Tunnel soft rock large deformation support reliability monitoring method and system

Abstract

The application provides a tunnel soft rock large deformation support reliability monitoring method and system, through acquiring and analyzing multiple support structure data of the tunnel soft rock, extracting a first deformation feature vector, and combining a monitoring intention and a data label, intelligently selecting a suitable deformation feature conversion strategy, and converting the first deformation feature vector into a second deformation feature vector which is more representative. This method not only realizes accurate capture and conversion of the deformation feature of the support structure, but also generates a reliability prediction result for the deformation support monitoring request through reliability index decision of the converted deformation feature, effectively improves the accuracy and efficiency of the tunnel soft rock large deformation support reliability monitoring, and thus helps safe construction and operation and maintenance management of the tunnel engineering.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, in particular, relates to a tunnel soft rock large deformation support reliability monitoring method and system. BACKGROUND

[0002] In tunnel engineering, the stability of the support structure of soft rock stratum is a key factor to ensure the safety of tunnel construction and operation. Due to the characteristics of soft rock such as easy deformation and low strength, the support structure often faces the risk of large deformation during the excavation process and the later operation. Therefore, it is particularly important to monitor the reliability of the tunnel soft rock large deformation support.

[0003] The traditional tunnel support reliability monitoring method mainly relies on experience judgment, field observation and simple data analysis. These methods have limitations in dealing with complex and variable soft rock support problems. On the one hand, experience judgment is greatly influenced by subjective factors, and it is difficult to accurately reflect the actual state of the support structure; on the other hand, field observation and simple data analysis can only provide limited information, and it is difficult to fully and deeply reveal the deformation characteristics and reliability of the support structure.

[0004] With the continuous development of information technology, data-driven intelligent monitoring methods have been gradually applied to tunnel support reliability monitoring. However, existing intelligent monitoring methods mostly focus on data collection and processing, and lack flexibility and pertinence in deformation feature extraction and conversion. Different support structures may exhibit different characteristics during deformation, and existing methods often use uniform feature extraction and conversion strategies, which are difficult to adapt to the particularity of different support structures. SUMMARY

[0005] In view of the above-mentioned problems, in combination with the first aspect of the present application, the embodiments of the present application provide a tunnel soft rock large deformation support reliability monitoring method, which comprises:

[0006] Obtaining a plurality of support structure data of the tunnel soft rock corresponding to the deformation support monitoring request, and extracting a first deformation feature vector corresponding to each of the plurality of support structure data, wherein the plurality of support structure data correspond to a plurality of structure performance modes;

[0007] Obtaining at least two reference deformation feature conversion strategies, each reference deformation feature conversion strategy including at least one of the steps of feature selection and feature focusing on the first deformation feature vector;

[0008] determine a target deformation feature conversion strategy from the at least two reference deformation feature conversion strategies according to the monitoring intention corresponding to the deformation support monitoring request and a data tag corresponding to the support structure data, the data tag representing a reference tag knowledge point corresponding to the support structure data under the deformation support monitoring request;

[0009] perform deformation feature conversion on the plurality of first deformation feature vectors by using the target deformation feature conversion strategy to generate a plurality of second deformation feature vectors;

[0010] perform reliability index decision-making on the plurality of second deformation feature vectors to generate reliability index decision-making data corresponding to the deformation support monitoring request, the reliability index decision-making data representing a reliability prediction result corresponding to the deformation support monitoring request generated by using the plurality of support structure data.

[0011] In a possible implementation of the first aspect, the determining of the target deformation feature conversion strategy from the at least two reference deformation feature conversion strategies according to the monitoring intention corresponding to the deformation support monitoring request and the data tag corresponding to the support structure data includes:

[0012] loading the monitoring intention corresponding to the deformation support monitoring request, the plurality of first deformation feature vectors, and the at least two reference deformation feature conversion strategies into an artificial intelligence network that has completed knowledge learning in advance to obtain the target deformation feature conversion strategy, the artificial intelligence network being configured to perform strategy matching on the at least two reference deformation feature conversion strategies according to the monitoring intention corresponding to the deformation support monitoring request and the plurality of first deformation feature vectors, and the target deformation feature conversion strategy including at least one of a target feature selection strategy and a target feature focusing strategy.

[0013] In a possible implementation of the first aspect, the artificial intelligence network includes a target matching module and a decision-making execution module.

[0014] The loading of the monitoring intention corresponding to the deformation support monitoring request, the plurality of first deformation feature vectors, and the at least two reference deformation feature conversion strategies into the artificial intelligence network that has completed knowledge learning in advance to obtain the target deformation feature conversion strategy includes:

[0015] loading the monitoring intention corresponding to the deformation support monitoring request, the plurality of first deformation feature vectors, and the at least two reference deformation feature conversion strategies into the target matching module to obtain matching confidence degrees corresponding to the at least two reference deformation feature conversion strategies respectively, the target matching module being configured to perform the strategy matching on the at least two reference deformation feature conversion strategies according to the monitoring intention corresponding to the deformation support monitoring request and the plurality of first deformation feature vectors;

[0016] determine the target feature selection strategy according to the matching confidence degrees respectively corresponding to the at least two reference deformation feature conversion strategies;

[0017] load the monitoring intention corresponding to the deformation support monitoring request, the plurality of first deformation feature vectors, and the matching confidence degrees respectively corresponding to the at least two reference deformation feature conversion strategies to the decision execution module to obtain the target feature focusing strategy, the decision execution module being configured to determine the target feature focusing strategy matching the monitoring intention based on the plurality of first deformation feature vectors and the matching confidence degrees respectively corresponding to the at least two reference deformation feature conversion strategies.

[0018] In a possible implementation of the first aspect, the matching confidence degrees respectively corresponding to the at least two reference deformation feature conversion strategies include first matching confidence degrees respectively corresponding to a plurality of feature selection strategies.

[0019] The determining the target feature selection strategy according to the matching confidence degrees respectively corresponding to the at least two reference deformation feature conversion strategies includes:

[0020] output, from the first matching confidence degrees respectively corresponding to the plurality of feature selection strategies, a feature selection strategy with the largest value of the first matching confidence degrees as the target feature selection strategy.

[0021] In a possible implementation of the first aspect, the matching confidence degrees respectively corresponding to the at least two reference deformation feature conversion strategies include second matching confidence degrees respectively corresponding to a plurality of feature focusing strategies.

[0022] The loading the monitoring intention corresponding to the deformation support monitoring request, the plurality of first deformation feature vectors, and the matching confidence degrees respectively corresponding to the at least two reference deformation feature conversion strategies to the decision execution module to obtain the target feature focusing strategy includes:

[0023] loading the monitoring intention corresponding to the deformation support monitoring request, the plurality of first deformation feature vectors, and the second matching confidence degrees respectively corresponding to the plurality of feature focusing strategies to the decision execution module to obtain the target feature focusing strategy.

[0024] In a possible implementation of the first aspect, before the loading the monitoring intention corresponding to the deformation support monitoring request, the plurality of first deformation feature vectors, and the at least two reference deformation feature conversion strategies to the target matching module to obtain the matching confidence degrees respectively corresponding to the at least two reference deformation feature conversion strategies, the method further includes:

[0025] obtain a plurality of template support structure data corresponding to the template deformation support monitoring request, and extract a first deformation feature vector corresponding to each of the plurality of template support structure data, wherein the plurality of template support structure data correspond to a plurality of structure performance morphologies;

[0026] obtain at least two template deformation feature conversion strategies, wherein each of the template deformation feature conversion strategies comprises at least one of a template feature selection strategy and a template feature focusing strategy;

[0027] load the monitoring intention corresponding to the template deformation support monitoring request, the plurality of first deformation feature vectors, and the at least two template deformation feature conversion strategies into a template matching module, and obtain a matching confidence corresponding to each of the at least two template deformation feature conversion strategies;

[0028] load the monitoring intention corresponding to the template deformation support monitoring request, the plurality of first deformation feature vectors, and the matching confidence corresponding to each of the at least two template deformation feature conversion strategies into the decision execution module, and generate a reliability index decision data of the feature focusing strategy;

[0029] generate a target incentive mechanism network according to the reliability index decision data of the feature focusing strategy;

[0030] optimize the template matching module according to the target incentive mechanism network, and generate the target matching module.

[0031] In a possible implementation of the first aspect, the deforming and converting the plurality of first deformation feature vectors by using the target deformation feature conversion strategy to generate a plurality of second deformation feature vectors comprises:

[0032] selecting the plurality of first deformation feature vectors by using the target feature selection strategy to generate a plurality of first sub-deformation feature vectors that meet a correlation requirement, wherein the correlation requirement represents a correlation between the first deformation feature vector and the monitoring intention;

[0033] focusing the plurality of first sub-deformation feature vectors by using the target feature focusing strategy to generate the plurality of second deformation feature vectors.

[0034] In a possible implementation of the first aspect, before the obtaining the at least two reference deformation feature conversion strategies, the method further comprises:

[0035] load the plurality of first deformation feature vectors to a feature processing network which has completed knowledge learning in advance, to generate a plurality of extended deformation feature vectors, the extended deformation feature vectors corresponding to a feature space field which is more than the feature space field corresponding to the first deformation feature vectors, the feature processing network being used to extract the extended deformation feature vectors corresponding to the first deformation feature vectors.

[0036] In a possible implementation of the first aspect, after the extracting the first deformation feature vectors corresponding to the plurality of support structure data respectively, the method further includes:

[0037] obtaining vector attributes corresponding to the plurality of first deformation feature vectors respectively, the vector attributes representing the feature space field corresponding to the first deformation feature vectors;

[0038] fusing the first deformation feature vectors with the same vector attributes in the first deformation feature vectors corresponding to the plurality of support structure data respectively, to generate a target deformation feature vector.

[0039] In another aspect, the embodiment of the present application further provides a tunnel soft rock large deformation support reliability monitoring system, comprising a processor and a machine readable storage medium, the machine readable storage medium is connected with the processor, the machine readable storage medium is used for storing programs, instructions or codes, and the processor is used for executing the programs, instructions or codes in the machine readable storage medium to realize the method described above.

[0040] Based on the above aspects, the embodiment of the present application extracts the first deformation feature vectors by obtaining and analyzing a plurality of support structure data of tunnel soft rock, and intelligently selects the applicable deformation feature conversion strategy by combining the monitoring intention and the data label, to convert the first deformation feature vectors into the second deformation feature vectors which are more representative. This method not only realizes the accurate capture and conversion of the deformation features of the support structure, but also generates the reliability prediction results for the deformation support monitoring request by making the reliability index decision on the converted deformation features, effectively improving the accuracy and efficiency of the tunnel soft rock large deformation support reliability monitoring, thereby helping the safe construction and operation management of the tunnel engineering. BRIEF DESCRIPTION OF DRAWINGS

[0041] Figure 1 is the execution flow diagram of the tunnel soft rock large deformation support reliability monitoring method provided by the embodiment of the present application.

[0042] Figure 2 is the hardware architecture diagram of the tunnel soft rock large deformation support reliability monitoring system provided by the embodiment of the present application. DETAILED DESCRIPTION

[0043] The present application will be described in detail below with reference to the accompanying drawings of the specification.Figure 1 is a flowchart of a tunnel soft rock large deformation support reliability monitoring method provided by an embodiment of the present application. The tunnel soft rock large deformation support reliability monitoring method will be described in detail below.

[0044] In step S110, a plurality of support structure data of the tunnel soft rock corresponding to the deformation support monitoring request is obtained, and a first deformation feature vector corresponding to each of the plurality of support structure data is extracted. The plurality of support structure data corresponds to a plurality of structural performance morphologies.

[0045] In this embodiment, in a tunnel project, in order to ensure the stability and safety of the soft rock part of the tunnel, it is often necessary to monitor the support structure. For example, in a mountainous tunnel project under construction, due to the complex soft rock geological conditions, the soft rock characteristics of different sections differ greatly, so a plurality of support structures are arranged at different positions of the tunnel. These support structures include anchor rods, sprayed concrete layers, steel arches, etc.

[0046] When the deformation support monitoring request is received, a plurality of support structure data of the tunnel soft rock corresponding to the deformation support monitoring request can be obtained. For the anchor rod support structure, its length, diameter, anchoring depth, spacing, etc. can be obtained. Taking the length data as an example, assuming that the design length of the anchor rod in a certain tunnel section is 3 meters, but the actual measured length may change due to the deformation of the soft rock, and this actual measured length is part of the support structure data. For the sprayed concrete layer, its thickness, strength grade, spraying angle, etc. can be obtained. The thickness data can be obtained by a special measuring instrument, such as an ultrasonic thickness gauge, and the measurement result is the support structure data of the sprayed concrete layer. The data of the steel arch includes the steel type of the arch, the spacing of the arch, the curvature of the arch, etc.

[0047] After obtaining these support structure data, the first deformation feature vector corresponding to each support structure data can be further extracted. Taking the length data of the anchor rod as an example, the first deformation feature vector corresponding thereto can contain the length change rate, the difference in length change with adjacent anchor rods, and other information. Assuming that the initial length of the anchor rod is 3 meters in a period of time, and the length is measured to be 3.05 meters after a period of time, the length change rate is (3.05-3) / 3=0.0167, and this 0.0167 is an element of the length change rate feature in the first deformation feature vector. For the thickness data of the sprayed concrete layer, if the initial thickness design is 15 centimeters, and the measured thickness is 14.5 centimeters, the thickness change amount -0.5 centimeters and the change rate -0.5 / 15, etc. constitute part of the elements of the first deformation feature vector corresponding to the thickness data of the sprayed concrete layer. The first deformation feature vectors corresponding to these different support structure data have different characteristics and meanings due to the various structural manifestations of the support structure, and they are the basis for subsequent analysis and decision-making.

[0048] In step S120, at least two reference deformation feature conversion strategies are obtained, and each reference deformation feature conversion strategy includes at least one of the steps of feature selection and feature focusing on the first deformation feature vector.

[0049] In the monitoring and analysis process of the tunnel soft rock support structure, in order to better process and analyze the first deformation feature vector, a plurality of reference deformation feature conversion strategies need to be obtained. For example, one reference deformation feature conversion strategy can focus on feature selection. Assuming that under this strategy, for the element of the anchor rod length change rate in the first deformation feature vector, if its fluctuation range in the past multiple monitoring data is very small, and the correlation with the overall deformation of the tunnel is low, this strategy can select to exclude it from subsequent analysis, which is the process of feature selection.

[0050] Another reference deformation feature conversion strategy can focus more on feature focusing. For example, for the first deformation feature vector of the sprayed concrete layer thickness, since the sprayed concrete layer plays a key role in resisting lateral pressure of soft rock in tunnel support, this reference deformation feature conversion strategy can focus on the thickness change rate feature. For example, by using a specific algorithm such as a weighting algorithm, the weight of the thickness change rate feature in subsequent analysis can be increased, so as to highlight the importance of this feature in judging the deformation of the support structure.

[0051] Another reference deformation feature conversion strategy can contain both feature selection and feature focusing. For example, for the first deformation feature vector element of the spacing of the steel arch, first, through the analysis of historical data, it is found that when the spacing of the arch changes within a certain range, the influence on the deformation of the tunnel soft rock is not significant, so feature selection is performed first, and the data corresponding to this change range is screened out. Then, for the remaining data related to the spacing of the arch, the feature focusing method is used, according to the correlation between the spacing of the arch and the settlement of the soft rock at the top of the tunnel, different weights are given to different degrees, so that the influence of the change of the spacing of the arch on the overall deformation of the tunnel soft rock supporting structure can be more accurately reflected in the subsequent analysis. These different reference deformation feature conversion strategies are constructed based on the deep understanding of the characteristics and deformation mechanism of the tunnel soft rock supporting structure, and provide multiple choices for subsequent determination of the most suitable target deformation feature conversion strategy.

[0052] In step S130, a target deformation feature conversion strategy is determined from the at least two reference deformation feature conversion strategies according to a monitoring intention corresponding to the deformation support monitoring request and a data tag corresponding to the support structure data, the data tag representing a reference tag knowledge point corresponding to the support structure data under the deformation support monitoring request.

[0053] In the actual scene of tunnel soft rock supporting structure monitoring, the monitoring intention of the deformation support monitoring request can be to focus on the settlement of the soft rock at the top of the tunnel to prevent collapse accidents. The data tag corresponding to the support structure data contains the reference tag knowledge points of each support structure data under this monitoring intention. For example, for the anchor support structure data, the data tag can indicate that the anchor mainly plays a role in controlling the settlement of the soft rock at the top of the tunnel through anchoring force, and the anchoring force is related to the length, diameter, and anchoring depth of the anchor.

[0054] Suppose there are two reference deformation feature conversion strategies, strategy A and strategy B. Strategy A focuses on the anchoring depth and diameter of the anchor as the first deformation feature vector elements in feature selection, and gives a higher weight to the thickness change rate of the sprayed concrete layer in feature focusing. Strategy B selects the length and spacing of the anchor and the thickness of the sprayed concrete layer as the first deformation feature vector elements in feature selection, and gives a higher weight to the change of the arc of the steel arch in feature focusing.

[0055] Since the monitoring intention is to focus on the settlement of the soft rock at the top of the tunnel, according to the data labels of the supporting structure data, it can be known that the anchoring depth of the anchor rod is more critical to control the settlement of the soft rock at the top, and the thickness change of the sprayed concrete layer also has an important influence on the settlement of the soft rock at the top. In contrast, strategy B, although it also involves some related elements, is not as good as strategy A in terms of matching the monitoring intention. Therefore, according to the monitoring intention and the data labels, strategy A is finally determined as the target deformation feature conversion strategy. This process is achieved by considering the focus of the monitoring intention and the importance of each element in the supporting structure data under the monitoring intention, ensuring that the selected target deformation feature conversion strategy can most effectively process the first deformation feature vector to meet the needs of deformation monitoring.

[0056] In step S140, the target deformation feature conversion strategy is used to perform deformation feature conversion on the plurality of first deformation feature vectors to generate a plurality of second deformation feature vectors.

[0057] After the target deformation feature conversion strategy is determined, the plurality of first deformation feature vectors can be subjected to deformation feature conversion. For example, the target deformation feature conversion strategy is the strategy A mentioned earlier. For the first deformation feature vector of the anchor rod supporting structure, since strategy A focuses on the anchoring depth and diameter of the anchor rod in feature selection, in the deformation feature conversion process, other first deformation feature vector elements of the anchor rod will be screened first, and only the data related to the anchoring depth and diameter will be retained. Assuming that the original first deformation feature vector of the anchoring depth of the anchor rod contains depth change data at multiple time points, such as [0.02 meters (1st week), 0.03 meters (2nd week), 0.025 meters (3rd week)], and the original first deformation feature vector of the diameter contains [0.002 meters (1st week), 0.0015 meters (2nd week), 0.0018 meters (3rd week)].

[0058] Then, according to the feature focusing part in strategy A, the thickness change rate of the sprayed concrete layer is given a higher weight. Assuming that the original thickness change rate first deformation feature vector of the sprayed concrete layer is [0.05 (1st week), 0.06 (2nd week), 0.04 (3rd week)], in the feature focusing process, through a specific weight calculation method, such as multiplying a coefficient greater than 1, it is converted to [0.1 (1st week), 0.12 (2nd week), 0.08 (3rd week)].

[0059] After such feature selection and feature focusing operations, the processed data of the anchoring depth and diameter of the anchor rod and the thickness variation rate of the shotcrete layer form a new second deformation feature vector. For the first deformation feature vectors of other support structures, similar conversion operations are performed according to the rules in the target deformation feature conversion strategy, thereby obtaining a plurality of second deformation feature vectors. These second deformation feature vectors are optimized and adjusted feature vectors, which can better reflect the key information related to the deformation of the tunnel soft rock, and provide more effective data support for the subsequent reliability index decision.

[0060] In step S150, reliability index decision is performed on the plurality of second deformation feature vectors to generate reliability index decision data corresponding to the deformation support monitoring request, which represents the reliability prediction result corresponding to the deformation support monitoring request generated by using the plurality of support structure data.

[0061] After obtaining the plurality of second deformation feature vectors, reliability index decision is performed. For example, in the tunnel soft rock support structure, the second deformation feature vector of the anchor rod support part contains the processed anchoring depth and diameter data. If the variation of the anchoring depth is always within a reasonable range, and the variation of the diameter does not exceed the design allowable fluctuation range, it indicates that the reliability of the anchor rod support at this stage is high. Assuming that according to historical experience and engineering standards, the reasonable variation range of the anchoring depth is ±0.05 meters, and the reasonable fluctuation range of the diameter is ±0.002 meters. When the anchoring depth variation value in the second deformation feature vector is within this range, and the diameter variation value is also within the corresponding range, a high reliability score, such as 80 points (out of 100 points), can be given to the anchor rod support part.

[0062] For the shotcrete layer, if the data of the thickness variation rate in the second deformation feature vector after feature focusing indicates that the thickness variation rate is within an acceptable range and there is no sharp change trend, it also indicates that the support reliability of the shotcrete layer is high. Assuming that the acceptable range of the thickness variation rate is ±0.1, when the actual thickness variation rate is within this range, a reliability score of 70 points is given to the shotcrete layer support part.

[0063] By performing such reliability analysis on the second deformation characteristic vectors corresponding to each supporting structure, and comprehensively considering the importance of all supporting structures in the tunnel soft rock support, such as the importance weight of the anchor rod being 0.4, the importance weight of the sprayed concrete layer being 0.3, and the other supporting structures such as the steel arch being analyzed for reliability according to the respective second deformation characteristic vectors and the corresponding standards and determining the weight, the reliability index decision data of the entire tunnel soft rock supporting structure under the deformation support monitoring request can be finally calculated. For example, the calculated reliability index decision data is 75 points, which represents the reliability prediction result corresponding to the deformation support monitoring request generated by using the data of multiple supporting structures. This result can provide intuitive information for tunnel engineering construction and maintenance personnel, so that they can take timely measures to ensure that the supporting structure of the tunnel soft rock part is always in a reliable working state and ensure the safety of the tunnel.

[0064] Based on the above steps, the embodiments of the present application extract the first deformation characteristic vector by acquiring and analyzing the multiple supporting structure data of the tunnel soft rock, and intelligently select the applicable deformation characteristic conversion strategy by combining the monitoring intention and the data label, and convert the first deformation characteristic vector into a more representative second deformation characteristic vector. This method not only realizes the accurate capture and conversion of the deformation characteristics of the supporting structure, but also generates a reliability prediction result for the deformation support monitoring request by performing reliability index decision on the converted deformation characteristics, effectively improving the accuracy and efficiency of the reliability monitoring of the large deformation support of the tunnel soft rock, thereby helping the safe construction and operation and maintenance management of the tunnel project.

[0065] In a possible implementation, step S130 can include:

[0066] The monitoring intention corresponding to the deformation support monitoring request, the multiple first deformation characteristic vectors, and the at least two reference deformation characteristic conversion strategies are loaded into an artificial intelligence network that has completed knowledge learning in advance, to obtain the target deformation characteristic conversion strategy. The artificial intelligence network is used to perform strategy matching on the at least two reference deformation characteristic conversion strategies according to the monitoring intention corresponding to the deformation support monitoring request and the multiple first deformation characteristic vectors, and the target deformation characteristic conversion strategy includes at least one of a target feature selection strategy and a target feature focusing strategy.

[0067] In a possible implementation, the artificial intelligence network includes a target matching module and a decision execution module.

[0068] The monitoring intention corresponding to the deformation support monitoring request, the multiple first deformation characteristic vectors, and the at least two reference deformation characteristic conversion strategies are loaded into an artificial intelligence network that has completed knowledge learning in advance, to obtain the target deformation characteristic conversion strategy. The artificial intelligence network is used to perform strategy matching on the at least two reference deformation characteristic conversion strategies according to the monitoring intention corresponding to the deformation support monitoring request and the multiple first deformation characteristic vectors, and the target deformation characteristic conversion strategy includes at least one of a target feature selection strategy and a target feature focusing strategy.

[0069] Step S131, load the monitoring intention corresponding to the deformation support monitoring request, the plurality of first deformation feature vectors and the at least two reference deformation feature conversion strategies to the target matching module, obtain the matching confidence degrees corresponding to the at least two reference deformation feature conversion strategies respectively, and the target matching module is used for performing the strategy matching of the at least two reference deformation feature conversion strategies according to the monitoring intention corresponding to the deformation support monitoring request and the plurality of first deformation feature vectors.

[0070] Step S132, determine the target feature selection strategy according to the matching confidence degrees corresponding to the at least two reference deformation feature conversion strategies respectively.

[0071] Step S133, load the monitoring intention corresponding to the deformation support monitoring request, the plurality of first deformation feature vectors and the matching confidence degrees corresponding to the at least two reference deformation feature conversion strategies respectively to the decision execution module, obtain the target feature focusing strategy, and the decision execution module is used for obtaining the target feature focusing strategy matching the monitoring intention based on the plurality of first deformation feature vectors and the matching confidence degrees corresponding to the at least two reference deformation feature conversion strategies respectively.

[0072] In this embodiment, a specific tunnel project is taken as an example, the monitoring intention of the deformation support monitoring request is to accurately evaluate the deformation of the supporting structure of the tunnel soft rock under the action of lateral pressure, so as to ensure the safety inside the tunnel and the stability of the structure. At this time, the plurality of first deformation feature vectors obtained contain the related deformation information of different supporting structures, such as the axial strain vector of the anchor rod, the surface crack propagation vector of the sprayed concrete layer and the bending degree change vector of the steel arch, etc. At the same time, there are a plurality of reference deformation feature conversion strategies to choose from, each reference deformation feature conversion strategy contains different ways of processing the first deformation feature vector, including feature selection and feature focusing operations.

[0073] Firstly, the monitoring intention corresponding to the deformation support monitoring request, the plurality of first deformation feature vectors and the at least two reference deformation feature conversion strategies are loaded to the target matching module. The target matching module performs strategy matching of the at least two reference deformation feature conversion strategies according to the monitoring intention corresponding to the deformation support monitoring request and the plurality of first deformation feature vectors, so as to obtain the matching confidence degrees corresponding to the at least two reference deformation feature conversion strategies respectively.

[0074] Assume that there are two reference deformation feature conversion strategies, strategy one and strategy two, in this tunnel project. For strategy one, in terms of feature selection, it tends to select the axial strain of the anchor rod and the surface crack propagation of the sprayed concrete layer as the two first deformation feature vector elements, and in terms of feature focusing, it focuses on the trend of the axial strain change; for strategy two, in terms of feature selection, it pays more attention to the bending degree change of the steel arch and the surface crack propagation of the sprayed concrete layer, and in terms of feature focusing, it focuses on the speed of surface crack propagation.

[0075] When the target matching module performs strategy matching, it will comprehensively consider the monitoring intention and the actual situation of each first deformation feature vector. Since the monitoring intention is to evaluate the deformation of the tunnel soft rock under the action of lateral pressure, from multiple first deformation feature vectors, the axial strain of the anchor rod is of great significance to reflect the deformation of the supporting structure under lateral pressure, and the surface crack propagation of the sprayed concrete layer is also a key representation under the action of lateral pressure. The target matching module evaluates strategy one and strategy two respectively through internal algorithms and learned knowledge, and obtains their respective matching confidence. Assume that the matching confidence of strategy one is 0.7 and the matching confidence of strategy two is 0.5.

[0076] According to the matching confidence of each of the at least two reference deformation feature conversion strategies, the target feature selection strategy is determined. In the above example, since the matching confidence of strategy one is 0.7, which is higher than that of strategy two, 0.5, the feature selection part of strategy one is determined as the target feature selection strategy, that is, the axial strain of the anchor rod and the surface crack propagation of the sprayed concrete layer are selected as the two first deformation feature vector elements as the content of the target feature selection strategy.

[0077] Then, the monitoring intention corresponding to the deformation support monitoring request, the multiple first deformation feature vectors, and the matching confidence of each of the at least two reference deformation feature conversion strategies are loaded to the decision execution module. The decision execution module obtains the target feature focusing strategy that matches the monitoring intention based on the multiple first deformation feature vectors and the matching confidence of each of the at least two reference deformation feature conversion strategies.

[0078] Continuing with this example, after receiving the relevant information, the decision execution module, due to having determined the first deformation feature vector elements of interest in the target feature selection strategy, will further determine the target feature focus strategy based on these elements and the matching confidence. In this process, the decision execution module will deeply analyze the relationship between the monitoring intention and these elements. For example, because the monitoring intention is to evaluate the deformation under lateral pressure, and the axial strain trend of the anchor rod plays an important role in reflecting this deformation, and the surface crack propagation speed of the shotcrete layer is also of key significance to the judgment of deformation under lateral pressure. Based on the matching confidence of strategy one and the importance of these elements, the decision execution module determines that in the target feature focus strategy, the axial strain trend of the anchor rod and the surface crack propagation speed of the shotcrete layer are given specific weights and focus methods to better match the monitoring intention.

[0079] Thus, by using the target matching module and the decision execution module in the artificial intelligence network, according to the monitoring intention corresponding to the deformation support monitoring request and the data label corresponding to the support structure data, the target deformation feature conversion strategy containing at least one of the target feature selection strategy and the target feature focus strategy is determined from at least two reference deformation feature conversion strategies. This target deformation feature conversion strategy can more accurately process the first deformation feature vector, thereby providing an effective basis for subsequent tunnel soft rock support structure deformation analysis and decision-making.

[0080] In this process, the knowledge learning of the artificial intelligence network is crucial, which is based on a large amount of tunnel soft rock support structure data, different monitoring intentions, and corresponding deformation feature conversion strategies. For example, in the learning process of the network, historical data of numerous tunnel projects may be used, which contains various first deformation feature vectors and corresponding effective deformation feature conversion strategies under different geological conditions and different support structure designs. Through learning of these massive data, the artificial intelligence network can accurately match the reference deformation feature conversion strategy under different monitoring intentions and first deformation feature vectors, derive a reasonable matching confidence, and determine the most suitable target deformation feature conversion strategy.

[0081] Meanwhile, the internal algorithms and logics of the target matching module and the decision execution module are also carefully designed. When calculating the matching confidence, the target matching module considers multiple factors, such as the relevance of each element in the first deformation feature vector to the monitoring intent, the adaptability of the feature selection and feature focusing operation in different reference deformation feature conversion strategies to the monitoring intent, and the like. When determining the target feature focusing strategy, the decision execution module not only considers the matching confidence, but also combines the actual physical meaning of the first deformation feature vector and the core requirements of the monitoring intent, to ensure that the determined target feature focusing strategy can meet the accurate evaluation requirements of the deformation of the tunnel soft rock supporting structure to the greatest extent.

[0082] In a possible implementation, the matching confidence corresponding to each of the at least two reference deformation feature conversion strategies includes first matching confidence corresponding to each of the multiple feature selection strategies.

[0083] Step S132 can include: from the first matching confidence corresponding to each of the multiple feature selection strategies, outputting the feature selection strategy with the largest value of the first matching confidence as the target feature selection strategy.

[0084] In a possible implementation, the matching confidence corresponding to each of the at least two reference deformation feature conversion strategies includes second matching confidence corresponding to each of the multiple feature focusing strategies.

[0085] Step S133 can include: loading the monitoring intent corresponding to the deformation support monitoring request, the multiple first deformation feature vectors, and the second matching confidence corresponding to each of the multiple feature focusing strategies into the decision execution module, to obtain the target feature focusing strategy.

[0086] In this embodiment, a certain large tunnel project is taken as an example. The tunnel passes through a complex soft rock stratum, and in order to ensure the safety and stability of the tunnel, the deformation monitoring of the supporting structure is crucial. In this process, there are multiple reference deformation feature conversion strategies, each strategy contains different feature selection strategies and feature focusing strategies, and the matching confidence corresponding to each reference deformation feature conversion strategy is calculated, which includes first matching confidence corresponding to each of the multiple feature selection strategies and second matching confidence corresponding to each of the multiple feature focusing strategies.

[0087] For the first matching confidence corresponding to each of the multiple feature selection strategies in the matching confidence corresponding to each of the at least two reference deformation feature conversion strategies, when determining the target feature selection strategy according to these matching confidences, the feature selection strategy with the largest value of the first matching confidence is outputted as the target feature selection strategy from the first matching confidence corresponding to each of the multiple feature selection strategies.

[0088] Suppose there are three reference deformation feature conversion strategies in this tunnel project, which are strategy A, strategy B and strategy C. The feature selection strategies in each strategy have different focuses. The feature selection strategy of strategy A focuses on the anchoring depth of the anchor rod, the thickness of the sprayed concrete layer and the spacing of the steel arch, which are the three first deformation feature vector elements. The feature selection strategy of strategy B focuses on the axial strain of the anchor rod, the compressive strength of the sprayed concrete layer and the steel material type of the steel arch, which are the three first deformation feature vector elements. The feature selection strategy of strategy C mainly focuses on the anchoring angle of the anchor rod, the elastic modulus of the sprayed concrete layer and the arch curvature of the steel arch, which are the three first deformation feature vector elements.

[0089] When calculating the first matching confidence corresponding to these feature selection strategies, the target matching module will comprehensively evaluate the monitoring intention corresponding to the deformation support monitoring request and the plurality of first deformation feature vectors. Suppose the monitoring intention is to accurately judge the stability of the tunnel soft rock under the action of vertical pressure. For strategy A, since the anchoring depth of the anchor rod is directly related to the resistance to vertical pressure, the thickness of the sprayed concrete layer also reflects the bearing capacity of the vertical pressure to some extent, and the spacing of the steel arch is related to the overall stability of the support structure. After complex calculation and evaluation, the first matching confidence corresponding to the feature selection strategy of strategy A is 0.7. For strategy B, although the axial strain of the anchor rod and the compressive strength of the sprayed concrete layer are related to the vertical pressure to some extent, but compared with strategy A, its direct correlation with the monitoring intention is slightly lower, and its first matching confidence is 0.5. For strategy C, the anchoring angle of the anchor rod, the elastic modulus of the sprayed concrete layer and the arch curvature of the steel arch have relatively weak relevance in reflecting the stability of the support structure under the action of vertical pressure, and its first matching confidence is 0.4.

[0090] According to the rules, since the first matching confidence of strategy A is 0.7, which is the largest value among the three strategies, the feature selection strategy in strategy A is determined as the target feature selection strategy, that is, the anchoring depth of the anchor rod, the thickness of the sprayed concrete layer and the spacing of the steel arch are selected as the three first deformation feature vector elements as the focus of the subsequent operation.

[0091] In the case where the matching confidence also includes the second matching confidence corresponding to a plurality of feature focusing strategies, the process of loading the monitoring intention corresponding to the deformation support monitoring request, the plurality of first deformation feature vectors and the matching confidence corresponding to the at least two reference deformation feature conversion strategies respectively into the decision execution module to obtain the target feature focusing strategy is as follows.

[0092] Continuing with the above tunnel engineering example, assume that the feature focusing strategy in each reference deformation feature conversion strategy is also different. The feature focusing strategy of strategy A gives higher weight to the thickness change rate of the sprayed concrete layer, aiming to highlight the importance of thickness change to the stability of the supporting structure under vertical pressure; the feature focusing strategy of strategy B focuses on the strength change coefficient corresponding to the steel type of the steel arch, because the strength change of the steel type has a greater impact on the overall performance of the supporting structure; the feature focusing strategy of strategy C focuses on the change trend of the anchoring angle of the anchor rod, considering that the change trend of the anchoring angle has special significance to the stability of the supporting structure under vertical pressure.

[0093] When calculating the second matching confidence corresponding to each feature focusing strategy, it is also based on the monitoring intention corresponding to the deformation supporting monitoring request, the plurality of first deformation feature vectors and other factors. For strategy A, since the thickness change rate of the sprayed concrete layer has higher relevance in reflecting the stability of the supporting structure under vertical pressure, after detailed analysis and calculation, the second matching confidence corresponding to the feature focusing strategy is 0.6. For strategy B, although the strength change coefficient corresponding to the steel type of the steel arch has an impact on the performance of the supporting structure, its importance in reflecting the stability under vertical pressure is slightly lower than that of the thickness change rate of the sprayed concrete layer, and its second matching confidence is 0.5. For strategy C, the change trend of the anchoring angle of the anchor rod has relatively lower importance in reflecting the stability of the supporting structure under vertical pressure, and its second matching confidence is 0.4.

[0094] Then, the monitoring intention corresponding to the deformation supporting monitoring request (i.e. accurately judging the stability of the supporting structure of the tunnel soft rock under the action of vertical pressure), the plurality of first deformation feature vectors (including the anchoring depth of the anchor rod, the thickness of the sprayed concrete layer, the spacing of the steel arch, and other related vector elements), and the second matching confidence corresponding to the plurality of feature focusing strategies (strategy A is 0.6, strategy B is 0.5, and strategy C is 0.4) are loaded into the decision execution module. Based on these input information, the decision execution module deeply analyzes the relationship between the monitoring intention and each feature focusing strategy. Since the second matching confidence of strategy A is the highest at 0.6, and the thickness change rate of the sprayed concrete layer has important significance in monitoring the stability of the supporting structure under vertical pressure, the decision execution module determines that the target feature focusing strategy is to give higher weight to the thickness change rate of the sprayed concrete layer to better match the monitoring intention.

[0095] Throughout the whole process, the work of the target matching module and the decision execution module is highly dependent on a large amount of engineering data and accurate algorithms. The target matching module needs to consider a large number of factors when calculating the first matching confidence and the second matching confidence. For the calculation of the first matching confidence, not only the direct correlation between the first deformation characteristic vector elements selected in the feature selection strategy and the monitoring intention needs to be considered, but also the interaction relationship between these elements in the whole supporting structure system needs to be considered. For example, although the anchoring depth of the anchor rod has a direct relationship with the resistance to vertical pressure, it has a synergistic relationship with the thickness of the sprayed concrete layer and the spacing of the steel arch, and this synergistic relationship also needs to be accurately considered when calculating the first matching confidence.

[0096] For the calculation of the second matching confidence, the decision execution module needs to consider the sensitivity and importance of the elements concerned in the feature focusing strategy in reflecting the monitoring intention. For example, the thickness change rate of the sprayed concrete layer is different in different engineering environments and supporting structure states, and the sensitivity of the thickness change rate of the sprayed concrete layer to the stability of the supporting structure under vertical pressure is different, which needs to be accurately analyzed and evaluated according to a large amount of historical data and engineering experience, so as to obtain an accurate second matching confidence.

[0097] In addition to the high and low of the second matching confidence, the decision execution module also needs to consider the overall situation of the first deformation characteristic vector and the core requirements of the monitoring intention when determining the target feature focusing strategy. For example, although the second matching confidence of strategy A is the highest, the decision execution module also needs to consider whether the weight given to the thickness change rate of the sprayed concrete layer can optimally reflect the stability of the supporting structure under vertical pressure on the basis of the target feature selection strategy having determined that the anchoring depth of the anchor rod, the thickness of the sprayed concrete layer, and the spacing of the steel arch are the elements to be concerned. This requires a deep understanding and accurate grasp of the mechanical properties, deformation mechanism, and change law of the monitoring data of the whole supporting structure.

[0098] Therefore, according to the specific monitoring intention, the most suitable strategy content can be accurately selected from a large number of reference deformation feature conversion strategies, thereby improving the accuracy and effectiveness of the deformation monitoring of the supporting structure, and providing reliable technical support for ensuring the safety and stability of the tunnel engineering.

[0099] In one possible implementation, before step S131, the method further includes:

[0100] Step A110, obtaining a plurality of template supporting structure data corresponding to the template deformation supporting monitoring request, and extracting a first deformation characteristic vector corresponding to each of the plurality of template supporting structure data, wherein the plurality of template supporting structure data correspond to a plurality of structural appearance forms.

[0101] Step A120: Obtain at least two template deformation feature transformation strategies, each of which includes at least one of a template feature selection strategy and a template feature focusing strategy.

[0102] Step A130: Load the monitoring intent corresponding to the template deformation support monitoring request, multiple first deformation feature vectors, and the at least two template deformation feature conversion strategies into the template matching module to obtain the matching confidence levels corresponding to the at least two template deformation feature conversion strategies respectively.

[0103] Step A140: Load the monitoring intent corresponding to the template deformation support monitoring request, multiple first deformation feature vectors, and the matching confidence corresponding to the at least two template deformation feature transformation strategies into the decision execution module to generate reliability index decision data of the feature focusing strategy.

[0104] Step A150: Generate a target incentive mechanism network based on the reliability index decision data of the feature focusing strategy.

[0105] Step A160: Optimize the template matching module based on the target incentive mechanism network to generate the target matching module.

[0106] In this embodiment, taking a typical tunnel project as an example, the template deformation support monitoring request is set to assess the stability of the support structure in soft rock of a specific area of ​​the tunnel under various complex working conditions. For this request, the corresponding template support structure data covers various types. For example, the relevant data for anchor bolts includes the depth values ​​of their different anchoring sections and the axial stress values ​​at different construction stages; the data for shotcrete layers includes thickness measurements at different locations and compressive strength values ​​at different curing times; the data for steel arch frames includes the arch frame radius at different locations, the elastic modulus of the steel, and the nodal stress values ​​at the arch frame connections. These data reflect the various structural performances of different support structures in the soft rock environment of the tunnel. From these template support structure data, the corresponding first deformation feature vectors are extracted according to a predetermined feature extraction algorithm. For the anchoring depth data of anchor bolts, the first deformation feature vector may include the rate of change of depth and the changing trend of the depth difference with adjacent anchor bolts; the first deformation feature vector corresponding to the shotcrete layer thickness data may include the rate of thickness reduction and the amplitude of thickness fluctuation within a specific time period.

[0107] Next, the template feature selection strategy of the first template deformation feature conversion strategy focuses on the two primary deformation feature vector elements: the axial stress of the anchor bolts and the compressive strength of the shotcrete layer. In terms of template feature focusing strategy, it assigns a high weight to the rate of change of the compressive strength of the shotcrete layer, because in soft rock tunnels, the change in the compressive strength of the shotcrete layer has a significant impact on the overall stability of the support structure. The other template deformation feature conversion strategy, the second one, focuses on the nodal stress of the steel arch frame and the anchorage depth of the anchor bolts in its template feature selection strategy. Its template feature focusing strategy focuses on the changing trend of the nodal stress of the steel arch frame, because the nodal stress of the steel arch frame can directly reflect the local stability of the support structure.

[0108] Then, the template matching module performs calculations based on preset algorithms and models. For template deformation feature conversion strategy one, since the monitoring intent is to assess the stability of the support structure under various complex working conditions, the axial stress of the anchor bolts and the compressive strength of the shotcrete layer selected in the template feature selection strategy are strongly correlated with this intent. When calculating the matching confidence level, the module considers factors such as the influence weight of these elements on the stability of the support structure under different working conditions and the correlation in historical data. Assuming that after detailed calculation, the matching confidence level of template deformation feature conversion strategy one is 0.7. For template deformation feature conversion strategy two, although the stress of the steel arch frame nodes and the anchor bolt anchorage depth are also important factors, their overall matching degree with the monitoring intent is slightly lower, and after evaluation, its matching confidence level is 0.5.

[0109] Subsequently, the decision execution module conducts in-depth analysis based on the input information. For template deformation feature conversion strategy one, due to its high matching confidence and the high weight given to the rate of change of compressive strength of the shotcrete layer in the template feature focusing strategy, the decision execution module evaluates the reliability of this feature focusing strategy in accurately reflecting the stability of the support structure under different working conditions based on a large amount of engineering data and empirical models. For example, by analyzing the relationship between the rate of change of compressive strength of the shotcrete layer and the actual stability of the support structure under different geological conditions and construction progress, a specific reliability index value is obtained. Assuming that the calculated reliability index decision data for the feature focusing strategy corresponding to template deformation feature conversion strategy one is 0.8. For template deformation feature conversion strategy two, the same principle is used for analysis and calculation, and the reliability index decision data for its feature focusing strategy is obtained as 0.6.

[0110] Further, the construction of the target incentive mechanism network is based on the reliability evaluation of the feature focusing strategy of different template deformation feature conversion strategies. Taking the previous results as an example, the reliability index decision data of template deformation feature conversion strategy one is 0.8, and the data of strategy two is 0.6. The target incentive mechanism network will set different incentive weights according to these data. For the template deformation feature conversion strategy one with higher reliability, a greater incentive weight is given in the target incentive mechanism network, which means that the feature selection and feature focusing operation related to strategy one will be more valued in the subsequent optimization process.

[0111] Finally, in the optimization process, the incentive weight of the target incentive mechanism network will affect the algorithm parameters and matching logic inside the template matching module. For the matching factors related to template deformation feature conversion strategy one, since it has a greater incentive weight in the target incentive mechanism network, the template matching module will adjust the internal algorithm, so that in the subsequent actual deformation support monitoring request, it is more inclined to identify a reference deformation feature conversion strategy similar to template deformation feature conversion strategy one, and give a higher matching confidence. For example, if there is a reference deformation feature conversion strategy similar to template deformation feature conversion strategy one in terms of feature selection and feature focusing in the actual deformation support monitoring request, the optimized template matching module will give it a relatively higher matching confidence, thereby improving the accuracy and rationality of the entire monitoring system in selecting the target deformation feature conversion strategy.

[0112] Through the processing of template related data and strategies, the optimized target matching module is finally generated, which lays a solid foundation for accurately obtaining the matching confidence of the reference deformation feature conversion strategy, thereby ensuring that in the tunnel soft rock support structure monitoring, the appropriate deformation feature conversion strategy can be selected more accurately according to the monitoring intention, and the effectiveness and reliability of the deformation evaluation and monitoring of the support structure are improved.

[0113] In one possible implementation, step S140 includes:

[0114] Step S141, using the target feature selection strategy to perform correlation selection on the plurality of first deformation feature vectors, to generate a plurality of first sub-deformation feature vectors satisfying the correlation requirement, wherein the correlation requirement represents the correlation between the first deformation feature vector and the monitoring intention.

[0115] Step S142, using the target feature focusing strategy to perform feature focusing on the plurality of first sub-deformation feature vectors, to generate the plurality of second deformation feature vectors.

[0116] In this embodiment, the tunnel engineering mentioned earlier is taken as an example, where the monitoring intention of the deformation support monitoring request is to accurately assess the deformation of the supporting structure of the tunnel soft rock under lateral pressure, to ensure the safety inside the tunnel and the stability of the structure. The target deformation feature conversion strategy is determined through the previous steps, which includes the target feature selection strategy and the target feature focusing strategy.

[0117] Firstly, the target feature selection strategy is used to select the correlation of multiple first deformation feature vectors, generating multiple first sub-deformation feature vectors that meet the correlation requirements. The correlation requirements here represent the correlation between the first deformation feature vector and the monitoring intention. For the supporting structure in this tunnel engineering, the first deformation feature vector contains multiple aspects of information, such as the axial strain vector of the anchor rod, the surface crack propagation vector of the sprayed concrete layer, and the bending degree change vector of the steel arch, etc.

[0118] The target feature selection strategy focuses on the first deformation feature vector elements closely related to the monitoring intention. In the monitoring intention of judging the influence of lateral pressure on the deformation of the supporting structure, the axial strain of the anchor rod is a key factor when resisting lateral pressure, and the change of the axial strain can directly reflect the size of the lateral force borne by the anchor rod and the deformation of the anchor rod itself, so the axial strain vector has a high correlation with the monitoring intention. The expansion direction and speed of the surface cracks of the sprayed concrete layer are greatly affected by the lateral pressure, and the expansion vector of the cracks can reflect the stress state and structural integrity of the sprayed concrete layer under lateral pressure, so the surface crack propagation vector is also closely related to the monitoring intention. For the steel arch, the bending degree change directly reflects the structural deformation under lateral pressure, and the bending degree change vector is also closely related to the monitoring intention.

[0119] Through the target feature selection strategy, the original multiple first deformation feature vectors are screened, and the elements with low correlation with the monitoring intention are removed, and the axial strain vector of the anchor rod, the surface crack propagation vector of the sprayed concrete layer, and the bending degree change vector of the steel arch, etc. elements that meet the correlation requirements are retained, thereby generating multiple first sub-deformation feature vectors. This process is based on the in-depth understanding of the mechanical behavior of the tunnel soft rock supporting structure under lateral pressure, and the accurate grasp of the internal relationship between each first deformation feature vector and the monitoring intention.

[0120] Then, the target feature focusing strategy is used to focus on the multiple first sub-deformation feature vectors, generating multiple second deformation feature vectors. The target feature focusing strategy aims to further highlight the importance of certain key information in the first sub-deformation feature vector, to better reflect the monitoring intention.

[0121] For the axial strain vector of the anchor rod, it is assumed that in the target feature focusing strategy, according to engineering experience and data analysis, the change trend of the axial strain in a certain period of time is more critical to the judgment of the stability of the supporting structure. For example, in the process of the lateral pressure of the tunnel soft rock gradually increasing, if the axial strain of the anchor rod changes sharply in a short time, it may indicate that the supporting structure is about to face the risk of instability. Therefore, the target feature focusing strategy focuses on this change trend in the axial strain vector, and enhances the influence of this change trend in the entire axial strain vector through a specific algorithm, such as a weighting algorithm or data transformation.

[0122] For the surface crack propagation vector of the sprayed concrete layer, the target feature focusing strategy may focus on the acceleration of crack propagation. Because under the action of lateral pressure, the acceleration of crack propagation can more sensitively reflect the structural change rate of the sprayed concrete layer. If the acceleration of crack propagation suddenly increases, it means that the structural integrity of the sprayed concrete layer is more seriously threatened. Therefore, the target feature focusing strategy focuses on the acceleration element in the crack propagation vector, and adjusts the data processing method to make the importance of this element in subsequent analysis increased.

[0123] For the bending degree change vector of the steel arch, the target feature focusing strategy may focus on the bending degree change at certain key positions. For example, at the connection position of the steel arch and the tunnel wall or the arch top position, the bending degree change of these positions has a greater impact on the stability of the entire supporting structure. The target feature focusing strategy may increase the weight of the bending degree change of these positions in the entire bending degree change vector through specific data processing means, such as amplification or weighting processing of the bending degree change data of these key positions.

[0124] After the target feature focusing strategy processes the features of the plurality of first deformation feature vectors, the result obtained is a plurality of second deformation feature vectors. These second deformation feature vectors have been optimized by correlation selection and feature focusing, and more accurately reflect the deformation of the supporting structure of the tunnel soft rock under the action of lateral pressure, providing a more targeted and effective data basis for subsequent reliability index decision-making.

[0125] This conversion process from the first deformation feature vector to the second deformation feature vector is an important link in the monitoring and analysis of the tunnel soft rock supporting structure, which can effectively extract key information from the original data, provide more accurate and valuable data support for the entire monitoring system, and improve the accuracy and reliability of the deformation monitoring of the tunnel soft rock supporting structure, and ensure the safety and stability of the tunnel project.

[0126] In one possible implementation, before the at least two reference deformation feature conversion strategies are acquired, the method further includes:

[0127] loading the plurality of first deformation feature vectors into a feature processing network that has previously completed knowledge learning, to generate a plurality of extended deformation feature vectors, the extended deformation feature vectors corresponding to a feature space field that is more than the feature space field corresponding to the first deformation feature vectors, the feature processing network being used to extract the extended deformation feature vectors corresponding to the first deformation feature vectors.

[0128] In the monitoring and analysis process of the tunnel soft rock supporting structure, before obtaining at least two reference deformation feature conversion strategies, there are some important pre-operations that help to enrich data features and optimize data structures, so as to more effectively determine the deformation feature conversion strategy subsequently. At the same time, after extracting a plurality of first deformation feature vectors corresponding to the supporting structure data, there are corresponding operations to further process these vectors to improve the availability and representativeness of the data.

[0129] First, before obtaining at least two reference deformation feature conversion strategies, load the plurality of first deformation feature vectors into a feature processing network that has previously completed knowledge learning, to generate a plurality of extended deformation feature vectors. Taking the tunnel project mentioned earlier as an example, in this project, for the monitoring of the tunnel soft rock supporting structure, a plurality of first deformation feature vectors have been obtained, which contain information such as the axial strain vector of the anchor rod, the surface crack expansion vector of the sprayed concrete layer, and the bending degree change vector of the steel arch, etc.

[0130] The feature processing network is a network that has learned a large amount of data and optimized algorithms, and it is designed to extract the extended deformation feature vectors corresponding to the first deformation feature vectors. When these first deformation feature vectors are loaded into the feature processing network, the network processes based on its pre-learned knowledge and algorithms. For example, for the axial strain vector of the anchor rod, the feature processing network may consider factors such as the relationship between axial strain and the stress-strain of the surrounding soft rock, the influence of the installation angle of the anchor rod on the axial strain, and the variation law of the axial strain at different time scales, etc., so as to add more feature information related to the axial strain on the basis of the original axial strain vector, to generate an extended deformation feature vector. These newly added feature information may include the interaction term between the elastic modulus of the soft rock around the anchor rod and the axial strain, the correction value of the axial strain in different seasons (taking into account factors such as temperature), etc.

[0131] The feature space field corresponding to the extended deformation feature vector is more than the feature space field corresponding to the first deformation feature vector. This means that the extended deformation feature vector contains more abundant information. For the surface crack propagation vector of the sprayed concrete layer, the feature processing network can introduce feature fields related to the influence of factors such as the mix proportion of concrete, curing conditions, and environmental humidity on crack propagation. For example, under different concrete mix proportions, the rate and mode of crack propagation will be different. By including these factors in the extended deformation feature vector, the surface crack propagation of the sprayed concrete layer can be more comprehensively described. Similarly, for the curvature change vector of the steel arch, the feature processing network can add feature fields related to factors such as the corrosion degree of the steel arch, the fatigue properties of the steel material, and the uneven distribution of lateral pressure on the arch, thereby obtaining a more complex and information-rich extended deformation feature vector.

[0132] In one possible implementation, after the first deformation feature vector corresponding to the plurality of support structure data is extracted, the method further comprises:

[0133] Obtaining vector attributes corresponding to the plurality of first deformation feature vectors, respectively, the vector attributes representing the feature space field corresponding to the first deformation feature vector.

[0134] Fusing the first deformation feature vectors with the same vector attributes in the first deformation feature vectors corresponding to the plurality of support structure data, respectively, to generate a target deformation feature vector.

[0135] In this embodiment, after extracting the first deformation feature vector corresponding to the plurality of support structure data, some subsequent operations are needed. First, the vector attributes corresponding to the plurality of first deformation feature vectors are obtained, respectively, the vector attributes representing the feature space field corresponding to the first deformation feature vector. Continuing to take the tunnel project as an example, for the axial strain vector of the anchor rod, the vector attributes can include the measurement direction of the strain (such as the longitudinal direction of the axial direction), the accuracy range of the strain measurement (for example, accurate to 0.001 millimeter / meter), the time resolution of the strain data (for example, measured once per hour), and the like. These vector attributes clearly define the specific description of the axial strain vector in the feature space and determine the corresponding feature space field.

[0136] For the surface crack propagation vector of the sprayed concrete layer, the vector attributes can include the plane direction of crack propagation (such as the vertical direction or the horizontal direction), the accuracy of crack width measurement (for example, accurate to 0.1 millimeter), the spatial resolution of crack propagation data (for example, the number of cracks per square meter), and the like. The vector attributes of the curvature change vector of the steel arch can include the reference point of the curvature measurement (such as the arch top or the arch foot), the error range of the curvature measurement (for example, the error is not more than 0.5 degrees), the update frequency of the curvature data (for example, updated once per day), and the like.

[0137] After obtaining these vector attributes, the first deformation feature vectors with the same vector attributes in the plurality of support structure data respectively corresponding first deformation feature vectors are fused to generate a target deformation feature vector. For example, in tunnel engineering, there may be multiple anchor rods in the same measurement accuracy range and the same measurement direction of the axial strain vector. Fusing these axial strain vectors with the same vector attributes can obtain a more representative target deformation feature vector. The fusion process may involve weighted average of data, data filtering or data merging operations.

[0138] For the surface crack propagation vector of the sprayed concrete layer, if there are multiple crack propagation vectors in the same crack plane direction and the same crack width measurement accuracy, by fusing these vectors, the crack propagation conditions at different positions or different time periods can be considered comprehensively to generate a target deformation feature vector that more comprehensively reflects the surface crack propagation conditions of the sprayed concrete layer. Similarly, for the curvature change vector of the steel arch, by fusing the curvature change vectors with the same curvature measurement reference point and the same measurement error range, a target deformation feature vector that more accurately reflects the overall curvature change of the steel arch can be obtained.

[0139] Such fusion operations help to reduce data redundancy while enhancing data representativeness. In the monitoring of tunnel soft rock support structures, by generating target deformation feature vectors, the subsequent analysis process can be simplified, and the efficiency and accuracy of data analysis can be improved. For example, in the subsequent deformation feature conversion strategy determination process, directly using the target deformation feature vector can avoid repeated processing of a large number of similar first deformation feature vectors. At the same time, since the target deformation feature vector integrates the information of multiple vectors with the same vector attributes, it can more accurately reflect the overall deformation of the support structure, thereby providing a more reliable basis for selecting a suitable deformation feature conversion strategy.

[0140] Figure 2 The hardware structure of the tunnel soft rock large deformation support reliability monitoring system 100 for implementing the tunnel soft rock large deformation support reliability monitoring method provided by the embodiment of the present application is shown, as shown in Figure 2 The tunnel soft rock large deformation support reliability monitoring system 100 can include a processor 110, a machine-readable storage medium 120, a bus 130, and a communication unit 140.

[0141] The machine readable storage medium 120 can store data and / or instructions. In some embodiments, the machine readable storage medium 120 can store data acquired from an external terminal. In some embodiments, the machine readable storage medium 120 can store data and / or instructions used by the tunnel soft rock large deformation support reliability monitoring system 100 to perform or use to complete the exemplary methods described in the present application.

[0142] In the implementation process, the one or more processors 110 execute the computer executable instructions stored in the machine readable storage medium 120, so that the processor 110 can perform the tunnel soft rock large deformation support reliability monitoring method of the method embodiments as described above. The processor 110, the machine readable storage medium 120 and the communication unit 140 are connected through the bus 130, and the processor 110 can be used to control the transceiving action of the communication unit 140.

[0143] The implementation process of the processor 110 can refer to the various method embodiments executed by the tunnel soft rock large deformation support reliability monitoring system 100 described above, which have similar implementation principles and technical effects, and will not be described here again.

[0144] In addition, the embodiment of the present application also provides a readable storage medium, wherein computer executable instructions are preset in the readable storage medium, and when the processor executes the computer executable instructions, the tunnel soft rock large deformation support reliability monitoring method as described above is realized.

[0145] It should be noted that, in order to simplify the description of the present application and to help the understanding of one or more embodiments of the present application, in the foregoing description of the embodiments of the present application, various features are sometimes combined into one embodiment, figure or description thereof.

Claims

1. A method for monitoring the reliability of tunnel soft rock support with large deformation, characterized in that, The method includes: Obtain multiple support structure data of soft rock in the tunnel corresponding to the deformation support monitoring request, and extract the first deformation feature vector corresponding to each of the multiple support structure data. The multiple support structure data correspond to multiple structural manifestations. Obtain at least two reference deformation feature transformation strategies, each of which includes at least one step of feature selection and feature focusing on the first deformation feature vector; Based on the monitoring intent corresponding to the deformation support monitoring request and the data tags corresponding to the support structure data, a target deformation feature transformation strategy is determined from the at least two reference deformation feature transformation strategies. The data tags represent the reference tag knowledge points corresponding to the support structure data under the deformation support monitoring request. The target deformation feature transformation strategy is used to transform the deformation features of multiple first deformation feature vectors to generate multiple second deformation feature vectors. Reliability index decision is performed on the plurality of second deformation feature vectors to generate reliability index decision data corresponding to the deformation support monitoring request. The reliability index decision data represents the reliability prediction result corresponding to the deformation support monitoring request generated using the plurality of support structure data. The step of determining the target deformation feature transformation strategy from the at least two reference deformation feature transformation strategies based on the monitoring intent corresponding to the deformation support monitoring request and the data label corresponding to the support structure data includes: The monitoring intent corresponding to the deformation support monitoring request, multiple first deformation feature vectors, and the at least two reference deformation feature transformation strategies are loaded into an artificial intelligence network that has already completed knowledge learning to obtain the target deformation feature transformation strategy. The artificial intelligence network is used to perform strategy matching on the at least two reference deformation feature transformation strategies based on the monitoring intent corresponding to the deformation support monitoring request and the multiple first deformation feature vectors. The target deformation feature transformation strategy includes at least one of a target feature selection strategy and a target feature focusing strategy.

2. The method for monitoring the reliability of tunnel soft rock large deformation support according to claim 1, characterized in that, The artificial intelligence network includes a target matching module and a decision execution module; The step of loading the monitoring intent corresponding to the deformation support monitoring request, multiple first deformation feature vectors, and the at least two reference deformation feature transformation strategies into an artificial intelligence network that has previously completed knowledge learning, to obtain the target deformation feature transformation strategy, includes: The monitoring intent corresponding to the deformation support monitoring request, multiple first deformation feature vectors, and the at least two reference deformation feature transformation strategies are loaded into the target matching module to obtain the matching confidence scores corresponding to the at least two reference deformation feature transformation strategies respectively. The target matching module is used to perform strategy matching on the at least two reference deformation feature transformation strategies based on the monitoring intent corresponding to the deformation support monitoring request and multiple first deformation feature vectors. The target feature selection strategy is determined based on the matching confidence scores corresponding to the at least two reference deformation feature transformation strategies. The monitoring intent corresponding to the deformation support monitoring request, multiple first deformation feature vectors, and the matching confidence scores corresponding to the at least two reference deformation feature transformation strategies are loaded into the decision execution module to obtain the target feature focusing strategy. The decision execution module is used to determine the target feature focusing strategy that matches the monitoring intent based on the matching confidence scores corresponding to the multiple first deformation feature vectors and the at least two reference deformation feature transformation strategies.

3. The method for monitoring the reliability of tunnel soft rock large deformation support according to claim 2, characterized in that, The matching confidence scores corresponding to the at least two reference deformation feature transformation strategies include the first matching confidence scores corresponding to multiple feature selection strategies. The step of determining the target feature selection strategy based on the matching confidence scores corresponding to the at least two reference deformation feature transformation strategies includes: The feature selection strategy with the largest first matching confidence value among the multiple feature selection strategies is output as the target feature selection strategy.

4. The method for monitoring the reliability of tunnel soft rock large deformation support according to claim 2, characterized in that, The matching confidence scores corresponding to the at least two reference deformation feature transformation strategies include the second matching confidence scores corresponding to multiple feature focusing strategies. The step of loading the monitoring intent corresponding to the deformation support monitoring request, multiple first deformation feature vectors, and the matching confidence scores corresponding to the at least two reference deformation feature transformation strategies into the decision execution module to obtain the target feature focusing strategy includes: The monitoring intent corresponding to the deformation support monitoring request, multiple first deformation feature vectors, and the second matching confidence scores corresponding to the multiple feature focusing strategies are loaded into the decision execution module to obtain the target feature focusing strategy.

5. The method for monitoring the reliability of tunnel soft rock large deformation support according to claim 2, characterized in that, Before loading the monitoring intent corresponding to the deformation support monitoring request, multiple first deformation feature vectors, and the at least two reference deformation feature transformation strategies into the target matching module, and obtaining the matching confidence scores corresponding to the at least two reference deformation feature transformation strategies respectively, the method further includes: Obtain multiple template support structure data corresponding to the template deformation support monitoring request, and extract the first deformation feature vector corresponding to each of the multiple template support structure data. The multiple template support structure data correspond to multiple structural manifestations. Obtain at least two template deformation feature transformation strategies, each of which includes at least one of a template feature selection strategy and a template feature focusing strategy; The monitoring intent corresponding to the template deformation support monitoring request, multiple first deformation feature vectors, and the at least two template deformation feature transformation strategies are loaded into the template matching module to obtain the matching confidence scores corresponding to the at least two template deformation feature transformation strategies respectively. The monitoring intent corresponding to the template deformation support monitoring request, multiple first deformation feature vectors, and the matching confidence corresponding to the at least two template deformation feature transformation strategies are loaded into the decision execution module to generate reliability index decision data for the feature focusing strategy. A target incentive mechanism network is generated based on the reliability index decision data of the feature-focusing strategy. The template matching module is optimized based on the target incentive mechanism network to generate the target matching module.

6. The method for monitoring the reliability of tunnel soft rock large deformation support according to any one of claims 1-5, characterized in that, The step of using the target deformation feature transformation strategy to perform deformation feature transformation on multiple first deformation feature vectors to generate multiple second deformation feature vectors includes: The target feature selection strategy is used to select the relevance of the plurality of first deformation feature vectors to generate a plurality of first sub-deformation feature vectors that meet the relevance requirements, wherein the relevance requirements characterize the relevance between the first deformation feature vectors and the monitoring intent; The target feature focusing strategy is used to focus the features of the plurality of first sub-deformation feature vectors to generate the plurality of second deformation feature vectors.

7. The method for monitoring the reliability of tunnel soft rock large deformation support according to any one of claims 1-5, characterized in that, Before acquiring at least two reference deformation feature transformation strategies, the method further includes: The multiple first deformation feature vectors are loaded into a feature processing network that has already completed knowledge learning to generate multiple extended deformation feature vectors. The feature space fields corresponding to the extended deformation feature vectors are more than those corresponding to the first deformation feature vectors. The feature processing network is used to extract the extended deformation feature vectors corresponding to the first deformation feature vectors.

8. The method for monitoring the reliability of tunnel soft rock large deformation support according to any one of claims 1-5, characterized in that, After extracting the first deformation feature vectors corresponding to the multiple support structure data respectively, the method further includes: Obtain the vector attributes corresponding to the plurality of first deformation feature vectors respectively, wherein the vector attributes characterize the feature space fields corresponding to the first deformation feature vectors; The first deformation feature vectors with the same vector attributes in the first deformation feature vectors corresponding to the multiple support structure data are fused to generate the target deformation feature vector.

9. A reliability monitoring system for tunnel soft rock large deformation support, characterized in that, The tunnel soft rock large deformation support reliability monitoring system includes a processor and a memory, the memory and the processor are connected, the memory is used to store programs, instructions or code, and the processor is used to execute the programs, instructions or code in the memory to implement the tunnel soft rock large deformation support reliability monitoring method according to any one of claims 1-8.

Images