crack tendency of the detection area through the discrete difference of the acoustic conduction characteristics. During the use of the bridge, the passage of vehicles will generate a certain vibration frequency and vibration amplitude on the bridge. At the same time, due to the increase in the service life and the influence of the external environment, the bridge will suffer certain losses. The abnormal crack situation is preliminarily judged by the changes in the vibration frequency and vibration amplitude, and the acoustic conduction characteristics of the current detection area are compared with several acoustic conduction characteristics corresponding to the same load in the historical database, wherein the acoustic conduction characteristics corresponding to the same load are selected for comparison to ensure the reliability and accuracy of the abnormal judgment, and the discrete difference of the acoustic conduction characteristics is used to provide data support for the subsequent division of the abnormal crack tendency category of the detection area. Furthermore, when targeting large-area bridges, the detection area with a higher abnormal tendency is identified, which is convenient for the subsequent targeted selection of the analysis method of the abnormal situation of the inspection area, and only the detection area with a higher abnormal tendency is used for refined analysis. In the face of the massive data brought by the inspection of large-area bridges, the reliability of the detection data is guaranteed, the accuracy of the detection is guaranteed, and computing resources are saved.

[0042] In particular, when the abnormal crack tendency category of the detection area is a high abnormal tendency category, the present invention adjusts the frequency response range of the detection module in the detection area under the high abnormal category to improve the data acquisition accuracy and obtain fine acoustic conduction characteristics, which is convenient for fine analysis of the fine acoustic conduction characteristics in the subsequent screening of specific clustering groups. In actual situations, the fine acoustic conduction characteristics in the specific clustering group correspond to larger loads, and the acoustic conduction characteristics caused by larger loads are stronger, and their data representativeness is stronger. When the fine acoustic conduction characteristic curve constructed at higher precision is fitted with the sample curve in the standard database, the accuracy is also higher, thereby providing data support for the subsequent determination of whether there is an abnormality in the detection area based on the degree of fit, and thereby, when real-time detection of gaps in large-area bridges is performed, the accuracy of detection is improved and computing resources are saved while ensuring the reliability of detection data. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 A schematic diagram of the steps of the automatic bridge crack detection method according to an embodiment of the invention;

[0044] Figure 2 A logical decision diagram for classifying abnormal crack tendency categories of the detection area according to the embodiment of the invention;

[0045] Figure 3 A logic decision diagram for adjusting detection parameters of the detection module and analyzing abnormal conditions in the detection area for an embodiment of the invention;

[0046] Figure 4 This is a logical decision diagram for determining whether there is an abnormality in the detection area according to an embodiment of the present invention. DETAILED DESCRIPTION

[0047] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0048] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0049] See also Figures 1 to 4 As shown, Figure 1 This is a schematic diagram of the steps of the automatic bridge crack detection method according to an embodiment of the present invention. Figure 2 A logical decision diagram for classifying abnormal crack tendency categories in the detection area according to the embodiment of the invention, Figure 3 A logical decision diagram for adjusting the detection parameters of the detection module and analyzing abnormal conditions in the detection area for an embodiment of the invention, Figure 4 The invention embodiment is a logic determination diagram for determining whether there is an abnormality in the detection area. The automatic detection method for bridge cracks in the embodiment of the invention includes:

[0050] Step S1, determining a number of detection areas at equal intervals along the bridge, and setting a detection module for collecting acoustic conduction characteristics and a load collection device for collecting the load in each of the detection areas, wherein the acoustic conduction characteristics include vibration frequency and vibration amplitude;

[0051] Step S2, based on the current load on the detection area, calling a number of acoustic conduction features corresponding to the same load from the historical database, comparing the current acoustic conduction feature with the called acoustic conduction feature, and obtaining the discrete difference of the acoustic conduction feature to classify the abnormal crack tendency category of the detection area;

[0052] Step S3, adjusting the detection parameters of the detection module according to the abnormal crack tendency category corresponding to the detection area and analyzing the abnormal situation of the detection area, including:

[0053] Adjusting the frequency response range of the detection module in the detection area, recording the fine acoustic conduction characteristics of the detection area and the corresponding load within a predetermined period after the adjustment is completed, and obtaining a fine feature set, so as to cluster the fine acoustic conduction feature set and screen a specific clustering group;

[0054] Extracting the fine acoustic conduction features in the specific cluster group to construct a corresponding fine acoustic conduction feature curve, and fitting it with the sample curve in the standard database, and determining whether there is an abnormality in the detection area based on the degree of fit;

[0055] Or, maintain the frequency response range of the detection module within the detection area.

[0056] Specifically, for large-area bridges, there is no limitation on the setting method of the detection area. For example, the bridge can be divided into several bridge sections, and a single bridge section can be used as the detection area. The length of the bridge section can be set within 50m to 100m, which will not be repeated here.

[0057] Specifically, there is no limitation on the specific structure of the detection module, which only needs to have the function of detecting and collecting the acoustic conduction characteristics of the detection area, for example, any vibration sensor with an adjustable frequency response range in the prior art, which will not be repeated here.

[0058] Specifically, there is no limitation on the specific structure of the load collection device, which only needs to have the function of detecting and collecting the load on the detection area. For example, a gravity sensor is set in the detection area to detect the load of the vehicle passing through, so as to obtain the load on the detection area. The setting position is preferably the middle position of the detection area. Of course, other forms can also be used, which will not be repeated here.

[0059] It can be understood that the load and the acoustic conduction characteristics correspond to each other. In implementation, the acoustic conduction characteristics in a predetermined time period before and after the load occurs are used as the acoustic conduction characteristics corresponding to the load, and the predetermined time period is selected within the interval [2.5s, 5s].

[0060] Specifically, the historical database is used to store the loads collected in each detection area and the corresponding acoustic conduction characteristics. The standard database can be obtained by filtering the data in the historical database. A predetermined number of acoustic conduction characteristics corresponding to different loads are filtered out from the historical database, and the acoustic conduction characteristic curves are synchronously constructed and stored in the standard database. The predetermined number must be greater than 10.

[0061] Specifically, in step S2, the discrete difference of the acoustic conduction characteristics is calculated according to formula (1):

[0062]

[0063] In formula (1), S represents the discrete difference of the acoustic conduction feature, ΔV represents the vibration frequency difference between the current vibration frequency and the average value of the called vibration frequency, V0 represents the standard deviation of the vibration frequency, Vei represents the difference between the current vibration frequency and the vibration frequency corresponding to the called ith acoustic conduction feature, ΔF represents the vibration amplitude difference between the current vibration amplitude and the average value of the called vibration amplitude, F0 represents the standard deviation of the vibration amplitude, Fei represents the difference between the current vibration amplitude and the vibration amplitude corresponding to the called ith acoustic conduction feature, and n represents the number of called acoustic conduction features.

[0064] Specifically, in this embodiment, the standard deviation value V0 of the vibration frequency and the standard deviation value F0 of the vibration amplitude are obtained in advance, and relevant data of several completed bridge crack detections are obtained, and the average value of the vibration frequency difference ΔVe between several vibration frequencies in the relevant data and the average value of the vibration frequencies called therefrom and the average value of the vibration amplitude difference ΔFe between several vibration amplitudes in the relevant data and the average value of the vibration amplitudes called therefrom are calculated and set, V0=ΔVe, F0=ΔFe.

[0065] It can be understood that the sound conduction characteristic is the vibration frequency over a period of time, the current vibration frequency is the average vibration frequency over a period of time, and similarly, the vibration amplitude is the average vibration amplitude over a period of time.

[0066] The discrete difference threshold S0 of the acoustic conduction characteristics is selected in the interval [0.23, 0.63].

[0067] The present invention characterizes the abnormal crack tendency of the detection area through the discrete difference of the acoustic conduction characteristics. During the use of the bridge, the passage of vehicles will generate a certain vibration frequency and vibration amplitude on the bridge. At the same time, due to the increase in the service life and the influence of the external environment, the bridge will suffer certain losses. The abnormal crack situation is preliminarily judged by the changes in the vibration frequency and vibration amplitude. The acoustic conduction characteristics of the current detection area are compared with several acoustic conduction characteristics corresponding to the same load in the historical database. Among them, the acoustic conduction characteristics corresponding to the same load are selected for comparison to ensure the reliability and accuracy of the abnormal judgment. The discrete difference of the acoustic conduction characteristics is used to provide data support for the subsequent division of the abnormal crack tendency category of the detection area. Furthermore, when targeting large-area bridges, the detection area with a higher abnormal tendency is identified, which is convenient for the subsequent targeted selection of the analysis method of the abnormal situation of the inspection area. Only for the detection area with a higher abnormal tendency, refined analysis is used. In the face of the massive data brought by the inspection of large-area bridges, the reliability of the detection data is guaranteed, the accuracy of the detection is guaranteed, and computing resources are saved.

[0068] Specifically, in step S2, the abnormal crack tendency categories of the detection area are divided into:

[0069] If the acoustic conduction characteristic discrete difference is greater than or equal to a preset acoustic conduction characteristic discrete difference threshold, the crack abnormal tendency category of the detection area is a high abnormal tendency category;

[0070] If the acoustic conduction characteristic discrete difference is less than a preset acoustic conduction characteristic discrete difference threshold, the crack abnormal tendency category of the detection area is a low abnormal tendency category.

[0071] Specifically, in step S3, adjusting the detection parameters of the detection module and analyzing the abnormality of the detection area include:

[0072] If the abnormal crack tendency category of the detection area is a high abnormal tendency category, the frequency response range of the detection module in the detection area is adjusted, and the fine acoustic conduction characteristics and the corresponding load of the detection area within a predetermined period after the adjustment are recorded to obtain a fine feature set, so as to cluster the fine acoustic conduction feature set and screen a specific clustering group;

[0073] Extracting the fine acoustic conduction features in the specific cluster group to construct a corresponding fine acoustic conduction feature curve, and fitting it with the sample curve in the standard database, and determining whether there is an abnormality in the detection area based on the degree of fit;

[0074] If the crack abnormal tendency category of the detection area is a low abnormal tendency category, the frequency response range of the detection module in the detection area is maintained.

[0075] Specifically, in step S3, adjusting the frequency response range of the detection module in the detection area includes:

[0076] The frequency response range of the detection module in the detection area is increased, and the increase in the frequency response range is positively correlated with the discrete difference of the acoustic conduction characteristics.

[0077] In this embodiment, optionally,

[0078] The acoustic conduction characteristic discrete difference S is compared with the first acoustic conduction characteristic discrete difference comparison threshold S1 and the second acoustic conduction characteristic discrete difference comparison threshold S2,

[0079] If S>S2, the frequency response range increase value is determined to be the first frequency response range increase value c1, and c1=0.45c0;

[0080] If S1≤S≤S2, the frequency response range increase value is determined to be the second frequency response range increase value c2, and c2=0.3c0 is set;

[0081] If S<S1, the frequency response range increase value is determined to be the third frequency response range increase value c3, and c3=0.15c0;

[0082] Wherein, c0 represents the initial frequency response range, S1=1.2S0, S2=1.3S0.

[0083] There is no limitation on the method of increasing the frequency response range. It is understandable that the vibration sensor can increase the range of detectable data by adjusting the frequency response range. For complex vibration signals containing multiple frequency components, a wide frequency response range can ensure the integrity of the signal and improve the accuracy of the measurement. This will not be elaborated here.

[0084] Specifically, in step S3, the process of clustering the fine acoustic conduction feature set and screening the specific clustering group includes:

[0085] Determine the load corresponding to each fine acoustic conduction feature;

[0086] Clustering each fine acoustic conduction feature according to the load to obtain cluster groups within different load ranges;

[0087] The cluster group corresponding to the load range with the largest upper limit value is selected as the specific cluster group;

[0088] The load difference between any two fine acoustic conduction features in the obtained cluster group must be less than a predetermined load difference threshold.

[0089] The load difference threshold K0 is selected in the interval [100kg, 300kg].

[0090] Specifically, in step S3, the process of extracting the fine acoustic conduction features in the specific cluster group and constructing the corresponding fine acoustic conduction feature curve includes:

[0091] extracting the vibration amplitude of the specific cluster group;

[0092] A rectangular coordinate system is constructed with time as the horizontal axis and vibration amplitude as the vertical axis;

[0093] Marking the coordinate points of the vibration amplitude at each moment in each rectangular coordinate system;

[0094] The coordinate points are connected with a smooth curve to obtain a fine sound conduction characteristic curve.

[0095] Specifically, in step S3, a number of sample curves corresponding to different loads are stored in the standard database.

[0096] Specifically, in step S3, the process of fitting the sample curve in the standard database includes:

[0097] The fine sound conduction characteristic curve is fitted with several sample curves corresponding to the same load in the standard database, and several fitting degrees between the fine sound conduction characteristic curve and the sample curve are obtained.

[0098] Specifically, there is no limitation on the method of solving the degree of fit. It can be understood that the fine sound conduction characteristic curve and the sample curve are both time domain curves. Technical personnel in this field can use any algorithm or model to solve the degree of fit of the time domain curve, for example, the mean square error algorithm, the root mean square error algorithm and other methods, which will not be repeated here.

[0099] Specifically, in step S3, judging whether there is an abnormality in the detection area based on the degree of fit includes:

[0100] Calculating the mean of the degrees of fit of each of the degrees of fit;

[0101] If the fitting degree mean fitting degree is less than or equal to a preset fitting degree threshold, it is determined that an abnormality exists in the detection area.

[0102] The fitting threshold B0 is preset. After pre-screening several bridges without abnormality, the fitting degrees of several fine sound conduction characteristic curves and sample curves corresponding to each detection area are taken as samples, and the sample fitting mean ΔB is solved. B0 is set to be g×ΔB, where g represents the offset coefficient, 0.85<g<0.95.

[0103] When the abnormal crack tendency category of the detection area is a high abnormal tendency category, the present invention adjusts the frequency response range of the detection module in the detection area under the high abnormal category to improve the data acquisition accuracy to obtain fine acoustic conduction characteristics, which is convenient for fine analysis of the fine acoustic conduction characteristics in the subsequent screening of specific clustering groups. In actual situations, the fine acoustic conduction characteristics in the specific clustering group correspond to larger loads, and the acoustic conduction characteristics caused by larger loads are stronger, and their data representativeness is stronger. When the fine acoustic conduction characteristic curve constructed at higher precision is fitted with the sample curve in the standard database, the accuracy is also higher, thereby providing data support for the subsequent determination of whether there is an abnormality in the detection area based on the degree of fit. Furthermore, when real-time detection of gaps in large-area bridges is performed, the accuracy of detection is improved and computing resources are saved while ensuring the reliability of detection data.

[0104] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A method for automatically detecting bridge cracks, characterized in that: include: Step S1, determining a number of detection areas at equal intervals along the bridge, and setting a detection module for collecting acoustic conduction characteristics and a load collection device for collecting the load in each of the detection areas, wherein the acoustic conduction characteristics include vibration frequency and vibration amplitude; Step S2, based on the current load on the detection area, calling a number of acoustic conduction features corresponding to the same load from the historical database, comparing the current acoustic conduction feature with the called acoustic conduction feature, and obtaining the discrete difference of the acoustic conduction feature to classify the abnormal crack tendency category of the detection area; Step S3, adjusting the detection parameters of the detection module according to the abnormal crack tendency category corresponding to the detection area and analyzing the abnormal situation of the detection area, including: Adjusting the frequency response range of the detection module in the detection area, recording the fine acoustic conduction characteristics of the detection area and the corresponding load within a predetermined period after the adjustment is completed, and obtaining a fine feature set, so as to cluster the fine acoustic conduction feature set and screen a specific clustering group; Extracting the fine acoustic conduction features in the specific cluster group to construct a corresponding fine acoustic conduction feature curve, and fitting it with the sample curve in the standard database, and determining whether there is an abnormality in the detection area based on the degree of fit; Or, maintain the frequency response range of the detection module within the detection area.

2. The automatic bridge crack detection method according to claim 1 is characterized in that: In step S2, the discrete difference of the acoustic conduction characteristics is calculated according to formula (1): In formula (1), S represents the discrete difference of the acoustic conduction feature, ΔV represents the vibration frequency difference between the current vibration frequency and the average value of the called vibration frequency, V0 represents the standard deviation of the vibration frequency, Vei represents the difference between the current vibration frequency and the vibration frequency corresponding to the called ith acoustic conduction feature, ΔF represents the vibration amplitude difference between the current vibration amplitude and the average value of the called vibration amplitude, F0 represents the standard deviation of the vibration amplitude, Fei represents the difference between the current vibration amplitude and the vibration amplitude corresponding to the called ith acoustic conduction feature, and n represents the number of called acoustic conduction features.

3. The automatic bridge crack detection method according to claim 1 is characterized in that: In step S2, the crack abnormality tendency category of the detection area is divided into: include, If the acoustic conduction characteristic discrete difference is greater than or equal to a preset acoustic conduction characteristic discrete difference threshold, the crack abnormal tendency category of the detection area is a high abnormal tendency category; If the acoustic conduction characteristic discrete difference is less than a preset acoustic conduction characteristic discrete difference threshold, the crack abnormal tendency category of the detection area is a low abnormal tendency category.

4. The automatic bridge crack detection method according to claim 1 is characterized in that: In step S3, the detection parameters of the detection module are adjusted and the abnormal conditions of the detection area are analyzed. include, If the abnormal crack tendency category of the detection area is a high abnormal tendency category, the frequency response range of the detection module in the detection area is adjusted, and the fine acoustic conduction characteristics and the corresponding load of the detection area within a predetermined period after the adjustment are recorded to obtain a fine feature set, so as to cluster the fine acoustic conduction feature set and screen a specific clustering group; Extracting the fine acoustic conduction features in the specific cluster group to construct a corresponding fine acoustic conduction feature curve, and fitting it with the sample curve in the standard database, and determining whether there is an abnormality in the detection area based on the degree of fit; If the crack abnormal tendency category of the detection area is a low abnormal tendency category, the frequency response range of the detection module in the detection area is maintained.

5. The automatic bridge crack detection method according to claim 1 is characterized in that: In the step S3, adjusting the frequency response range of the detection module in the detection area includes: The frequency response range of the detection module in the detection area is increased, and the increase in the frequency response range is positively correlated with the discrete difference of the acoustic conduction characteristics.

6. The automatic bridge crack detection method according to claim 1 is characterized in that: In step S3, the process of clustering the fine acoustic conduction feature set and screening the specific clustering group includes: Determine the load corresponding to each fine acoustic conduction feature; Clustering each fine acoustic conduction feature according to the load to obtain cluster groups within different load ranges; The cluster group corresponding to the load range with the largest upper limit value is selected as the specific cluster group; The load difference between any two fine acoustic conduction features in the obtained cluster group must be less than a predetermined load difference threshold.

7. The automatic bridge crack detection method according to claim 1 is characterized in that: In step S3, the process of extracting the fine acoustic conduction features in the specific cluster group and constructing the corresponding fine acoustic conduction feature curve includes: extracting the vibration amplitude of the specific cluster group; A rectangular coordinate system is constructed with time as the horizontal axis and vibration amplitude as the vertical axis; Marking the coordinate points of the vibration amplitude at each moment in each rectangular coordinate system; The coordinate points are connected with a smooth curve to obtain a fine sound conduction characteristic curve.

8. The automatic bridge crack detection method according to claim 1 is characterized in that: In step S3, a number of sample curves corresponding to different loads are stored in the standard database.

9. The automatic bridge crack detection method according to claim 1, characterized in that: In step S3, the process of fitting the sample curve in the standard database includes: The fine sound conduction characteristic curve is fitted with several sample curves corresponding to the same load in the standard database, and several fitting degrees between the fine sound conduction characteristic curve and the sample curve are obtained.

10. The automatic bridge crack detection method according to claim 9, characterized in that: In step S3, judging whether there is an abnormality in the detection area based on the degree of fit includes: Calculating the mean of the degrees of fit of each of the degrees of fit; If the fitting degree mean fitting degree is less than or equal to a preset fitting degree threshold, it is determined that an abnormality exists in the detection area.

Citation Information

Cited By

  • Road bridge displacement monitoring method and system based on Internet of Things

    CN120832518A