High-strength concrete construction crack detection system and method

Abstract

The invention belongs to the technical field of concrete crack detection, and particularly discloses and provides a high-strength concrete construction crack detection system and method.The system comprises a multi-parameter cooperative sensing module, a construction crack recognition module, a crack risk analysis module and a crack early warning processing module. According to the invention, the temperature, humidity, strain and multispectral image data from a plasticity stage to a hardening stage are collected in real time through the multi-parameter cooperative sensing module, and crack characteristics are extracted by combining staged differential signal processing and a deep convolutional neural network, so that the detection precision of micro cracks and internal defects is remarkably improved. Meanwhile, multi-physics field data are fused, a crack evolution risk rate model is innovatively constructed, dynamic probability prediction before crack initiation is achieved, a grading disposal scheme is automatically triggered through risk grade division and a progressive early warning mechanism, drying shrinkage and temperature crack causes are effectively distinguished, the prevention and control efficiency is remarkably improved, and the method is suitable for large-scale popularization and application. And full-life-cycle accurate monitoring of the high-strength concrete is supported.

Description

Technical Field

[0001] The present invention belongs to the technical field of concrete crack detection. Specifically, it relates to a high-strength concrete construction crack detection system and method. Background Art

[0002] Due to its high compressive strength and good durability, high-strength concrete is widely used in scenarios with demanding material performance requirements such as super high-rise buildings, long-span bridges, nuclear power plants, etc. Although it has high strength, due to the large amount of cementitious materials used and low water-cement ratio, it is prone to microcracks caused by temperature stress, shrinkage deformation, etc. during the construction stage, thus highlighting the necessity of crack detection during construction.

[0003] Existing technologies such as a civil construction monitoring intelligent sensor information detection method and system under signal mismatch disclosed in the Chinese invention patent application with the application number 202411431893.5 mainly extract the features of temperature / strain signals through wavelet denoising combined with the singular value decomposition algorithm, and reconstruct the internal field distribution based on numerical simulation to solve the problem of accurate identification of concrete internal defects under signal mismatch, achieve high-precision monitoring and defect location, and improve the reliability of structural health assessment.

[0004] Existing technologies such as a detection method for the construction quality of rockfill concrete disclosed in the Chinese invention patent application with the application number 202311748375.1 mainly uses a dynamic detection mechanism that triggers image acquisition through temperature thresholds, combines steady-state temperature analysis with real-time image processing to solve the problem of lag in real-time monitoring of cracks caused by abnormal temperatures during the construction period of rockfill concrete, and achieves the effect of quickly identifying cracks and giving early warnings to ensure construction quality and structural safety.

[0005] In view of the above technical solutions, obviously, there are still the following deficiencies in the current concrete construction crack detection: 1. There is no detection mechanism that dynamically matches the plastic-hardening stage. For example, fixed wavelet bases (db4) and static temperature thresholds lead to the loss of signal features and misjudgment, missing the intervention window in the plastic stage and lacking sensitivity in the hardening stage.

[0006] 2. It has low sensitivity to early microcracks and internal hidden cracks. At the same time, image recognition is easily interfered by surface floating slurry, resulting in a high missed detection rate.

[0007] 3. There is a lack of a dynamic probability prediction model and a hierarchical disposal strategy before crack initiation, and it is impossible to actively prevent and control large-area avoidable cracks.

[0008] 4. It relies on single / double parameters, does not integrate the collaborative perception of multiple physical fields, and ignores the inducing effects of the humidity field and construction vibration on cracks, resulting in misjudgment of the causes. For example, drying shrinkage is misjudged as a temperature crack. Summary of the Invention

[0009] In view of this, to solve the problems raised in the above-mentioned background technology, a high-strength concrete construction crack detection system and method are proposed herein.

[0010] The object of the present invention can be achieved through the following technical solutions: The present invention provides a high-strength concrete construction crack detection system, which includes: a multi-parameter collaborative sensing module that collects temperature, humidity, strain, and multi-spectral image data in real time from the plastic stage to the hardening stage of the concrete.

[0011] A construction crack identification module, which consists of an image analysis unit and a risk prediction unit.

[0012] The image analysis unit uses a deep convolutional neural network to extract crack features from the multi-spectral image and outputs the number, width, and location of the cracks.

[0013] The risk prediction unit, when no cracks are detected, synthesizes the temperature, humidity, and strain data, outputs the crack evolution risk rate in the time series, and constructs a crack evolution risk rate growth curve.

[0014] A crack risk analysis module that conducts risk level division based on the number, width, and location of the cracks and the set risk level division rules.

[0015] A crack early warning and processing module, which consists of a multi-level disposal strategy library and a dynamic early warning unit.

[0016] The multi-level disposal strategy library stores disposal plans linked to the risk level.

[0017] The dynamic early warning unit calls the disposal plan corresponding to the risk level in the multi-level disposal strategy library to process the cracks, and triggers a progressive early warning based on the curve. When the curve slope exceeds the threshold, a crack risk emergency response is initiated.

[0018] The present invention also provides a high-strength concrete construction crack detection method, which includes: A1. Multi-parameter collaborative data collection: Collect temperature, humidity, strain, and multi-spectral image data in real time from the plastic stage to the hardening stage of the concrete.

[0019] A2. Construction crack identification: Use a deep convolutional neural network to extract crack features from the multi-spectral image, output the number, width, and location of the cracks, and when no cracks are detected, synthesize the temperature, humidity, and strain data, output the crack evolution risk rate in the time series, and construct a crack evolution risk rate growth curve.

[0020] A3. Crack risk analysis: Conduct risk level division based on the number, width, and location of the cracks and the set risk level division rules.

[0021] A4. Crack warning and treatment: Based on the risk level, call the treatment plan corresponding to the risk level in the multi-level treatment strategy library for crack treatment, trigger progressive warnings based on the crack evolution risk rate growth curve, and initiate crack risk emergency responses when the curve slope exceeds the threshold.

[0022] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention uses a multi-parameter collaborative sensing module to collect temperature, humidity, strain, and multi-spectral image data in the plastic to hardening stage in real time. By combining stage-differentiated signal processing and deep convolutional neural networks to extract crack features, the detection accuracy of micro-cracks and internal defects is significantly improved. At the same time, by integrating multi-physical field data, an innovative crack evolution risk rate model is constructed to achieve dynamic probability prediction before crack initiation. Through risk level classification and progressive warning mechanisms, the hierarchical treatment plan is automatically triggered, effectively distinguishing the causes of drying shrinkage and temperature cracks, and significantly improving the prevention and control efficiency.

[0023] (2) In the plastic stage, the present invention uses the sym8 wavelet basis for high-frequency filtering to capture transient temperature fluctuations, and switches to the db6 wavelet basis for low-frequency filtering to suppress noise in the hardening stage. By combining stage-differentiated signal processing, the problem of feature loss caused by a fixed wavelet basis such as db4 is solved. The recognition accuracy of the intervention window in the plastic stage is significantly improved, and the sensitivity in the hardening stage is also greatly enhanced.

[0024] (3) By integrating four-dimensional data of temperature-humidity-strain-multi-spectral, the present invention uses a polarization filter, morphological filtering, and frequency domain fusion technology to eliminate the interference of surface floating slurry. The projection positioning error of internal defects is significantly reduced. At the same time, through the gradient layout of distributed humidity sensors combined with diffusion model compensation, the coupling effect of drying shrinkage and temperature strain can be identified, which is convenient for cause identification and also reduces the misjudgment rate of causes.

[0025] (4) By integrating multiple physical parameters to construct a crack evolution risk rate model based on time series, the present invention realizes accurate risk probability prediction before crack initiation. Combined with the three-level risk level classification and progressive warning, the prevention and control rate of avoidable cracks is significantly improved. Description of the Drawings

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0027] Figure 1 It is a schematic diagram of the connection of the system modules of the present invention.

[0028] Figure 2This is a schematic diagram of the implementation steps of the method of the present invention.

[0029] Figure 3 This is a schematic diagram of the composition structure of the construction crack identification module.

[0030] Figure 4 This is a schematic diagram of the composition structure of the signal preprocessing unit.

[0031] Figure 5 This is a schematic diagram of the composition structure of the crack early warning processing module. Specific implementation manners

[0032] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present invention.

[0033] Please refer to Figure 1 As shown, the present invention provides a high-strength concrete construction crack detection system, which includes: a multi-parameter collaborative sensing module, a construction crack identification module, a crack risk analysis module, and a crack early warning processing module.

[0034] Among the above, the construction crack identification module is respectively connected to the multi-parameter collaborative sensing module and the crack risk analysis module, and the crack risk analysis module is connected to the crack early warning processing module.

[0035] The multi-parameter collaborative sensing module collects temperature, humidity, strain, and multi-spectral image data in real time from the plastic stage to the hardening stage of the concrete.

[0036] Specifically, the multi-parameter collaborative sensing module includes: collecting three-dimensional distribution data of the temperature field from the plastic stage to the hardening stage in real time through a distributed optical fiber temperature sensor array buried inside the concrete along a three-dimensional orthogonal grid of the concrete structure, denoted as temperature data.

[0037] Capacitive humidity sensor arrays are installed in a gradient distribution on the surface layer, 1 / 2 thickness layer, and bottom layer of the concrete structure to collect humidity data from the plastic stage to the hardening stage in real time.

[0038] FBG strain sensor arrays are orthogonally arranged at the beam ends and plate corners of the concrete to measure axial strain and shear strain from the plastic stage to the hardening stage in real time, and the strain data is integrated.

[0039] Through a multispectral imager installed on a construction robot, dual-band scanning is carried out in real time using visible light and near-infrared, and combined with a three-way polarization filter to eliminate the interference of surface floating slurry reflection, so as to obtain multispectral image data corresponding to the plastic stage to the hardening stage after eliminating the interference.

[0040] Understandably, high-strength concrete usually has a low water-cement ratio and contains mineral admixtures, so it has high early strength, but this may also lead to large autogenous shrinkage and is prone to cracking. Cracks may appear in different stages, such as the plastic stage, the hardening stage, or load cracks after long-term use. Therefore, the present invention focuses on targeted crack detection in specific stages.

[0041] Please refer to Figure 3 As shown, the construction crack identification module consists of an image analysis unit, a risk prediction unit, and a signal preprocessing unit.

[0042] The image analysis unit uses a deep convolutional neural network to extract crack features from the multispectral image and outputs the number, width, and location of the cracks.

[0043] Specifically, the deep convolutional neural network includes: a multi-scale feature extraction layer configured to parallelly extract crack texture features using dilated convolutional kernels (3×3, 5×5, 7×7).

[0044] An attention enhancement module configured to enhance the feature weights of the crack region through the SE-Net channel attention mechanism.

[0045] A transfer learning framework, a pre-trained model, such as constructed based on 5000 groups of concrete CT scan images and synchronous surface images.

[0046] It should be added that using a deep convolutional neural network to extract crack features from the multispectral image is a relatively common technical means in the art, and its more detailed feature extraction process will not be elaborated here.

[0047] When no cracks are detected, the risk prediction unit synthesizes temperature, humidity, and strain data, outputs the crack evolution risk rate under the time series, and constructs a crack evolution risk rate growth curve.

[0048] Specifically, the crack evolution risk rate output under the time series is used to output the crack evolution risk rate corresponding to the plastic stage under the time series, including: B1. For the plastic stage, traverse each temperature sensor, calculate the absolute value of the temperature difference between it and all adjacent points, and take the maximum absolute temperature difference as the target temperature difference.

[0049] B2. Calculate the power spectral density integrals of the stress signals in the beam end direction and the shear strain signals in the orthogonal direction respectively, output the axial strain energy and the shear strain energy, calculate the axial strain energy by synthesizing the axial strain energy and the shear strain energy, and calculate the total energy ratio.

[0050] Further, the specific calculation formula for the axial strain energy is: , represents the axial strain energy, represents the frequency. During the integration process, will traverse all the frequency values between 10 Hz and 50 Hz, represents the power spectral density of the corresponding stress signal along the beam end direction at the frequency and reflects the energy distribution of this frequency component.

[0051] The specific calculation formula for the shear strain energy is: , represents the shear strain energy, represents the power spectral density of the corresponding shear strain signal along the orthogonal direction at the frequency .

[0052] The specific calculation formula for calculating the axial strain energy by synthesizing the axial strain energy and the shear strain energy and calculating the total energy ratio is: , represents the interval duration, which represents the time length corresponding to calculating the changes in the axial strain energy and the shear strain energy. The strain data is collected and analyzed with a 1-minute period, takes a value of 60 seconds, and is a dynamic value. When the structure contraction intensifies, will decrease, such as being set to 20 seconds. And the higher the energy ratio, the greater the strain energy accumulated in the structure, and the increased risk of plastic deformation.

[0053] B3. Calculate the standard deviations of the humidity data corresponding to the surface layer, the 1 / 2 thickness layer, and the bottom layer of the concrete structure respectively, and perform a weighted summation calculation to output the target humidity fluctuation degree.

[0054] It can be understood that the humidity distribution of the concrete structure may vary at different depths due to factors such as environmental exposure and moisture migration. The surface layer is directly in contact with the external environment and has large humidity fluctuations. The middle layer belongs to the transition zone and is less affected. The bottom layer has different humidity conditions due to the humidity changes in the placement area. Exemplarily, the weights of the standard deviations of the humidity data corresponding to the surface layer, the 1 / 2 thickness layer, and the bottom layer of the concrete structure can be taken as 0.4, 0.2, and 0.4 respectively.

[0055] B4. Statistically obtain the microtrace density, bleeding channel length, aggregate dispersion, and interference fringe intensity from the multispectral image data, and construct an image feature vector.

[0056] Understandably, the microtrace density is positively correlated with the crack risk rate. Corresponding to plastic shrinkage cracks, in the areas where the microtrace density exceeds the threshold, it is very likely to develop into visible cracks with a width ≥ 0.1 mm during the subsequent hardening stage. The bleeding channel length corresponds to dry shrinkage cracks. The bleeding channels accelerate the local humidity decline rate, resulting in the concentration of dry shrinkage stress. The aggregate dispersion corresponds to interfacial cracks, and the interference fringe intensity corresponds to temperature stress cracks. Therefore, the four indicators of microtrace density, bleeding channel length, aggregate dispersion, and interference fringe intensity are selected for risk analysis in the image dimension.

[0057] Among them, the microtrace density is calculated by processing the dark lines in the polarization image, which involves edge detection and morphological processing. The bleeding channel length is calculated by using the Hough transform. The aggregate dispersion involves image segmentation, such as the SLIC algorithm, to distinguish the aggregate and paste regions. The interference fringe intensity may be obtained by analyzing color differences or light intensity changes, such as calculating the ΔE value in the CIELab color space.

[0058] Exemplarily, the microtrace density is obtained by performing Canny edge detection based on the degree of linear polarization (DoLP) channel and morphological closing operation, and counting the number of shrinkage dark lines per unit area to obtain the microtrace density. The bleeding channel length is obtained by detecting the linear bright band using the Hough transform of the near-infrared angle of polarization (AoP) image and calculating the geometric length of the continuous bright band as the bleeding channel length. The aggregate dispersion is obtained by performing SLIC superpixel segmentation on the visible light multispectral image and calculating the proportion of the aggregate-paste interface region as the aggregate dispersion. The interference fringe intensity is obtained by extracting the polarization degree difference of the RGB three channels and calculating the color difference in the CIELab color space to obtain the interference fringe intensity.

[0059] B5. For each acquisition time point, comprehensively consider the target temperature difference, total energy ratio, target humidity fluctuation degree, and the image feature vector, calculate the crack evolution risk rate, and further construct the crack evolution risk rate corresponding to the time series in the plastic stage.

[0060] Specifically, the calculation process of calculating the crack evolution risk rate is as follows: perform normalization processing on the target temperature difference, total energy ratio, and target humidity fluctuation degree respectively, and output the normalized target temperature difference, total energy ratio, and target humidity fluctuation degree, which are respectively denoted as 、 、 。

[0061] Normalize the bleeding channel length and microtrace density by the Min-Max normalization method, and standardize the aggregate dispersion and interference fringe intensity by the Z-Score standardization. Perform a weighted summation calculation on the processed bleeding channel length, microtrace density, aggregate dispersion, and interference fringe intensity, and output the comprehensive image feature value, denoted as .

[0062] Based on , , and , a weighted summation is used to obtain the crack evolution risk rate, denoted as , , , , and represent the target temperature difference, total energy ratio, target humidity fluctuation degree, and the weights corresponding to the comprehensive image feature value, respectively.

[0063] Understandably, the weights of the bleeding channel length, microtrace density, aggregate dispersion, and interference fringe intensity, as well as , , and are all set by fitting historical data or expert experience.

[0064] Specifically, the crack evolution risk rate in the time series is output, including: C1, the crack evolution risk rate based on the time series corresponding to the plastic stage, and a hardening stage risk compensation factor is set.

[0065] Furthermore, setting the hardening stage risk compensation factor includes: screening out the maximum crack evolution risk rate from the crack evolution risk rates in the time series corresponding to the plastic stage.

[0066] Statistically record the duration during which the crack evolution risk rate is greater than the set reference interference value as the crack evolution interference duration.

[0067] Normalize the maximum crack evolution risk rate and the crack evolution interference duration, and input them into the sigmoid function to output the hardening stage risk compensation factor.

[0068] Specifically, the core reason for using the sigmoid function in the calculation of the concrete hardening stage risk compensation factor is that it can transform the non-linear and asymmetric risk characteristics into continuous compensation values within the 0-1 interval, which meets the threshold effect of crack evolution and the risk control requirements.

[0069] C2. Calculate the temperature rise rate, strain acceleration value, humidity gradient value, and polarization interference fringe density within each time window during the hardening stage based on the temperature, humidity, strain, and multi-spectral image data in the hardening stage.

[0070] Understandably, during the concrete hardening stage, the temperature rise rate can be calculated through time series analysis of temperature data, obtained by dividing the temperature difference between adjacent time points by the time interval. The strain acceleration value requires second-order time differentiation of the strain data and uses the numerical difference method to calculate the change amplitude of the strain change rate per unit time. The humidity gradient value is based on multi-position humidity sensor data and calculates the ratio of the humidity difference between adjacent points in space to the distance. The polarization interference fringe density is quantified by analyzing multi-spectral images and statistically calculating the number of interference fringes per unit area or the reciprocal of the spacing. These parameters respectively reflect the dynamic characteristics of thermodynamics, mechanical deformation, moisture migration, and microstructural evolution during the hardening process. Therefore, analyze the temperature rise rate, strain acceleration value, humidity gradient value, and polarization interference fringe density.

[0071] C3. Normalize the temperature rise rate, strain acceleration value, humidity gradient value, and polarization interference fringe density, and obtain the crack evolution risk rate within each time window through weighted summation.

[0072] It should be added that the time window can be set to 30 minutes.

[0073] It also should be added that the normalization process is to uniformly scale parameters with different dimensions such as the temperature rise rate, strain acceleration value, humidity gradient value, and polarization interference fringe density to [0, 1] to eliminate the influence of dimension differences on comprehensive evaluation, usually using extreme value normalization. The weight setting needs to be determined according to the contribution degree of each parameter to the crack risk. It is recommended that the temperature rise rate and strain acceleration be given higher weights of 0.3 because they are directly related to thermal stress and deformation mutations. The humidity gradient reflects the shrinkage stress caused by moisture migration and is set to 0.25. The polarization interference fringe density, as an index of microstructural evolution, has a lower weight and is set to 0.15.

[0074] C4. Correct the crack evolution risk rate during the hardening stage based on the hardening stage risk compensation factor, and output the crack evolution risk rate in the time series based on the corrected crack evolution risk rate values within each time window.

[0075] Specifically, the larger the hardening stage risk compensation factor, the corresponding increase in the crack evolution risk rate during the hardening stage. Exemplarily, correcting the crack evolution risk rate during the hardening stage based on the hardening stage risk compensation factor satisfies the following correction formula: , represents the corrected crack evolution risk rate value, represents the crack evolution risk rate during the hardening stage, Represents the risk compensation factor in the hardening stage.

[0076] In real time, the present invention constructs a crack evolution risk rate model based on time series by fusing multiple physical parameters, realizes accurate risk probability prediction before crack initiation, and combines three-level risk level division and progressive early warning, thus significantly improving the prevention and control rate of avoidable cracks.

[0077] Please refer to Figure 4 As shown, the signal preprocessing unit is specifically configured as follows: a time series data processing channel that performs the following staged differential processing on temperature, humidity, and strain data: In the plastic stage, the temperature data is subjected to high-frequency band-pass filtering using the sym8 wavelet basis, the strain data is subjected to dynamic wide-band pass filtering, and the humidity data is subjected to baseline drift correction.

[0078] In the hardening stage, the temperature data switches to the db6 wavelet basis for low-frequency band-pass filtering, the strain data is subjected to narrow-band pass filtering, and the humidity data starts a diffusion model established based on Fick's second law for phase compensation.

[0079] An image data processing channel. In the plastic stage, a multi-angle polarization fusion technology is used to suppress the reflection interference on the liquid surface of the concrete, and morphological filtering is used to eliminate the artifacts generated by air bubbles and floating slurry in the image. In the hardening stage, a multi-spectral frequency domain fusion technology is enabled to enhance the edge features of microcracks, and combined with the strain distribution heat map, the projection area of internal defects in the concrete on the surface is located.

[0080] Specifically, high-frequency band-pass filtering based on the sym8 wavelet basis is adopted, and the passband frequency range is set to 50 - 200 Hz to suppress the disturbance noise with a frequency greater than 500 Hz generated during operations such as vibration and pumping during pouring, and retain the high-frequency characteristic signals reflecting the early state changes of the concrete. Through dynamic wide-band pass filtering, the pulse characteristics of early shrinkage of the concrete are captured. The frequency range of this dynamic wide-band pass is dynamically adjusted according to the time after concrete pouring. The initial range is 30 - 300 Hz, and as the plastic stage progresses, the upper limit frequency gradually decreases to 200 Hz to accurately capture the early shrinkage characteristics. Baseline drift correction is performed by real-time monitoring of the environmental humidity and the change trend of the internal humidity of the concrete, establishing a reference baseline for humidity change, and using a linear regression algorithm to correct the humidity measurement values to ensure the accuracy of the humidity data.

[0081] Furthermore, the pulse characteristics of early shrinkage of the concrete are captured through dynamic wide-band pass filtering, and the frequency range of this dynamic wide-band pass is dynamically adjusted according to the time after concrete pouring. The specific adjustment process is as follows: 1) Within 0 - 1 hour after concrete pouring, the initial frequency range is set to 30 - 300 Hz.

[0082] At this stage, the concrete has just been poured and is in a state of relatively high fluidity. The internal cement particles are in a dispersed and suspended state, with sufficient moisture and relatively uniform distribution. Due to the strong fluidity of the concrete, the internal structure has not yet started to coagulate and solidify significantly, so the frequency distribution of the shrinkage characteristic signals is relatively broad.

[0083] To comprehensively capture these early shrinkage signals caused by different factors, a relatively wide frequency band is used for filtering. In actual operation, the strain signals of the concrete are collected at a sampling frequency of 1000 times per second to ensure that the high-frequency shrinkage pulse characteristics can be captured. The collected signals are converted from analog to digital and then enter a dynamic wide-band pass filter. The filter screens the signals according to the preset frequency range of 30 - 300 Hz, removing the noise signals below 30 Hz and above 300 Hz, so as to retain various possible early shrinkage signals.

[0084] 2) During the 1 - 3 hours after pouring, the upper limit frequency is decreased at a rate of 50 Hz per hour.

[0085] Understandably, as time goes by, the concrete gradually loses its fluidity, the high-frequency components of the shrinkage characteristic signals decrease, and decreasing the upper limit frequency can more accurately focus on the effective shrinkage characteristic frequency band. For example, 2 hours after pouring, the frequency range is adjusted to 30 - 250 Hz. At this time, the system will automatically adjust the parameters of the dynamic wide-band pass filter according to the progress of time, decreasing the upper limit frequency from 300 Hz to 250 Hz. This can more accurately focus on the effective shrinkage characteristic frequency band and reduce the interference of high-frequency noise on the shrinkage characteristic signals.

[0086] 3) Within 3 - 6 hours after pouring, continue to decrease the upper limit frequency at a rate of 25 Hz per hour until the upper limit frequency reaches 200 Hz.

[0087] At this stage, the plasticity of the concrete gradually decreases, the internal structure further solidifies and stabilizes, and the shrinkage characteristics are more stable, and the rising rate can be slow.

[0088] 4) After the upper limit frequency reaches 200 Hz, finally fix the frequency range at 30 - 200 Hz to accurately capture the early shrinkage characteristics of this stage.

[0089] This stage will continuously filter the strain signals of the concrete with this fixed frequency range to ensure that the characteristic information related to early shrinkage can be accurately extracted.

[0090] It should be added that the low-frequency band-pass filtering based on the db6 wavelet basis is switched to, and the cut-off frequency is set to be less than 10 Hz, which is convenient for extracting the trend characteristics of the temperature rise during concrete hydration later, and separating the long-period and low-frequency temperature change signals generated by the cement hydration reaction. The passband frequency range of the narrow-band pass filtering is 2 - 5 Hz, separating the response component caused by creep and removing the strain fluctuations caused by other interference factors. The diffusion model established based on Fick's second law takes into account factors such as the pore structure and humidity gradient of the hardened concrete body, compensates for the hysteresis effect in humidity measurement, and accurately reflects the humidity change inside the concrete.

[0091] Furthermore, the multi-angle polarization fusion technology is adopted to collect multi-spectral images from different polarization angles, and the weighted average algorithm is used to fuse the images of each angle to reduce the influence of surface reflected light on the image quality. The weights can be set correspondingly in combination with empirical data. The morphological filtering adopts the morphological operation combined with opening and closing operations, and the size of the structural element is determined according to the image resolution and the approximate sizes of air bubbles and floating mortar, generally 3×3 to 7×7 pixels, removing small-sized noise and artifacts. The multi-spectral frequency-domain fusion technology converts the images of different spectral bands to the frequency domain, and by adjusting the weights of each band image in the frequency domain, highlights the characteristic information of micro-cracks in the frequency domain, and then converts the processed frequency-domain image back to the spatial domain to enhance the clarity of the micro-crack edges. The strain distribution heat map is generated based on the data collected by the strain sensors. The projection area of the internal defects of the concrete on the surface is located by aligning the heat map with the multi-spectral image using image registration technology. By analyzing the strain anomaly area in the heat map, the possible projection area of the internal defects on the surface can be located in the multi-spectral image.

[0092] In the embodiment of the present invention, the sym8 wavelet basis high-frequency filtering is used to capture transient temperature fluctuations in the plastic stage, and the db6 wavelet basis low-frequency filtering is switched to in the hardening stage to suppress noise. By combining the stage-differentiated signal processing, the problem of feature loss caused by a fixed wavelet basis such as db4 is solved. The recognition accuracy of the intervention window in the plastic stage is significantly improved, and the sensitivity in the hardening stage is also greatly improved. By integrating the four-dimensional data of temperature-humidity-strain-multi-spectral, the interference of surface floating mortar is eliminated through the polarization filter, morphological filtering and frequency-domain fusion technology, and the projection positioning error of internal defects is significantly reduced. At the same time, through the distributed gradient layout of humidity sensors combined with the diffusion model compensation, the coupling effect of drying shrinkage and temperature strain can be identified, which is convenient for cause identification and also reduces the cause misjudgment rate.

[0093] Exemplarily, for further analysis, the signal-to-noise ratio, early warning lead time, crack misjudgment rate, strain sensitivity, hardware survival rate, cross-modal response time and long-term monitoring accuracy attenuation are used as comparative technical indicators for comparison. The specific comparison results are shown in Table 1.

[0094] Table 1 Schematic Table of Comparison of Effects of This Technical Solution

[0095]

[0096] The crack risk analysis module divides the risk level based on the number, width, and location of cracks and the set risk level division rules.

[0097] Specifically, the set risk level division rules are specifically implemented as follows: Extract the concrete apparent image from the multispectral image data, and accordingly define the key stress-bearing areas, secondary key stress-bearing areas, and non-critical areas of the concrete.

[0098] If the number of cracks and the maximum crack width are less than the corresponding first trigger thresholds, and at the same time the crack positions are all in the non-critical areas, a first-level risk level is triggered.

[0099] If the number of cracks or the maximum crack width exceeds the corresponding first trigger threshold and is less than or equal to the corresponding second trigger threshold, or the crack position is in the secondary key stress-bearing area, a second-level risk level is triggered.

[0100] If the number of cracks or the maximum crack width exceeds the corresponding second trigger threshold, or the crack position is in the key stress-bearing area, a third-level risk level is triggered.

[0101] It should be added that the first trigger thresholds corresponding to the number of cracks and the maximum crack width can be taken as 2 and 0.1 mm respectively, and the second trigger thresholds corresponding to the number of cracks and the maximum crack width can be taken as 5 and 0.3 mm respectively. Non-critical areas such as non-load-bearing walls and decorative surfaces, secondary key stress-bearing areas such as the edges of floors and non-node areas of shear walls, and key stress-bearing areas such as beam-column joints, the roots of load-bearing walls, and prestressed anchorage areas.

[0102] Please refer to Figure 5 As shown, the crack early warning processing module consists of a multi-level disposal strategy library and a dynamic early warning unit.

[0103] The multi-level disposal strategy library stores disposal plans linked to the risk level.

[0104] The dynamic early warning unit calls the disposal plan corresponding to the risk level in the multi-level disposal strategy library to process cracks, and triggers a progressive early warning based on the curve. When the curve slope exceeds the threshold, a crack risk emergency response is initiated.

[0105] Specifically, the specific trigger of the progressive early warning is as follows: Calculate the real-time slope of the crack evolution risk rate growth curve by the central difference method, denoted as the risk change slope , denote the real-time fission risk rate in the curve as , represents the time point number, .

[0106] Locate the starting time point and the ending time point from the growth curve of the crack evolution risk rate, form a time interval, and calculate the integral value of the growth curve of the crack evolution risk rate within the time interval, which is denoted as the cumulative risk amount and denoted as .

[0107] Based on , and and the pre-set warning trigger rules to trigger the corresponding warning. The warning trigger rules are as follows: If there is and , and at the same time , trigger a level-I warning, and are respectively set as the first risk rate threshold and the second risk rate threshold, and are respectively set as the first risk change slope and the second risk amount threshold.

[0108] If there is or or , trigger a level-II warning, is set as the third risk rate threshold, and are respectively set as the second risk change slope and the second risk amount threshold.

[0109] If there is or or , trigger a level-II warning.

[0110] When the risk rate change slope continuously exceeds the set value within multiple time windows, automatically raise the warning level.

[0111] In the embodiment of the present invention, the multi-parameter collaborative sensing module is used to collect temperature, humidity, strain and multi-spectral image data in real time from the plastic to the hardening stage, and combined with the stage-differentiated signal processing and the deep convolutional neural network to extract crack features, significantly improving the detection accuracy of micro-cracks and internal defects. At the same time, by fusing multi-physical field data, an innovative crack evolution risk rate model is constructed to realize the dynamic probability prediction before the crack initiation, and through the risk level division and the progressive warning mechanism, the hierarchical disposal plan is automatically triggered, effectively distinguishing the causes of drying shrinkage and temperature cracks, and significantly improving the prevention and control efficiency.

[0112] Please refer to Figure 2 as shown. The present invention also provides a method for detecting construction cracks in high-strength concrete. The method includes: A1. Multi-parameter collaborative data collection: Collect temperature, humidity, strain and multi-spectral image data in real time from the plastic stage to the hardening stage of the concrete.

[0113] A2. Construction crack identification: A deep convolutional neural network is used to extract crack features from multispectral images, and the number, width and location of cracks are output. When no cracks are detected, the temperature, humidity and strain data are integrated to output the crack evolution risk rate in the time series, and a crack evolution risk rate growth curve is constructed.

[0114] A3. Crack risk analysis: Risk level classification is carried out based on the number, width and location of cracks and the set risk level classification rules.

[0115] A4. Crack early warning processing: Based on the risk level, call the treatment plan corresponding to the risk level in the multi-level treatment strategy library to handle the cracks, and trigger a progressive early warning based on the crack evolution risk rate growth curve, and initiate a crack risk emergency response when the slope of the curve exceeds the threshold.

[0116] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they shall all fall within the protection scope of the present invention.

Claims

1. A high-strength concrete construction crack detection system, characterized in that Comprising: A multi-parameter collaborative sensing module for real-time collection of temperature, humidity, strain and multi-spectral image data from the plastic stage to the hardening stage of concrete; A construction crack identification module composed of an image analysis unit and a risk prediction unit; The image analysis unit uses a deep convolutional neural network to extract crack features from multi-spectral images and outputs the number, width and location of cracks; The risk prediction unit, when no cracks are detected, synthesizes temperature, humidity and strain data, outputs the crack evolution risk rate in the time series, and constructs a crack evolution risk rate growth curve; A crack risk analysis module for risk level classification based on the number, width and location of cracks and the set risk level classification rules; A crack early warning processing module composed of a multi-level disposal strategy library and a dynamic early warning unit; The multi-level disposal strategy library stores disposal plans linked to risk levels; The dynamic early warning unit calls the disposal plan corresponding to the risk level in the multi-level disposal strategy library to process cracks, and triggers a progressive early warning based on the curve. When the curve slope exceeds the threshold, a crack risk emergency response is initiated.

2. The high-strength concrete construction crack detection system according to claim 1, wherein: The multi-parameter collaborative sensing module includes: The distributed optical fiber temperature sensor array embedded in the concrete along the three-dimensional orthogonal grid of the concrete structure is used to collect the three-dimensional distribution data of the temperature field from the plastic stage to the hardening stage in real time, denoted as temperature data; Capacitive humidity sensor arrays are installed in a gradient distribution on the surface layer, 1 / 2 thickness layer and bottom layer of the concrete structure to collect humidity data from the plastic stage to the hardening stage in real time; FBG strain sensor arrays are orthogonally arranged at the beam ends and plate corners of the concrete to measure the axial strain and shear strain from the plastic stage to the hardening stage in real time, and the strain data is integrated; Through a multi-spectral imager mounted on a construction robot, visible light and near-infrared are used for dual-band scanning in real time, and a three-way polarization filter is combined to eliminate the reflection interference of the surface floating slurry, and multi-spectral image data corresponding to the elimination of interference from the plastic stage to the hardening stage is obtained.

3. The high-strength concrete construction crack detection system according to claim 1, characterized in that: The construction crack identification module further includes a signal preprocessing unit, and its specific configuration is as follows: The time series data processing channel performs the following stage-differentiated processing on temperature, humidity and strain data: In the plastic stage, the temperature data is filtered by a high-frequency band-pass filter using the sym8 wavelet basis, the strain data is filtered by a dynamic wide-band filter, and the humidity data is corrected for baseline drift; In the hardening stage, the temperature data is switched to the db6 wavelet basis for low-frequency band-pass filtering, the strain data is filtered by a narrow-band filter, and the humidity data starts a diffusion model based on Fick's second law for phase compensation; The image data processing channel, in the plastic stage, uses the multi-angle polarization fusion technology to suppress the reflection interference of the liquid surface of the concrete, and eliminates the artifacts generated by bubbles and floating slurry in the image through morphological filtering. In the hardening stage, the multi-spectral frequency domain fusion technology is enabled to enhance the edge features of micro-cracks, and combined with the strain distribution heat map, the projection area of the internal defects of the concrete on the surface is located.

4. The high-strength concrete construction crack detection system according to claim 1, characterized in that: The output of the crack evolution risk rate in the time series is used to output the crack evolution risk rate in the corresponding time series of the plastic stage, including: For the plastic stage, traverse each temperature sensor, calculate the absolute value of the temperature difference between it and all adjacent points, and take the maximum absolute temperature difference as the target temperature difference; Perform power spectral density integration calculations on the stress signal in the beam end direction and the shear strain signal in the orthogonal direction respectively, output the axial strain energy and the shear strain energy, calculate the comprehensive axial strain energy, and calculate the total energy ratio; Calculate the standard deviations of the humidity data corresponding to the surface layer, 1 / 2 thickness layer, and bottom layer of the concrete structure respectively, and perform weighted summation calculation to output the target humidity fluctuation degree; Statistically obtain the microtrace density, bleeding channel length, aggregate dispersion, and interference fringe intensity from the multispectral image data, and construct an image feature vector; For each acquisition time point, comprehensively calculate the crack evolution risk rate based on the target temperature difference, total energy ratio, target humidity fluctuation degree, and image feature vector, and then construct the crack evolution risk rate corresponding to the time series in the plastic stage.

5. The high-strength concrete construction crack detection system according to claim 4, characterized in that: The output crack evolution risk rate under the time series includes: Based on the crack evolution risk rate corresponding to the time series in the plastic stage, set the risk compensation factor for the hardening stage; Based on the temperature, humidity, strain, and multispectral image data in the hardening stage, calculate the temperature rise rate, strain acceleration value, humidity gradient value, and polarization interference fringe density in each time window of the hardening stage; Normalize the temperature rise rate, strain acceleration value, humidity gradient value, and polarization interference fringe density, and obtain the crack evolution risk rate of the hardening stage in each time window through weighted summation; Correct the crack evolution risk rate of the hardening stage based on the risk compensation factor for the hardening stage, and output the crack evolution risk rate under the time series based on the corrected crack evolution risk rate values in each time window.

6. The high-strength concrete construction crack detection system according to claim 5, wherein: The setting of the risk compensation factor for the hardening stage includes: Select the maximum crack evolution risk rate from the crack evolution risk rate corresponding to the time series in the plastic stage; Statistically record the duration during which the crack evolution risk rate is greater than the set reference interference value as the crack evolution interference duration; Normalize the maximum crack evolution risk rate and the crack evolution interference duration, and input them into the sigmoid function to output the risk compensation factor for the hardening stage.

7. The high-strength concrete construction crack detection system according to claim 1, wherein: The specific implementation of the set risk level classification rules is as follows: Extract the concrete apparent image from the multispectral image data, and accordingly define the key stress-bearing area, secondary key stress-bearing area, and non-key area of the concrete; If the number of cracks and the maximum crack width are less than the corresponding first trigger threshold, and at the same time the crack positions are all in the non-key area, trigger the first-level risk level; If the number of cracks or the maximum crack width exceeds the corresponding first trigger threshold and is less than or equal to the corresponding second trigger threshold, or the crack position is in the secondary key stress-bearing area, trigger the second-level risk level; If the number of cracks or the maximum crack width exceeds the corresponding second trigger threshold, or the crack position is in the key stress-bearing area, trigger the third-level risk level.

8. The high-strength concrete construction crack detection system according to claim 1, wherein: The specific trigger of the progressive warning is as follows: The real-time slope of the growth curve of the crack evolution risk rate is calculated by the central difference method, denoted as the risk change slope , and the real-time fission risk rate in the curve is denoted as , represents the time point number, ; Locate the starting time point and the ending time point from the growth curve of the crack evolution risk rate to form a time interval, and calculate the integral value of the growth curve of the crack evolution risk rate within the time interval, which is denoted as the cumulative risk amount, denoted as ; Based on 、 and and trigger corresponding warnings according to the preset warning trigger rules.

9. The high-strength concrete construction crack detection system according to claim 8, characterized in that: The specific warning trigger rules are as follows: If there exists and At the same time , trigger a level-I early warning, and are respectively set as the first risk rate threshold and the second risk rate threshold, and are respectively set as the first risk change slope and the second risk quantity threshold; If there exists or or , a level II early warning is triggered. is the set third risk rate threshold, and are the set second risk change slope and second risk quantity threshold respectively; If there exists or or else , a level-II early warning will be triggered; When the slope of the risk rate changes Continuously exceeds the set value within multiple time windows, the warning level is automatically raised.

10. A method for detecting construction cracks in high-strength concrete, characterized in that: The method includes: A1. Multi-parameter collaborative data acquisition: Real-time acquisition of temperature, humidity, strain, and multispectral image data from the plastic stage to the hardening stage of the concrete; A2. Construction crack identification: Use a deep convolutional neural network to extract crack features from multi-spectral images, output the number, width, and location of cracks, and when no cracks are detected, comprehensively consider temperature, humidity, and strain data, output the crack evolution risk rate under the time series, and construct a crack evolution risk rate growth curve; A3. Crack risk analysis: Based on the number, width, and location of cracks and the set risk level classification rules, conduct risk level classification; A4. Crack early warning and treatment: Based on the risk level, call the treatment plan corresponding to the risk level in the multi-level disposal strategy library to handle cracks, trigger a progressive early warning based on the crack evolution risk rate growth curve, and initiate a crack risk emergency response when the curve slope exceeds the threshold.

Images