Grouting quality detection method, system, electronic device, medium and program product
By combining a twin network model with grouting pressure and flow rate sequences, the difference and geological similarity of grouting anchor points are calculated, which solves the problem of inaccurate grouting quality evaluation and achieves a more accurate assessment of grouting effect.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HUANENG QINGYANG COAL POWER CO LTD HETAOYU COAL MINE
- Filing Date
- 2021-12-15
- Publication Date
- 2026-06-16
AI Technical Summary
Existing technologies are not accurate in evaluating grouting quality. Conventional monitoring methods, such as permeability measurement and the PQt method, have low accuracy and are difficult to effectively assess the grouting effect.
Using a twin network model, ground-penetrating radar images and standard images of grouting anchor points are acquired after grouting. Combined with grouting pressure and flow rate sequences, the difference and geological similarity are calculated, the contrast loss function is adjusted, and the grouting quality is evaluated.
It improves the accuracy of grouting quality assessment, comprehensively considers the influence of geological and grouting factors, and enhances the precision of grouting analysis.
Smart Images
Figure CN114399469B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of grouting quality evaluation technology, specifically to a grouting quality testing method, system, electronic equipment, medium, and program product. Background Technology
[0002] Grouting technology has become the preferred technical means to solve many problems such as water hazards in engineering projects, water seepage in mines, karst filling, soft rock reinforcement, and leakage detection of hydropower dams. However, how to effectively evaluate its engineering effects has not yet been reasonably resolved.
[0003] Conventional grouting monitoring typically uses traditional methods such as permeability measurement, geophysical exploration, or the PQt method, which have low accuracy. Summary of the Invention
[0004] To address the aforementioned technical problems, the present invention aims to provide a grouting quality detection method, system, electronic device, medium, and program product.
[0005] This invention provides a method for detecting grouting quality, comprising the following steps:
[0006] Acquire post-grouting ground-penetrating radar images of each grouting anchor point in the target grouting area and a standard post-grouting ground-penetrating radar image of the target grouting area;
[0007] The ground-penetrating radar image after grouting at any grouting anchor point and the standard ground-penetrating radar image after grouting are input into the first twin network model to obtain the grouting quality index of any grouting anchor point.
[0008] The grouting quality of the target grouting area is determined based on the grouting quality indicators of each grouting anchor point.
[0009] The contrastive loss of the first Siamese network model is determined based on the following steps:
[0010] During grouting, the grouting pressure sequence and grouting flow rate sequence of each grouting anchor point in the target grouting area are collected. Based on the grouting pressure sequence and grouting flow rate sequence of each grouting anchor point, the difference degree between any two grouting anchor points is calculated. Based on the difference degree between any two grouting anchor points, the difference degree sequence of each grouting anchor point is determined.
[0011] Based on the difference sequence, each grouting anchor point is paired to obtain several matching pairs;
[0012] The consistency of the grouting process of any matching pair is calculated based on the grouting pressure sequence and grouting flow rate sequence of each grouting anchor point within any matching pair.
[0013] Based on the ground-penetrating radar images before grouting corresponding to each grouting anchor point of any matching pair, the geological similarity between the two ground-penetrating radar images before grouting is determined.
[0014] The contrast loss of any matching pair is determined based on the degree of consistency and the geological similarity.
[0015] Furthermore, the contrastive loss of the first twin network model is:
[0016] LOSS=U i *loss*J i ;
[0017] Where loss is a preset contrastive loss function, U i J represents the consistency of the i-th matching pair. i Let be the geological similarity of the i-th matching pair, where each matching pair includes two grouting anchor points.
[0018] Furthermore, the formula for calculating the degree of consistency is as follows:
[0019]
[0020] Where, p A This represents the sequence of grouting pressure changes during grouting at the A-th grouting anchor point within the matching pair, v A p represents the sequence of grouting flow rate changes during grouting at the A-th grouting anchor point within the matching pair. B This represents the sequence of grouting pressure changes during grouting at the B-th grouting anchor point within the matching pair, v B This represents the grouting flow rate change sequence during grouting at the Bth grouting anchor point within the matching pair. DTW indicates the similarity between the two sequences, and RT is the rise time calculation function.
[0021] Further, determining the geological similarity between two pre-grouting ground-penetrating radar images based on the pre-grouting ground-penetrating radar images corresponding to each grouting anchor point of any matching pair includes:
[0022] The ground-penetrating radar images before grouting corresponding to each grouting anchor point of any matching pair are input into the second twin network model to obtain the geological similarity of the two ground-penetrating radar images before grouting output by the second twin network model.
[0023] Furthermore, the standard post-grouting ground-penetrating radar image is a post-grouting ground-penetrating radar image that meets the grouting quality standards under similar geological conditions.
[0024] Furthermore, the formula for calculating the degree of difference is:
[0025]
[0026] Where X represents the difference distance between the a-th grouting anchor point and the b-th grouting anchor point, p a v represents the sequence of grouting pressure changes during grouting at the a-th grouting anchor point. a p represents the sequence of grouting flow rate changes during grouting at the a-th grouting anchor point. b v represents the sequence of grouting pressure changes during grouting at the b-th grouting anchor point. b DTW represents the sequence of grouting flow rate changes during grouting at the b-th grouting anchor point, RT represents the similarity between the two sequences, and RT is the rise time calculation function.
[0027] The present invention also provides a grouting quality detection system, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of any of the grouting quality detection methods described above.
[0028] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of any of the above-described grouting quality detection methods.
[0029] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the above-described grouting quality detection methods.
[0030] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of any of the above-described grouting quality detection methods.
[0031] The present invention has the following beneficial effects:
[0032] The grouting quality detection method, system, electronic equipment, medium, and program products provided by this invention, through the establishment of a twin network model, use the acquired post-grouting ground-penetrating radar images and standard post-grouting ground-penetrating radar images of each grouting anchor point as input to the twin network model to obtain the grouting quality index of each grouting anchor point. This grouting quality index is the similarity between the post-grouting ground-penetrating radar image of the corresponding grouting anchor point and the standard post-grouting ground-penetrating radar image, which serves as the quality index for the current grouting anchor point and is used to judge the grouting quality of that anchor point. The loss function of the twin network model incorporates the consistency during grouting and the geological similarity before grouting, used to correct the contrast loss function. This adjusts the impact of different geological conditions and grouting factors on the grouting quality of the grouting area, allowing for a more comprehensive consideration of grouting quality, facilitating the analysis of grouting quality, further improving the accuracy of grouting analysis, and ultimately enhancing the accuracy of grouting quality assessment. Attached Figure Description
[0033] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments of the present invention or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0034] Figure 1 This is a schematic flowchart of the grouting quality testing method provided by the present invention;
[0035] Figure 2 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation
[0036] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following detailed description, in conjunction with the accompanying drawings and preferred embodiments, provides a detailed account of the specific implementation methods, structures, features, and effects of the solutions according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.
[0037] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.
[0038] The following section uses a grouting area in a certain location as an example, specifically the target grouting area, to illustrate the specific scheme of the grouting quality detection method provided by this invention in conjunction with the accompanying drawings.
[0039] Specifically, please refer to Figure 1 The diagram illustrates a flowchart of a grouting quality testing method according to an embodiment of the present invention, which includes the following steps:
[0040] Step 110: Obtain post-grouting ground-penetrating radar images of each grouting anchor point in the target grouting area and a standard post-grouting ground-penetrating radar image of the target grouting area.
[0041] The ground-penetrating radar images in this embodiment of the invention are obtained by ground-penetrating radar detection.
[0042] Specifically, in this embodiment, the target grouting area is pre-divided into multiple partitions, and a corresponding grouting anchor point is set for each partition, with one grouting anchor point corresponding to one partition; radar survey lines are arranged at each grouting anchor point in the grouting area to detect each radar wave under the grouting area, so as to obtain the ground-penetrating radar image of each grouting anchor point before grouting and the ground-penetrating radar image of each grouting anchor point after grouting.
[0043] It should be noted that, for the target grouting area, the distribution of underground soil layers varies, so the specific distribution of grouting anchor points in different zones needs to be determined by the implementer based on the local conditions. This embodiment of the invention does not limit this.
[0044] Before determining the ground-penetrating radar image before grouting and the ground-penetrating radar image after grouting (i.e., the ground-penetrating radar image after grouting), the collected radar waves are processed, such as filtering, removing direct waves, background removal, and linear gain. Since the signal processing method of radar waves is a known method of information processing, it will not be elaborated on here.
[0045] During the detection radar measurement, the parameters of the detection radar (time window parameters, antenna frequency, scanning speed, number of sampling points, resolution) need to be adjusted according to the medium characteristics in the detection area and the depth to be detected. The specific adjustment method is well known and will not be described here.
[0046] In this embodiment, the standard post-grouting ground-penetrating radar (GPR) image is a GPR image acquired under similar geological conditions that meets the grouting quality standards. That is, the standard GPR image is a GPR image acquired under similar geological conditions showing good grouting quality. The standard GPR image can be determined based on historical data, or it can be selected from normal post-grouting GPR images acquired from different zones to obtain GPR images that meet the grouting quality standards. The similar geological conditions refer to geological conditions similar to those corresponding to the GPR images of each grouting anchor point.
[0047] In addition, each target grouting area corresponds to a standard post-grouting ground-penetrating radar image.
[0048] Of course, as another implementation, the standard ground-penetrating radar image after grouting of the present invention can also be obtained through the following steps: calculating the cosine similarity of the normal ground-penetrating radar images after grouting in different zones, and obtaining the sum of the similarity between each normal ground-penetrating radar image and the other remaining normal ground-penetrating radar images respectively; then, based on the sum of the similarity of each normal ground-penetrating radar image, determining the normal ground-penetrating radar image corresponding to the highest sum of similarity, and using the normal ground-penetrating radar image corresponding to the highest sum of similarity as the standard ground-penetrating radar image after grouting.
[0049] Step 210: Input the post-grouting ground-penetrating radar image of any grouting anchor point and the standard post-grouting ground-penetrating radar image into the first twin network model to obtain the grouting quality index of any grouting anchor point.
[0050] The training process of the first Siamese network model mentioned above is as follows:
[0051] 1) Acquiring Tag Data: Using radar detection technology, acquire several ground-penetrating radar images after grouting. Add tags to each echo image (i.e., ground-penetrating radar image after grouting) to identify echo images where the grouting result is normal (I). normal Echo image I of abnormal grouting results unusal Echo image I showing normal grouting results normal Echo image I of abnormal grouting results unusal As label data.
[0052] The distinction between normal echo images (echo images with normal grouting results) and abnormal echo images (echo images with abnormal grouting results) mentioned above is based on the following: If the grouting quality of the grouting anchor point is good, that is, its radar image shows relatively uniform gloss changes in depth, such images are labeled as normal grouting and recorded as echo images with normal grouting results (I). normal If voids appear at the grouting anchor points, indicating poor grouting quality, a bright area will appear at a certain depth in the radar image. This type of image is labeled as having poor grouting quality and recorded as an echo image (I) indicating abnormal grouting results. unusal .
[0053] 2) Input the labeled data into the constructed first Siamese network model, and train the constructed first Siamese network model by combining the improved contrastive loss to obtain the trained first Siamese network model.
[0054] The contrastive loss of the first Siamese network model is determined based on the following steps:
[0055] During grouting, the grouting pressure sequence and grouting flow rate sequence of each grouting anchor point in the target grouting area are collected. Based on the grouting pressure sequence and grouting flow rate sequence of each grouting anchor point, the difference degree between any two grouting anchor points is calculated. Based on the difference degree between any two grouting anchor points, the difference degree sequence of each grouting anchor point is determined.
[0056] Based on the difference sequence, each grouting anchor point is paired to obtain several matching pairs;
[0057] The consistency of the grouting process of any matching pair is calculated based on the grouting pressure sequence and grouting flow rate sequence of each grouting anchor point within any matching pair.
[0058] Based on the ground-penetrating radar images before grouting corresponding to each grouting anchor point of any matching pair, the geological similarity between the two ground-penetrating radar images before grouting is determined.
[0059] The contrast loss of any matching pair is determined based on the degree of consistency and the geological similarity.
[0060] The contrastive loss of the first Siamese network model, i.e. the improved contrastive loss, is:
[0061] LOSS=U i *loss*J i ;
[0062] Where loss is a preset contrastive loss function, U i J represents the consistency of the i-th matching pair. i Let be the geological similarity of the i-th matching pair, where each matching pair includes two grouting anchor points.
[0063] Since the preset contrastive loss function is an existing technology, namely the inherent contrastive loss function of Siamese networks, it will not be introduced here.
[0064] Among them, the degree of consistency U i J represents the consistency of the grouting process between the two grouting anchors in the i-th matching pair, and is the loss weight applied to the sample matching within this matching pair. i This represents the similarity between the two grouting anchor points in the i-th matching pair before grouting in the underground soil layer. It affects the difference between each feature in the imaging result. LOSS should converge to 0 between normal samples, thus enabling better identification of normal grouting conditions.
[0065] The improved contrast loss function introduced above adjusts the contrast loss by incorporating two parameters: grouting parameters and geological environment parameters. This allows for a more comprehensive assessment of the grouting quality in the grouting area, thereby improving the accuracy of the assessment. Here, the grouting parameters include, but are not limited to, grouting pressure and grouting flow rate.
[0066] It should be noted that the above contrastive loss is for a single matching pair. When there are multiple matching pairs, the corresponding contrastive loss should be calculated for each matching pair. Therefore, during training, it is necessary to consider the contrastive loss for all matching pairs.
[0067] Among them, the degree of consistency U i The method for obtaining it is as follows:
[0068] 1) Collect the grouting pressure sequence and grouting flow rate sequence of each grouting anchor point in the target grouting area during the grouting process, and calculate the difference between any two grouting anchor points based on the grouting pressure sequence and grouting flow rate sequence of each grouting anchor point. Based on the difference between any two grouting anchor points, determine the difference sequence of each grouting anchor point.
[0069] When grouting the anchor points of each zone, the grouting pressure change sequence p and the grouting flow rate change sequence v over time are collected, thus obtaining the grouting data of each zone during the grouting process: grouting pressure sequence p and grouting flow rate sequence v.
[0070] Based on the grouting pressure sequence p and grouting flow rate sequence v of each grouting anchor point, the difference X between any two grouting anchor points is obtained. The formula for calculating the difference X is as follows:
[0071]
[0072] Where X represents the difference distance between the a-th grouting anchor point and the b-th grouting anchor point; p a v represents the sequence of grouting pressure changes during grouting at the a-th grouting anchor point. a p represents the sequence of grouting flow rate changes during grouting at the a-th grouting anchor point. b v represents the sequence of grouting pressure changes during grouting at the b-th grouting anchor point. b This represents the sequence of grouting flow rate changes during grouting at the b-th grouting anchor point. DTW indicates the similarity between two sequences; the closer the changes are, the smaller the DTW value and the smaller the difference distance. Conversely, the less similar the changes are, the larger the DTW value and the larger the difference distance.
[0073] The RT function mentioned above is a rise time calculation function, used to statistically analyze the time it takes for pressure to rise during grouting. Its calculation method can refer to the method for calculating the rise time of a pulse signal in electrical engineering (the interval between the initial arrival of the pulse instantaneous value at a specified lower and upper limit. Unless otherwise specified, the lower and upper limits are set at 10% and 90% of the pulse peak amplitude, respectively). By using a rise time correction, the similarity between the two grouting anchor points can be better reflected.
[0074] The Max and Min functions represent functions that define the maximum and minimum values, respectively.
[0075] The difference in similarity between the grouting flow rate change sequence and the grouting pressure change sequence between the two grouting anchor points is due to the difference in the required accuracy of the data comparison. The grouting flow rate change is relatively small compared to the grouting pressure change, so the exponential function can fall quickly, while the inverse proportional function falls slower. Therefore, the exponential function with a faster falling speed is chosen for the grouting flow rate change, while the inverse proportional function is chosen for the grouting pressure change.
[0076] 2) Based on the difference sequence of each grouting anchor point, pair each grouting anchor point to obtain several matching pairs;
[0077] In this embodiment, the difference sequence of each grouting anchor point is sorted by size, and one of the grouting anchor points is matched with other grouting anchor points using KM optimal pairing to obtain the matching pair of the grouting anchor point; in the same way, the remaining grouting anchor points are matched; thus, the matching results of all grouting anchor points are obtained, that is, there are multiple matching pairs.
[0078] Furthermore, this embodiment also includes a matching verification of the paired grouting anchor points to determine the matching accuracy of the two grouting anchor points.
[0079] Specifically, in this embodiment, the tag data of the post-grouting ground-penetrating radar images corresponding to each grouting anchor point obtained above is used for matching verification. That is, if the feature tags of two post-grouting anchor points within the matching pair are different, i.e., there is one tag I... normal A label is I unusal This situation is relatively rare. Therefore, when optimizing consistency and geological similarity, the magnitude of the contrast loss should be divergent, i.e., the difference between the two should be amplified. If both samples in a matching pair are anomalous, the matching pair is split and recombined with samples from other normal matching pairs to form new matching pairs for Siamese network training.
[0080] 3) Calculate the consistency of the grouting process of any matching pair based on the grouting pressure sequence and grouting flow sequence of each grouting anchor point.
[0081] The formula for calculating this degree of consistency is:
[0082]
[0083] In the formula, p A This represents the sequence of grouting pressure changes during grouting at the A-th grouting anchor point within the matching pair, v A p represents the sequence of grouting flow rate changes during grouting at the A-th grouting anchor point within the matching pair. B This represents the sequence of grouting pressure changes during grouting at the B-th grouting anchor point within the matching pair, v B DTW represents the sequence of grouting flow rate changes during grouting at the Bth grouting anchor point within the matching pair, and indicates the degree of similarity between the two sequences.
[0084] The RT function mentioned above is the rise time calculation function, and the Max and Min functions represent the maximum and minimum values, respectively.
[0085] In this embodiment, the difference in the grouting process between the two grouting anchor points within the matching pair is taken into account, which will result in a certain difference in the quality after grouting. That is, it is determined that the current parameters (i.e., grouting parameters: grouting pressure and grouting flow rate) will affect the grouting quality. If these parameters are not added, the change in grouting quality may be caused by other factors, such as an unsuitable water-cement ratio in the grout, which may cause the difference in grouting quality.
[0086] The method for obtaining geological similarity is as follows:
[0087] Based on the ground-penetrating radar images before grouting corresponding to each grouting anchor point of any matching pair, the geological similarity between the two ground-penetrating radar images before grouting is determined.
[0088] Specifically, ground-penetrating radar (GPR) images before grouting are acquired, and GPR images corresponding to each grouting anchor point of any matching pair are selected before grouting to obtain the geological similarity between the two GPR images before grouting. That is, using pre-acquired GPR images before grouting, combined with matching pairs, GPR images corresponding to each grouting anchor point of the matching pair are selected before grouting, and the feature vectors R of the two images are extracted using a CNN network. A and R B .
[0089] Calculate the geological similarity between the two grouting anchor points within any matching pair:
[0090]
[0091] Among them, R A Let R be the geological feature vector of the A-th grouting anchor point within any matching pair. BTo match the geological feature vector of the Bth grouting anchor point, the abs function is an absolute value function. The closer the two are, the smaller the abs function becomes, and the larger the value of J becomes.
[0092] The SIM function mentioned above is the cosine similarity function, which compares the degree of difference between two feature vectors. The more similar the two are, the closer the value is to 1, and the greater the difference, the closer it is to 0.
[0093] Of course, as another implementation, the present invention can also input the ground-penetrating radar images before grouting corresponding to each grouting anchor point of any matching pair into the second twin network model to obtain the geological similarity of the two ground-penetrating radar images before grouting output by the second twin network model.
[0094] The second Siamese network model mentioned above was trained using an existing contrastive loss function.
[0095] Specifically, the training process of the second Siamese network model is as follows:
[0096] The ground-penetrating radar image Pic_p obtained before grouting is mirrored to obtain Pic_p'. The mirrored result has the same characteristics as the ground-penetrating radar image obtained before grouting, thus achieving a double sample amplification effect.
[0097] Then, Pic_p and Pic_p' are input into the second Siamese network model to train the Siamese network to output the sample similarity between Pic_p and Pic_p', which is close to 1.
[0098] The training process of the Siamese network employs contrastive loss; since the method for training the Siamese network is a classic contrastive learning strategy, it will not be elaborated upon here.
[0099] In this embodiment, the same operation is performed on the radar detection maps corresponding to all grouting anchor points to continue training the twin network, thereby enabling it to learn the characteristic differences of different geology.
[0100] This invention, through obtaining the degree of geological difference J between two grouting anchor points within a matching pair, reflects the geological difference information between the two grouting anchor points and adjusts the impact of geological differences on post-grouting imaging. Simultaneously, due to differences in geological structural characteristics, such as rock strata and sand layers, the grouting pressure and flow rate during the grouting process will differ, resulting in variations.
[0101] Step 310: Determine the grouting quality of the target grouting area based on the grouting quality indicators of each grouting anchor point.
[0102] Specifically, when the grouting quality index of a grouting anchor point is greater than or equal to a set threshold, the grouting quality of that grouting anchor point is good; when the grouting quality index of a grouting anchor point is less than the set threshold, the grouting quality of that grouting anchor point after grouting is not good.
[0103] In this embodiment, the threshold value is set to 0.9. That is, when the grouting quality index reaches 0.9 or higher, it proves that the post-grouting echo map of the grouting area corresponding to the current grouting anchor point is similar to the standard echo map, indicating that the grouting quality of the grouting area meets the standard. The threshold value mentioned above is not limited to 0.9; the implementer can determine it according to the grouting requirements.
[0104] The embodiments of the present invention can train the first twin network model by introducing geological parameters and grouting parameters, so as to adjust the influence of different geological conditions and grouting factors on the grouting quality of the grouting area, making it more comprehensively consider the grouting quality, which helps in the analysis of grouting quality, thereby improving the accuracy of grouting analysis and thus improving the accuracy of grouting quality assessment.
[0105] This invention provides a grouting quality detection method. By setting up a twin network model, the method uses post-grouting ground-penetrating radar (GPR) images of each grouting anchor point and a standard post-grouting GPR image as input to obtain a grouting quality index for each anchor point. This index is the similarity between the post-grouting GPR image of the corresponding anchor point and the standard post-grouting GPR image, serving as the quality index for the current anchor point and used to determine its grouting quality. The twin network model incorporates the consistency during grouting and the geological similarity before grouting into its loss function. This loss function is used to adjust for the influence of different geological conditions and grouting factors on the grouting quality of the grouting area, allowing for a more comprehensive consideration of grouting quality. This improves the accuracy of grouting analysis and ultimately enhances the accuracy of grouting quality assessment.
[0106] The present invention also provides a technical solution for a grouting quality detection system, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the steps of the grouting quality detection method as described in any of the above claims.
[0107] Since the steps for grouting quality testing methods described above have been explained in detail, they will not be repeated here.
[0108] Figure 2 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 2As shown, the electronic device may include: a processor 210, a communication interface 220, a memory 230, and a communication bus 240, wherein the processor 210, the communication interface 220, and the memory 230 communicate with each other through the communication bus 240. The processor 210 can call logical instructions in the memory 230 to execute a grouting quality detection method. This method includes: acquiring post-grouting ground-penetrating radar images of each grouting anchor point in the target grouting area and a standard post-grouting ground-penetrating radar image of the target grouting area; inputting the post-grouting ground-penetrating radar image of any grouting anchor point and the standard post-grouting ground-penetrating radar image into a first twin network model to obtain the grouting quality index of the any grouting anchor point; and determining the grouting quality of the target grouting area based on the grouting quality index of each grouting anchor point. The contrast loss of the first twin network model is determined based on the following steps: collecting the grouting pressure sequence and grouting flow sequence of each grouting anchor point in the target grouting area during grouting. The process involves: calculating the difference between any two grouting anchors based on their grouting pressure and flow rates; determining a difference sequence for each grouting anchor based on the difference sequences; pairing the grouting anchors to obtain several matching pairs; calculating the consistency of the grouting process for any matching pair based on the grouting pressure and flow rates of each grouting anchor within that matching pair; determining the geological similarity between two pre-grouting ground-penetrating radar images based on the pre-grouting ground-penetrating radar images corresponding to each grouting anchor in any matching pair; and determining the contrast loss of any matching pair based on the consistency and geological similarity.
[0109] Furthermore, the logical instructions in the aforementioned memory 230 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0110] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the grouting quality detection method provided by the above methods. The method includes: acquiring post-grouting ground-penetrating radar images of each grouting anchor point in a target grouting area and a standard post-grouting ground-penetrating radar image of the target grouting area; inputting the post-grouting ground-penetrating radar image of any grouting anchor point and the standard post-grouting ground-penetrating radar image into a first twin network model to obtain the grouting quality index of the any grouting anchor point; and determining the grouting quality of the target grouting area based on the grouting quality index of each grouting anchor point; wherein the contrast loss of the first twin network model is determined based on the following steps. The process involves: collecting grouting pressure and flow rate sequences at each grouting anchor point in the target grouting area during grouting; calculating the difference between any two grouting anchor points based on these sequences; determining the difference sequence for each grouting anchor point based on the difference sequence; pairing the grouting anchor points according to the difference sequence to obtain several matching pairs; calculating the consistency of the grouting process for any matching pair based on the grouting pressure and flow rate sequences of each grouting anchor point within that pair; determining the geological similarity between two pre-grouting ground-penetrating radar images based on the pre-grouting ground-penetrating radar images corresponding to each grouting anchor point of any matching pair; and determining the contrast loss of any matching pair based on the consistency and geological similarity.
[0111] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the grouting quality detection method provided by the above methods. The method includes: acquiring post-grouting ground-penetrating radar images of each grouting anchor point in a target grouting area and a standard post-grouting ground-penetrating radar image of the target grouting area; inputting the post-grouting ground-penetrating radar image of any grouting anchor point and the standard post-grouting ground-penetrating radar image into a first twin network model to obtain a grouting quality index for any grouting anchor point; and determining the grouting quality of the target grouting area based on the grouting quality index of each grouting anchor point; wherein the contrast loss of the first twin network model is determined based on the following steps: collecting data from the target grouting area during grouting... The grouting pressure sequence and grouting flow rate sequence of each grouting anchor point are calculated. Based on the grouting pressure sequence and grouting flow rate sequence of each grouting anchor point, the difference degree between any two grouting anchor points is calculated. Based on the difference degree between any two grouting anchor points, the difference degree sequence of each grouting anchor point is determined. Based on the difference degree sequence, each grouting anchor point is paired to obtain several matching pairs. Based on the grouting pressure sequence and grouting flow rate sequence of each grouting anchor point in any matching pair, the consistency degree of the grouting process of any matching pair is calculated. Based on the ground penetrating radar image before grouting corresponding to each grouting anchor point of any matching pair, the geological similarity between the two ground penetrating radar images before grouting is determined. Based on the consistency degree and the geological similarity, the contrast loss of any matching pair is determined.
[0112] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.
[0113] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.
[0114] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, specific embodiments have been described above. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps described in the claims can be performed in a different order than that shown in the embodiments and still achieve the desired result. Additionally, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0115] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.
[0116] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A method for detecting grouting quality, characterized in that, Includes the following steps: Acquire post-grouting ground-penetrating radar images of each grouting anchor point in the target grouting area and a standard post-grouting ground-penetrating radar image of the target grouting area; The ground-penetrating radar image after grouting at any grouting anchor point and the standard ground-penetrating radar image after grouting are input into the first twin network model to obtain the grouting quality index of any grouting anchor point. The grouting quality of the target grouting area is determined based on the grouting quality indicators of each grouting anchor point. The contrastive loss of the first Siamese network model is determined based on the following steps: During grouting, the grouting pressure sequence and grouting flow rate sequence of each grouting anchor point in the target grouting area are collected. Based on the grouting pressure sequence and grouting flow rate sequence of each grouting anchor point, the difference degree between any two grouting anchor points is calculated. Based on the difference degree between any two grouting anchor points, the difference degree sequence of each grouting anchor point is determined. Based on the difference sequence, each grouting anchor point is paired to obtain several matching pairs; The consistency of the grouting process of any matching pair is calculated based on the grouting pressure sequence and grouting flow rate sequence of each grouting anchor point within any matching pair. Based on the ground-penetrating radar images before grouting corresponding to each grouting anchor point of any matching pair, the geological similarity between the two ground-penetrating radar images before grouting is determined. The contrast loss of any matching pair is determined based on the degree of consistency and the geological similarity.
2. The grouting quality testing method according to claim 1, characterized in that, The contrastive loss of the first twin network model is: LOSS=U i *loss*J i ; Where loss is a preset contrastive loss function, U i J represents the consistency of the i-th matching pair. i Let be the geological similarity of the i-th matching pair, where each matching pair includes two grouting anchor points.
3. The grouting quality testing method according to claim 1, characterized in that, The formula for calculating the degree of consistency is: Where, p A This represents the sequence of grouting pressure changes during grouting at the A-th grouting anchor point within the matching pair, v A p represents the sequence of grouting flow rate changes during grouting at the A-th grouting anchor point within the matching pair. B This represents the sequence of grouting pressure changes during grouting at the B-th grouting anchor point within the matching pair, v B This represents the grouting flow rate change sequence during grouting at the Bth grouting anchor point within the matching pair. DTW indicates the similarity between the two sequences, and RT is the rise time calculation function.
4. The grouting quality testing method according to claim 1, characterized in that, The determination of the geological similarity between two pre-grouting ground-penetrating radar images based on the pre-grouting ground-penetrating radar images corresponding to each grouting anchor point of any matching pair includes: The ground-penetrating radar images before grouting corresponding to each grouting anchor point of any matching pair are input into the second twin network model to obtain the geological similarity of the two ground-penetrating radar images before grouting output by the second twin network model.
5. The grouting quality testing method according to claim 1, characterized in that, The standard ground-penetrating radar image after grouting is a ground-penetrating radar image after grouting that meets the grouting quality standards and is collected under similar geological conditions.
6. The grouting quality testing method according to claim 1, characterized in that, The formula for calculating the degree of difference is: Where X represents the difference distance between the a-th grouting anchor point and the b-th grouting anchor point, p a v represents the sequence of grouting pressure changes during grouting at the a-th grouting anchor point. a p represents the sequence of grouting flow rate changes during grouting at the a-th grouting anchor point. b v represents the sequence of grouting pressure changes during grouting at the b-th grouting anchor point. b DTW represents the sequence of grouting flow rate changes during grouting at the b-th grouting anchor point, RT represents the similarity between the two sequences, and RT is the rise time calculation function.
7. A grouting quality detection system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the computer program is executed by the processor, it implements the steps of the grouting quality detection method as described in any one of claims 1-6.
8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the grouting quality detection method as described in any one of claims 1 to 6.
9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the grouting quality detection method as described in any one of claims 1 to 6.
10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the grouting quality detection method as described in any one of claims 1 to 6.