Non-uniform deformation prediction method and system suitable for composite stratum tunnel construction period

Abstract

The application discloses a non-uniform deformation prediction method and system suitable for composite stratum tunnel construction period, improves a traditional tunnel surrounding rock deformation displacement monitoring method, extracts geological sketch data and full-section scanning diagram data, classifies and quantifies tunnel deformation monitoring data, defines average deformation coefficients, non-uniform deformation coefficients, sectional deformation coefficients and abnormal deformation coefficients to measure the non-uniform deformation degree of the tunnel, and establishes a non-uniform deformation grading method for the composite stratum tunnel construction period; based on a PSO-LSSVM combined algorithm and a numerical simulation method, taking a stratum boundary line inclination, a stratum proportion, a ground stress and a pore water pressure as inputs and taking index grades as outputs, the correlation degree of each grading coefficient and the tunnel non-uniform deformation grade and the deformation grade discrimination standard are calculated, and a non-uniform deformation prediction model for the composite stratum tunnel surrounding rock is established; and the collected data are input into the PSO-LSSVM composite stratum tunnel deformation intelligent prediction system to obtain a prediction result.

Description

Technical Field

[0001] This invention belongs to the field of soil and rock classification and deformation level prediction, and relates to a method and system for predicting non-uniform deformation during the construction period of tunnels in composite strata. Background Technology

[0002] The lithology of surrounding rocks in complex strata changes frequently, and the physical and mechanical properties of different surrounding rocks vary significantly. During excavation, uneven stress release between soft and hard surrounding rocks can easily lead to non-uniform large deformations. Frequent strata changes also pose a serious safety threat to TBM equipment and workers. Real-time monitoring, effective classification, and intelligent prediction of tunnel deformation in complex strata provide a reference for construction personnel to optimize the initial foundation support strength, determine the scope of local reinforcement support, and provide a basis for the prevention and control of tunnel deformation and failure.

[0003] According to the inventors, traditional monitoring and measurement data only reflect vertical settlement and horizontal convergence, but cannot reflect the nonlinear changes in the tunnel axis and the radial direction of the tunnel face. In other words, it is difficult to monitor the non-uniform deformation of tunnels in composite strata in real time. The surrounding rock deformation classification method also rarely considers engineering geological strata parameters and the non-uniform deformation of each segment. Traditional tunnel deformation level prediction mainly uses empirical formulas and numerical simulation methods. However, since the non-uniform large deformation of tunnels in composite strata is affected by a variety of factors such as engineering geological factors and construction factors, it is a nonlinear problem. Traditional empirical formulas and numerical simulation methods have certain limitations and it is difficult to predict the degree of non-uniformity of tunnel deformation. Summary of the Invention

[0004] To address the aforementioned problems, this invention proposes a method and system for predicting non-uniform deformation during the construction of tunnels in composite strata. This invention enables real-time monitoring, effective classification, and intelligent prediction of non-uniform deformation in tunnels in composite strata. The method has been validated in actual tunnel engineering projects, and the classification results show good agreement with the actual monitoring levels, demonstrating good real-time performance and applicability.

[0005] According to some embodiments, the present invention adopts the following technical solution:

[0006] In a first aspect, the present invention provides a method for classifying non-uniform deformation during the construction of tunnels in composite strata, the steps of which are as follows:

[0007] Step 1: Use the tunnel monitoring system to obtain deformation data of each existing tunnel, and obtain various deformation coefficients based on the deformation data.

[0008] Step 2: Obtain simulated geological deformation data using numerical simulation data;

[0009] Step 3: Use the data from Step 1 and Step 2 to build the initial database;

[0010] Step 4: Based on the Particle Swarm Optimization-Least Squares Support Vector Machine (PSO-LSSVM) joint intelligent algorithm, construct an LSSVM learning model under the classification theory of non-uniform deformation of composite strata tunnels, and input the learning sample data from the initial database into the LSSVM learning model.

[0011] Step 5: Optimize the parameters of the LSSVM learning model using the PSO algorithm, and use the optimized parameters to establish a PSO-LSSVM prediction model as a real-time online classification model for non-uniform deformation of tunnels in composite strata. Input the training set of samples to be tested into the PSO-LSSVM prediction model to obtain real-time classification output results, and predict the non-uniform deformation of tunnels in composite strata during construction based on the output results.

[0012] As a further technical solution, the complete deformation data of the tunnel cross section in the composite strata is obtained by comparing the difference between the geological sketch and the tunnel cross section scanning data after the initial excavation and the tunnel deformation stabilization in the full-section scanning image.

[0013] The method for establishing step one is as follows: based on the deformation data of the composite stratum tunnel, determine the average deformation coefficient. The non-uniform deformation coefficient ξ, piecewise deformation coefficient ζ, and abnormal deformation coefficient ψ replace the traditional deformation evaluation indices. Among them, the average deformation coefficient... The overall situation of non-uniform deformation is characterized by the non-uniform deformation coefficient ξ, which measures the non-uniformity of cross-sectional deformation. The segmented deformation coefficient ζ determines the location of abnormal encroachment deformation in the tunnel, and the abnormal deformation coefficient ψ measures the magnitude of the encroachment deformation.

[0014] In step three, part of the deformation data in the initial database is obtained from the field monitoring data in step one, and the other part is modeled and calculated by the numerical simulation software in step two to reflect the influence of different levels of factors such as tunnel dip angle, proportion, ground stress, and pore water pressure on deformation.

[0015] In step four, the LSSVM learning model transforms inequality constraints into equality constraints, converts the quadratic programming problem into a linear equation problem, and integrates LSSVM parameter optimization and prediction into the PSO algorithm to improve the algorithm's accuracy.

[0016] The optimization method of the PSO algorithm in step five is as follows: use the particle swarm algorithm to find the optimal penalty factor c and kernel parameter g, then iterate until the global fitness of the particle swarm is optimal, obtain the optimal penalty factor c and kernel parameter g, and then predict each index.

[0017] In step five, the obtained optimization parameters, penalty factor c and kernel parameter g, are used to establish a prediction model as a real-time classification model for non-uniform deformation of tunnels in composite strata. The accurate monitoring data of crown settlement and horizontal convergence, geological sketch data and full-section scanning data sample training set obtained in real time in steps one and two are used. The classification coefficients of non-uniform deformation of tunnels in composite strata mentioned in step one are input into the LSSVM prediction model to obtain the real-time classification output results.

[0018] Secondly, the present invention provides a system for predicting the level of non-uniform deformation during the construction of tunnels in composite strata, comprising:

[0019] The sample database construction module is configured to use the tunnel monitoring system to acquire deformation data of each existing tunnel, obtain various deformation coefficients based on the deformation data, acquire simulated geological deformation data using numerical simulation data, and establish an initial database based on the deformation coefficients and simulated geological deformation data.

[0020] The prediction model construction module is configured to construct an LSSVM learning model based on the particle swarm optimization-least square support vector machine (PSO-LSSVM) joint intelligent algorithm for the classification theory of non-uniform deformation of tunnels in composite strata. The learning sample data from the initial database is input into the LSSVM learning model. The parameters of the LSSVM learning model are optimized using the PSO algorithm, and the optimized parameters are used to establish a PSO-LSSVM prediction model as a real-time online classification model for non-uniform deformation of tunnels in composite strata.

[0021] The prediction module inputs the training set of the sample to be tested into the PSO-LSSVM prediction model to obtain real-time graded output results. Based on the output results, it predicts the non-uniform deformation during the construction period of the composite stratum tunnel. In the third aspect, the present invention also provides a computer-readable storage medium on which a computer program is stored. The instruction is loaded and executed by the processor of the terminal device. The method for predicting the non-uniform deformation level of the composite stratum tunnel is described above.

[0022] Fourthly, the present invention also provides a terminal device, including a processor and a computer-readable storage medium, wherein the processor is used to implement various instructions; the computer-readable storage medium is used to store multiple instructions, which are adapted to be loaded by the processor and executed by the processor to perform the steps of the method for predicting non-uniform deformation of tunnels in composite strata.

[0023] Compared with the prior art, the beneficial effects of this disclosure are as follows:

[0024] 1. This disclosure, based on actual tunnel engineering projects, extensively collects geological information, construction data, and monitoring measurement data from these projects to enrich the data sample set of non-uniform deformation in tunnels in composite strata, thereby establishing a widely applicable evaluation system for non-uniform deformation in tunnels in composite strata. It summarizes the variation patterns of tunnel non-uniform deformation indices under the influence of different stratum parameters, establishing an initial sample library with broad coverage and real-time performance. An optimized sample database is established, and the Particle Swarm Optimization (PSO) algorithm is scientifically sound and significantly improves the accuracy of prediction results from the Least Squares Support Vector Machine (LSSVM) learning model. This constructs a prediction model for non-uniform deformation in tunnels in composite strata, which, combined with subsequent real-time monitoring data, exhibits unique advantages and high prediction accuracy.

[0025] 2. When predicting non-uniform deformation of tunnels in complex strata, the PSO-LSSVM model can be obtained simply by inputting the calculated average deformation coefficient, non-uniform deformation coefficient, segmented deformation coefficient, and abnormal deformation coefficient based on the geological conditions, construction conditions, and deformation monitoring data of the tunnel to be predicted. This method is simple, reliable, and can accurately predict the deformation of tunnels in complex strata, providing a reference for construction personnel to optimize the initial foundation support strength and determine the scope of local reinforcement.

[0026] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0027] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0028] Figure 1 This is a schematic diagram of a method for predicting non-uniform deformation during the construction period of a tunnel in composite strata, provided by an embodiment of the present invention.

[0029] Figure 2 This is a schematic diagram of the non-uniform deformation curve during the construction period of a composite stratum tunnel, provided in an embodiment of the present invention. Detailed Implementation

[0030] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0031] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0032] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments of the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. Furthermore, it should be understood that the terms “comprising” and “having”, and any variations thereof, are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0033] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0034] Based on extensive collection of real-time acquired surrounding rock geological parameters and deformation data from existing and under-construction composite stratum tunnels, this invention proposes a method for predicting non-uniform deformation during the construction period of composite stratum tunnels, combining particle swarm optimization (PSO) and least squares support vector machine (LSV) methods. This invention comprehensively summarizes and analyzes the non-uniform deformation patterns and geological variation characteristics of composite stratum tunnels by combining actual engineering geological conditions, construction conditions, and deformation data. It defines average deformation coefficient, non-uniform deformation coefficient, segmented deformation coefficient, and abnormal deformation coefficient to replace traditional deformation evaluation indicators, establishing an initial sample database for non-uniform deformation of composite stratum tunnels. Through the LSSVM learning model, inequality constraints are transformed into equality constraints, and the quadratic programming problem is transformed into a linear equation problem. Simultaneously, LSSVM parameter optimization and prediction are integrated into the PSO algorithm to improve algorithm accuracy and optimize the original sample database. Using the calculated grading coefficient index as input parameters and the non-uniform deformation level as output parameters, a PSO-LSSVM prediction model for non-uniform large deformation of composite stratum tunnels is constructed. This method relies on actual engineering tunnel projects to obtain historical data corresponding to deformation indices, as well as geological conditions, construction conditions, and deformation data of existing tunnels. It utilizes field data and numerical simulation data to acquire geological change characteristics, establishing a sample database with broad applicability and representativeness. The particle swarm optimization algorithm is used to find the globally optimal penalty factor and kernel parameters. The optimized result shows high accuracy, demonstrating that particle swarm optimization significantly improves the accuracy of least squares support vector machines (LSSVMs), greatly enhancing the accuracy of LSSVM learning model predictions. The sample data used in this method is authentic, rich in information, and representative. The database optimization method is scientifically sound, and the prediction method possesses unique advantages, resulting in high prediction accuracy. Real-time monitoring, effective classification, and intelligent prediction of deformation in tunnels in composite strata provide a reference for construction personnel to optimize initial foundation support strength and determine the scope of local reinforcement, offering a basis for preventing tunnel deformation and damage. It has significant guiding significance for predicting non-uniform deformation in tunnels in composite strata and for safe construction.

[0035] This embodiment provides a method for predicting non-uniform deformation during the construction period of tunnels in composite strata, such as... Figure 1 As shown, it specifically includes:

[0036] (1) Select the real-time acquired geological parameters and deformation data of the surrounding rock to obtain the deformation coefficient, wherein the deformation coefficient includes the average deformation coefficient. The non-uniform deformation coefficient ξ, piecewise deformation coefficient ζ, and abnormal deformation coefficient ψ replace the traditional deformation evaluation indices. Among them, the average deformation coefficient... The overall situation of non-uniform deformation is characterized by the non-uniform deformation coefficient ξ, which measures the non-uniformity of cross-sectional deformation. The segmented deformation coefficient ζ determines the location of abnormal encroachment deformation in the tunnel, and the abnormal deformation coefficient ψ measures the magnitude of the encroachment deformation.

[0037] The average deformation coefficient in step one for:

[0038]

[0039] In the formula, L is the circumferential length of the tunnel (excluding the invert arch), δ(l) is the tunnel deformation function, which represents the distance between the actual tunnel cross-section and the initial cross-section at the corresponding position, and R is the initial radius of the tunnel.

[0040] The non-uniform deformation coefficient ξ in step one is:

[0041]

[0042] In the formula, S 侵 S represents the encroachment area of ​​the cross section. 扩 n1 is the area of ​​the cross-section expansion, and l is the number of encroachment areas of the cross-section. n1 Let δ1(l) be the arc length of the encroaching portion of the cross-section, δ1(l) be the deformation function of the encroaching portion of the cross-section, and n2 be the number of outward expansion regions of the cross-section. n2 Let δ2(l) be the arc length of the outer portion of the cross section, and let δ2(l) be the deformation function of the outer portion of the cross section.

[0043] The segmented deformation coefficient ζ in step one is:

[0044]

[0045] In the formula, This represents the average intrusion limit for the corresponding area. R represents the average expansion amount of the corresponding region's outer expansion portion. n L represents the equivalent radius of the tunnel in the corresponding area, d represents the tunnel's allowable deformation, and L represents the equivalent radius of the tunnel's deformation. n This represents the arc length of the corresponding region.

[0046] The abnormal deformation coefficient ψ in step one is:

[0047]

[0048] In the formula, S 侵 L represents the area of ​​the encroached portion. 侵 This indicates the arc length corresponding to the encroachment limit.

[0049] (2) Based on the above-mentioned field data and numerical simulation data, geological change characteristics were obtained, and an initial database was established. The initial database was divided into the following levels: the strata ratio was divided into three different categories according to the soft rock to hard rock ratio of 1:1, 1:2, and 2:1. The applied geostress was divided into three levels: 0 MPa, 4 MPa, and 8 MPa. The groundwater condition was determined by adjusting the pore water pressure, which was divided into three levels: 0 MPa, 0.5 MPa, and 1 MPa. Taking the intersection of the strata boundary line and the central horizontal line of the tunnel height as the origin, with the horizontal plane as 0°, counterclockwise rotation was positive and clockwise rotation was negative, and the dip angle of the rock strata interface was divided into five angles: 0°, 45°, 90°, 135°, and 180°.

[0050] (3) Based on the Particle Swarm Optimization-Least Squares Support Vector Machine (PSO-LSSVM) joint intelligent algorithm, construct an LSSVM learning model under the classification theory of non-uniform deformation of composite strata tunnels, and input the obtained learning sample data into the LSSVM learning model.

[0051] In step three, the LSSVM learning model is as follows: LSSVM transforms inequality constraints into equality constraints, and the quadratic programming problem into a linear equation problem. When utilizing the structural risk principle, the optimization problem of LSSVM becomes:

[0052]

[0053] The corresponding constraints are as follows:

[0054] y i [(w T *x i )+b]=1-ξ i i = 1, 2, g, n

[0055] In the formula, ξ i γ represents the relaxation factor, γ is the regularization parameter, also known as the penalty coefficient, and n is the population size.

[0056] Establish the Lagrange equation:

[0057]

[0058] Minimize L with respect to ω, b, ξ, α, and the partial derivative is 0:

[0059]

[0060] Based on the four conditions in the above equation, a system of linear equations can be established to solve for a and b:

[0061]

[0062] Where Ω is the kernel matrix, I is the identity matrix, and y and a are vectors. We obtain:

[0063]

[0064] The final classification decision model of LSSVM is as follows:

[0065]

[0066] In step three, the learning sample data is the learning samples selected from the initial database. The samples in the database are combined into a sample set, considering n particles X = (X1, X2, ..., X...). n ), where the position of the i-th particle is X. i =(x i1 ,x i2 ,…,x id ) T X i Introducing the fitness function f(X) i ), calculate the fitness value of the particle's position. The velocity of the i-th particle is V. i =(v i1 ,v i2 ,…,v id ) T The individual extreme value of the i-th particle is P. i =(P i1 ,P i2 ,…,P id ) T The global extremum is P. g =(P g1 ,P g2 ,…,P gd ) T During the optimization process, the velocity and position of particles are updated by tracking the individual optimal particle and the swarm optimal particle using the following equation.

[0067] In step three: n is the number of samples, X = (X1, X2, ..., X...) n ) represents a four-dimensional vector of deformation grading coefficients, including the average deformation coefficient. Uneven deformation coefficient ξ, piecewise deformation coefficient ζ, and abnormal deformation coefficient ψ, f(X) i f(X) is the output vector and its value is the corresponding deformation level; the deformation level of the learned sample is f(X). iThe result was obtained through the non-uniform deformation classification method for composite strata tunnels in step one.

[0068] In step three, the fitness function f(X) i ):

[0069]

[0070] Where k is the current iteration number, n is the population size, and X k Let be the position of the k-th particle.

[0071] The positions of the sample particles in step three are:

[0072] X k+1 id =X k id +V k id

[0073] Where X k id and V k id This represents the individual extreme values ​​of the position and velocity of the i-th particle in the (k+1)-th iteration.

[0074] The velocity of the i-th sample particle in the (k+1)-th iteration of step three. for:

[0075]

[0076] Where ω is the inertia weight, controlling how much velocity the i-th particle inherits. Let be the velocity of the i-th sample particle after the k-th iteration. and Represents the individual extremum and global extremum of the i-th particle. and Let c1 and c2 be the individual and global positions of the k-th particle, respectively. c1 and c2 are acceleration factors controlling the velocity calculation. r1 and r2 are two random numbers ranging from 0 to 1, which can increase the randomness of the search. To prevent blind searching of particles, position and velocity are usually limited to [-X]. max X max ]、[-V max V max ].

[0077] (4) The parameters of the LSSVM learning model are optimized using the PSO algorithm. The optimized parameters are used to establish a prediction model as a real-time online classification model for non-uniform deformation of composite strata tunnels. The sample training set composed of real-time sample data is input into the LSSVM prediction model to obtain the real-time classification output results.

[0078] Table 1. Criteria for Deformation Levels

[0079]

[0080] Furthermore, in step (1), the required data includes historical data corresponding to each graded index, geological conditions, construction conditions and deformation data of each existing tunnel, definition of average deformation coefficient, uneven deformation coefficient, segmented deformation coefficient and abnormal deformation coefficient, and establishment of the deformation level discrimination criteria of the prediction model in Table 1.

[0081] Furthermore, in step (2), geological change characteristics are obtained using field data to establish an initial sample database of non-uniform deformation of tunnels in composite strata. Actual engineering section data is shown in Table 2.

[0082] Table 2. Actual Engineering Tunnel Measurement Data Set

[0083]

[0084]

[0085] Furthermore, in step (3), the on-site monitoring data and geological model data of the project are substituted into the LSSVM learning model, configured to use the calculated average deformation coefficient, uneven deformation coefficient, segmented deformation coefficient and abnormal deformation coefficient as input parameters, and the deformation level of each segment as output parameters, to establish an optimized sample database and construct a composite stratum tunnel non-uniform deformation prediction model.

[0086] Furthermore, in step (4), the sample dataset composed of real-time sample data to be tested is input into the LSSVM prediction model to obtain the real-time hierarchical output results, as shown in Table 3.

[0087] Table 3 Dataset of Prediction Results for Actual Engineering Tunnels

[0088]

[0089] Furthermore, in step (4), the prediction levels obtained from Table 3 show the following accuracy rates in the training model: average deformation level 93.33% (14 / 15), abnormal deformation level 86.67% (13 / 15), uneven deformation level 93.33% (14 / 15), and segmented deformation level 93.33% (14 / 15). The classification accuracy of the training set meets the testing requirements. When 15 sets of test sample data are input into the SVM prediction model, and their classification results are compared with traditional surrounding rock classification results, the accuracy reaches 100%, indicating good testing performance.

[0090] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0091] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0092] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0093] The above description is merely a preferred embodiment of this disclosure and is not intended to limit this disclosure. Various modifications and variations can be made to this disclosure by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

[0094] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A method for predicting non-uniform deformation during the construction period of tunnels in composite strata, characterized in that: Includes the following steps: Step 1 involves using a tunnel monitoring system to acquire deformation data for each existing tunnel, and then obtaining various deformation coefficients based on this data; these deformation coefficients include the average deformation coefficient. Coefficient of non-uniform deformation Segmented deformation coefficient and abnormal deformation coefficient ; The average deformation coefficient for: In the formula, L is the circumferential length of the tunnel, δ(l) is the tunnel deformation function, which represents the distance between the actual tunnel cross-section and the initial cross-section at the corresponding position, and R is the initial radius of the tunnel. The coefficient of non-uniform deformation for: In the formula, S 侵 The encroachment area is the cross-sectional area. S 扩 For the area of ​​the cross-section expansion, n 1 represents the number of encroachment zones in the cross-section. l n1 The arc length of the section encroaching on the boundary. δ 1( l ) represents the deformation function of the encroaching portion of the cross-section. n 2 represents the number of areas extending beyond the cross-section. l n2 The arc length of the outer portion of the cross-section. δ 2( l ) represents the deformation function of the outer portion of the cross-section; The segmented deformation coefficient for: In the formula, This represents the average intrusion limit for the corresponding area. This represents the average outward expansion of the corresponding area. R n This represents the equivalent radius of the tunnel in the corresponding area. d Allowing for deformation in the tunnel, L n This represents the arc length of the corresponding region; The abnormal deformation coefficient for: In the formula, S 侵 Indicates the area of ​​the encroached portion. L 侵 Indicates the arc length corresponding to the encroachment limit; Step 2: Obtain simulated geological deformation data using numerical simulation data; Step 3: Use the data from Step 1 and Step 2 to build the initial database; Step 4: Based on the Particle Swarm Optimization-Least Squares Support Vector Machine (PSO-LSSVM) joint intelligent algorithm, construct an LSSVM learning model under the classification theory of non-uniform deformation of composite strata tunnels, and input the learning sample data from the initial database into the LSSVM learning model. Step 5: Optimize the parameters of the LSSVM learning model using the PSO algorithm, and use the optimized parameters to establish a PSO-LSSVM prediction model as a real-time online classification model for non-uniform deformation of tunnels in composite strata. Input the training set of samples to be tested into the PSO-LSSVM prediction model to obtain real-time classification output results, and predict the non-uniform deformation of tunnels in composite strata during construction based on the output results.

2. The method for predicting non-uniform deformation during tunnel construction in composite strata as described in claim 1, characterized in that: The method for obtaining deformation data in step one is as follows: The complete deformation data of the tunnel cross-section is obtained by comparing the geological sketch with the full-section scanning data of the tunnel after initial excavation and after the tunnel deformation has stabilized.

3. The method for predicting non-uniform deformation during tunnel construction in composite strata as described in claim 1, characterized in that: The PSO-LSSVM prediction model in step five is as follows: The PSO algorithm optimizes the LSSVM learning model, where: X =( X 1, X 2, …, X n ) represents a four-dimensional vector of deformation grading coefficients, including the average deformation coefficient. Coefficient of non-uniform deformation Segmented deformation coefficient and abnormal deformation coefficient , f ( X i ) is the output vector and its value is the corresponding deformation level; fitness function f ( X i ): Where k is the current iteration number, For population size, This represents the position of the k-th particle. The positions of the sample particles are: in and This represents the individual extreme values ​​of the position and velocity of the i-th particle in the (k+1)-th iteration; During the optimization process, the velocity and position of particles are updated by tracking the individual optimal particle and the swarm optimal particle using the following equation: the velocity of the i-th sample particle in the (k+1)-th iteration. for: ； Where ω is the inertia weight, controlling how much velocity the i-th particle inherits. Let be the velocity of the i-th sample particle after the k-th iteration. and Represents the individual extremum and global extremum of the i-th particle. and Let be the individual and global positions of the k-th particle. As an acceleration factor, it controls the speed calculation; These are two random numbers, ranging from 0 to 1, which can increase the randomness of the search.

4. The method for predicting non-uniform deformation during tunnel construction in composite strata as described in claim 1, characterized in that: In step four, the LSSVM learning model transforms inequality constraints into equality constraints, converts the quadratic programming problem into a linear equation problem, and integrates LSSVM parameter optimization and prediction into the PSO algorithm to improve the algorithm's accuracy.

5. The method for predicting non-uniform deformation during tunnel construction in composite strata as described in claim 1, characterized in that: In step five, the particle swarm optimization algorithm is used to find the optimal penalty factor. c and kernel parameters g Then iterate until the particle swarm fitness is globally optimal, and stop when the optimal penalty factor is obtained. c and kernel parameters g Then, predictions are made for each indicator, and the obtained optimization parameters and penalty factors are used. c and kernel parameters g Establish a PSO-LSSVM prediction model.

6. A system for predicting the level of non-uniform deformation during the construction of tunnels in composite strata, characterized in that: include: The sample database construction module is configured to use the tunnel monitoring system to acquire deformation data of each existing tunnel and obtain various deformation coefficients based on the deformation data. Numerical simulation data is used to obtain simulated geological deformation data. An initial database was established based on deformation coefficients and simulated geological deformation data; The deformation coefficient includes the average deformation coefficient. Coefficient of non-uniform deformation Segmented deformation coefficient and abnormal deformation coefficient ; The average deformation coefficient for: In the formula, L is the circumferential length of the tunnel, δ(l) is the tunnel deformation function, which represents the distance between the actual tunnel cross-section and the initial cross-section at the corresponding position, and R is the initial radius of the tunnel. The coefficient of non-uniform deformation for: In the formula, S 侵 The encroachment area is the cross-sectional area. S 扩 For the area of ​​the cross-section expansion, n 1 represents the number of encroachment zones in the cross-section. l n1 The arc length of the section encroaching on the boundary. δ 1( l ) represents the deformation function of the encroaching portion of the cross-section. n 2 represents the number of areas extending beyond the cross-section. l n2 The arc length of the outer portion of the cross-section. δ 2( l ) represents the deformation function of the outer portion of the cross-section; The segmented deformation coefficient for: In the formula, This represents the average intrusion limit for the corresponding area. This represents the average outward expansion of the corresponding area. R n This represents the equivalent radius of the tunnel in the corresponding area. d Allowing for deformation in the tunnel, L n This represents the arc length of the corresponding region; The abnormal deformation coefficient for: In the formula, S 侵 Indicates the area of ​​the encroached portion. L 侵 Indicates the arc length corresponding to the encroachment limit; The prediction model building module is configured to construct an LSSVM learning model based on the particle swarm optimization-least square support vector machine (PSO-LSSVM) joint intelligent algorithm for the classification theory of non-uniform deformation of tunnels in composite strata. The learning sample data from the initial database is input into the LSSVM learning model. The parameters of the LSSVM learning model are optimized using the PSO algorithm, and the optimized parameters are used to establish a PSO-LSSVM prediction model as a real-time online classification model for non-uniform deformation of tunnels in composite strata. The prediction module inputs the training set of the samples to be tested into the PSO-LSSVM prediction model to obtain real-time hierarchical output results, and predicts the non-uniform deformation during the construction period of the tunnel in the composite strata based on the output results.

7. A computer-readable storage medium, characterized in that: It stores a computer program, characterized in that the computer program is loaded and executed by the processor of the terminal device, which is a method for predicting the non-uniform deformation level of tunnels in composite strata as described in any one of claims 1-6.

8. A terminal device, characterized in that: The device includes a processor and a computer-readable storage medium, the processor being used to implement various instructions; the computer-readable storage medium being used to store multiple instructions adapted for loading by the processor and executing the steps of the method for predicting non-uniform deformation of tunnels in composite strata as described in any one of claims 1-6.

Images