Diversion tunnel soft rock supporting structure multi-objective intelligent optimization method

Abstract

The invention relates to the technical field of tunnel support optimization engineering, in particular to a diversion tunnel soft rock support structure multi-target intelligent optimization method, which comprises the following steps of: establishing a diversion tunnel excavation support three-dimensional model, and establishing a vault settlement proxy model by taking vault settlement and support material cost as target functions; then, a double-target optimization model of the diversion tunnel soft rock support structure is established in combination with the support material cost, and the double-target optimization model is solved by adopting a multi-target optimization algorithm; and finally, an optimal support parameter is decided by adopting entropy weight-TOPSIS. The method has the beneficial effects that the method is used for guiding diversion tunnel support optimization, and the problem of how to effectively guarantee the tunnel safety and greatly reduce the cost of support materials is solved, so that the rationality and scientificity of tunnel support construction are guaranteed.

Description

Technical Field

[0001] The invention relates to the field of tunnel support optimization engineering, and in particular to a multi-objective intelligent optimization method for a soft rock support structure of a water diversion tunnel. Background Art

[0002] Water diversion tunnels occupy an important position in the construction of water conservancy projects in my country. In recent years, water diversion tunnels have gradually developed in the direction of length, size and depth. For water diversion tunnels, their stability during construction is closely related to the overall project safety. In order to ensure the safety of the tunnel, it is often necessary to support the rock around the tunnel. Overly conservative design of support parameters will lead to economic waste, and vice versa will cause safety problems. How to balance safety and economy and scientifically and rationally optimize tunnel support parameters is crucial. Tunnel support optimization methods can be divided into traditional optimization methods and intelligent optimization methods. Among them, the intelligent optimization method deeply integrates optimization theory with computer technology. It not only has fast convergence speed and high calculation accuracy, but also has strong universality.

[0003] At present, most methods for tunnel support optimization use single-objective optimization, or convert multiple objectives into single-objective methods. If only single-objective optimization is considered, a single calculation can only obtain a single solution to a certain problem. In fact, the surrounding rock of the tunnel construction section is complex and changeable. Engineers hope to select suitable support parameters from multiple sets of feasible solutions according to actual conditions to adapt to the changes in surrounding rock. If the above method is used, the complex relationship between multiple objectives is ignored, and the calculation results are of low accuracy.

[0004] Therefore, there is an urgent need for a more scientific and intelligent method that can perform multi-objective optimization to solve the above problems. Summary of the invention

[0005] The purpose of the present invention is to provide a multi-objective intelligent optimization method for soft rock support structure of a water diversion tunnel to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: a multi-objective intelligent optimization method for soft rock support structure of a water diversion tunnel, characterized in that the method comprises the following steps: establishing a three-dimensional model for numerical analysis of tunnel excavation support for numerical simulation, and using an orthogonal test method to generate a sample data set;

[0007] The sample data was trained using the support vector machine (SVM) supervised learning method to identify the highly nonlinear relationship between tunnel support parameters and arch settlement.

[0008] A proxy model of the arch settlement G(x) is established, that is, G(x)=SVM(x), which can reflect the safety of tunnel excavation support.

[0009] Preferably, the method further comprises:

[0010] The support material cost model J(x) is established to evaluate and improve the economic efficiency of tunnel support, where:

[0011] J(x)=J1(x)+J2(x)+J3(x)

[0012] In the formula: J1(x) represents the cost of steel; J2(x) represents the cost of concrete; J3(x) represents the cost of other materials; x represents different support parameters.

[0013] Preferably, the method further comprises:

[0014] A dual-objective optimization model for the soft rock support structure of the water diversion tunnel is established. This model is a multi-objective optimization model that is very helpful in improving the safety and economy of tunnel support at the same time. The specific dual-objective optimization model and support parameter value range of the soft rock support structure of the water diversion tunnel are as follows:

[0015] minF(x)=[G(x),J(x)] T ,x=[s a ,s b ,s c ,s d ] T ∈R 4

[0016]

[0017] Where G(x) is the established vault settlement proxy model, J(x) is the cost of support materials,

[0018] s a is the longitudinal spacing of anchor rods (m), s b is the circumferential spacing of anchor rods (m), s c is the steel support spacing (m), s d is the thickness of the concrete spray layer (cm).

[0019] Preferably, the method adopts the multi-objective multiverse algorithm MOMVO to solve the established multi-objective optimization model that takes into account the vault settlement G(x) and the support material cost J(x), and obtains the Pareto solution set; the entropy weight method is used to determine the weight of each objective, and then the TOPSIS comprehensive ranking comparison is used to finally decide the optimal support plan.

[0020] Compared with the prior art, the present invention has the following beneficial effects:

[0021] The multi-objective intelligent optimization method for the soft rock support structure of a water diversion tunnel proposed in the present invention establishes a three-dimensional model of excavation and support of a water diversion tunnel, and establishes a vault settlement proxy model with vault settlement and support material cost as objective functions; then, a dual-objective optimization model of the soft rock support structure of the water diversion tunnel is established in combination with the support material cost, and a multi-objective optimization algorithm is used to solve the above dual-objective optimization model; finally, the entropy weight-TOPSIS is used to decide the optimal support parameters; the method of the present invention is used to guide the optimization of water diversion tunnel support, and solves the problem of how to effectively ensure the safety of the tunnel while significantly reducing the cost of support materials, thereby ensuring the rationality and scientificity of the tunnel support construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0023] In order to make the purpose and technical solution of the present invention clearly and completely described, and the advantages more clearly understood, the embodiments of the present invention are further described in detail with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are part of the embodiments of the present invention, rather than all of the embodiments, and are only used to explain the embodiments of the present invention, and are not used to limit the embodiments of the present invention. All other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0024] For example, see Figure 1 The present invention provides a technical solution: a multi-objective intelligent optimization method for a soft rock support structure of a water diversion tunnel, the method comprising the following steps:

[0025] First, a three-dimensional model of tunnel excavation support numerical analysis is established for numerical simulation, and a sample data set is generated using the orthogonal test method. Then, the sample data is trained using the support vector machine (SVM) supervised learning method to identify the highly nonlinear relationship between tunnel support parameters and vault settlement, and a proxy model of vault settlement G(x) is established, i.e., G(x) = SVM(x). This proxy model can reflect the safety of tunnel excavation support.

[0026] The support material cost model J(x) is established to evaluate and improve the economic efficiency of tunnel support.

[0027] J(x)=J1(x)+J2(x)+J3(x)(1)

[0028] In the formula: J1(x) represents the cost of steel; J2(x) represents the cost of concrete; J3(x) represents the cost of other materials; x represents different support parameters.

[0029] In order to take into account the safety and economy of tunnel support engineering, a dual-objective optimization model of the soft rock support structure of the water diversion tunnel is established. This model is a multi-objective optimization. Compared with the single-objective optimization form, it makes the objective function more comprehensive and the optimization result more accurate, which is very helpful to improve the safety and economy of tunnel support at the same time. The specific dual-objective optimization model and parameter value range of the soft rock support structure of the water diversion tunnel are as follows:

[0030]

[0031] Where G(x) is the established vault settlement proxy model, J(x) is the cost of support materials, s a is the longitudinal spacing of anchor rods (m), s b is the circumferential spacing of anchor rods (m), s c is the steel support spacing (m), s d is the thickness of the concrete spraying layer (m).

[0032] The multi-objective multiverse algorithm (MOMVO) is used to solve the multi-objective optimization model established above that takes into account both the vault settlement G(x) and the support material cost J(x), and the Pareto solution set is obtained. The entropy weight method is used to determine the weight of each objective, and then the optimal support scheme is finally decided based on the TOPSIS comprehensive ranking comparison.

[0033] Among them, a multi-objective optimization algorithm is used to solve the above dual-objective optimization model to obtain a set of support parameter solution sets. The algorithm is based on theoretical knowledge such as the Big Bang and the multiverse. It is assumed that the multiverse originated from the Big Bang. There are black holes, white holes and wormholes in the multiverse. Black holes can absorb all objects in the universe, white holes release all objects, and wormholes are channels connecting different universes and transmitting objects. The specific solution steps are as follows:

[0034] (1) Initialization: The initialization of a multiverse is described as:

[0035]

[0036] In the formula: U represents a multiverse, which represents the set of the entire support scheme; n represents the number of parallel universes contained in the multiverse U, corresponding to different support schemes; d represents the number of substances contained in each universe, corresponding to different support parameters such as the circumferential spacing of anchor rods, the longitudinal spacing of anchor rods, the spacing of steel supports, and the thickness of concrete spraying layer.

[0037] (2) Iteration and update:

[0038] In one iteration, P j,j There are two ways to process it, one is based on the roulette mechanism, and the other is to keep the original value. The formula is as follows:

[0039]

[0040] Where: P i,j represents the jth substance in the i-th universe, representing a set of support parameters; Norm(U i ) represents the normalized expansion rate of the ith universe; P k,j Represents the jth substance in the kth universe determined by the roulette mechanism, and represents the new support parameter combination formed after random combination; the random number rand1 takes values ​​between [0,1].

[0041] White holes would be searched in a spiral pattern with reference to the normalized expansion rate of the Universe, with objects with higher expansion rates more likely to be transported through white holes, while objects with lower expansion rates more likely to be transported through black holes.

[0042] Assuming that any solution can exchange objects with the current optimal solution through a wormhole, the corresponding solution update formula is as follows:

[0043]

[0044] Where: b u,j , b l,j They represent the maximum and minimum values ​​of the jth variable, i.e., the upper and lower limits of the support parameter; the range of random numbers rand2, rand3 and rand4 is [0,1]; P WEP It represents the probability of the existence of a wormhole model in the search space, that is, the probability of combining different support methods in the optimization. As the iteration proceeds, P WEP Increase linearly to gradually improve the algorithm's development capabilities, thereby preventing the algorithm from failing to converge; r TDR represents the travel distance rate, which represents the refined search near the high-quality solution and gradually decreases during the iteration process in order to perform a more accurate local search around the optimal solution obtained; P WEP and r TDR The update formula is as follows:

[0045]

[0046] Where: and Respectively represent P WEP The minimum and maximum values ​​of t cu Indicates the current iteration number, t max is the total number of iterations; P e It is the exploitation degree when the algorithm is running. The larger its value is, the higher the exploitation degree of the algorithm is, that is, the more accurate the iteration result is.

[0047] (3) If the maximum number of iterations t is satisfied max, then output the Pareto solution. If it is not satisfied, return to step (1).

[0048] Among them, the specific steps of combining entropy weight-TOPSIS to determine the optimal support parameters are as follows:

[0049] (1) According to the Pareto solution set obtained by the MOMVO algorithm, the decision matrix F = [f ij ], where f ij is the jth objective function value corresponding to the i-th Pareto solution, which is standardized as:

[0050]

[0051] (2) Determine the negative ideal solution Y - With positive ideal solution Y + :

[0052]

[0053] (3) Based on the entropy weight method, calculate the weight ω of each optimization objective:

[0054]

[0055] Where: H j is the entropy value of the jth optimization target; I is the number of Pareto solution sets; d j is the information entropy redundancy of the jth optimization target; ω j is the weight value of the jth optimization target; M is the number of optimization targets.

[0056] (4) Calculate the weighted Euclidean distance between each solution and the positive and negative ideal solutions

[0057]

[0058] (5) Calculate the ranking score Z of the i-th solution i :

[0059]

[0060] (6) According to Z i The plans are sorted according to the scores, and the one with the highest score is selected as the optimal support plan.

[0061] Example 2, based on Example 1, takes a tunnel project as an example to further illustrate the implementation steps of the present invention. A tunnel is of Class IV surrounding rock, and the preliminary support plan is as follows: the side arch is sprayed with 20cm thick C20 coarse fiber concrete; φ25 hollow grouting anchor rods are used, with an anchor rod length of 6m, annular spacing of 1.5m, and longitudinal spacing of 1.5m; the tunnel section is provided with I20a steel supports with a spacing of 1.1m.

[0062] 1. Establish a three-dimensional model for numerical analysis of tunnel excavation support for numerical simulation. Establish a proxy model for arch settlement G(x), set multiple constraints according to the layout of each support parameter, and establish a dual-objective optimization model for the soft rock support structure of the water diversion tunnel in combination with the support material cost J(x). The parameter value range is as follows:

[0063] min F(x)=[G(x),J(x)] T ,x=[s a ,s b ,s c ,s d ] T ∈R 4

[0064]

[0065] 2. The orthogonal experimental method and numerical simulation method are used to generate a sample data set; the support vector machine (SVM) supervised learning method is used to train the sample data, in which the algorithm randomly selects 20 groups of training set samples (accounting for 80%) and 5 groups of test set samples (accounting for 20%), and identifies the highly nonlinear relationship between tunnel support parameters and vault settlement, and obtains the proxy model G(x) = SVM(x) of vault settlement G(x).

[0066] The cost of establishing support materials J(x) is as follows:

[0067]

[0068] Taking the market price in a certain period of time as an example, the unit price of C20 concrete is 295 yuan / m 3 The unit price of I20a I-beam is 4,000 yuan / ton, and the unit price of structural seamless steel pipe used in anchor rods is 4,300 yuan / ton.

[0069] 3. A multi-objective optimization algorithm is used to solve the above dual-objective optimization model and obtain a set of Pareto solutions of support parameter schemes.

[0070] 4. Combined entropy weight - TOPSIS determines the optimal support parameters and obtains the optimal support plan.

[0071] In order to analyze whether the arch settlement meets the safety requirements, in accordance with the provisions of Article 9.2.8 of the "Technical Specifications for Geotechnical Anchors and Shotcrete Support Engineering" GB50086-2015, combined with the engineering geological conditions, this paper takes the relative convergence value of the tunnel as 1.2%, and calculates that the maximum allowable settlement of the tunnel is 118.3mm.

[0072] The comparison results of the support schemes before and after optimization are shown in Table 1 and Table 2. After optimization, the cost of support materials per linear meter was reduced from 5536.3 yuan to 5084.1 yuan, that is, 452 yuan was saved per linear meter, a decrease of 8.16%. The settlement of the tunnel vault was 92.3 mm, which only increased by 0.7 mm and was far less than the maximum allowable settlement of 118.3 mm, with sufficient safety margin. It can be seen that the support cost per linear meter can be greatly reduced under the premise of ensuring the safety of the tunnel surrounding rock, which shows that the optimization method of the present invention is completely feasible.

[0073] Table 1 Comparison of support parameters before and after optimization

[0074]

[0075] Table 2 Comparison of objective functions before and after optimization

[0076]

[0077] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A multi-objective intelligent optimization method for soft rock support structure of a water diversion tunnel, characterized by: The method comprises the following steps: A three-dimensional model for numerical analysis of tunnel excavation support was established for numerical simulation, and the orthogonal test method was used to generate sample data sets; The sample data was trained using the support vector machine (SVM) supervised learning method to identify the highly nonlinear relationship between tunnel support parameters and arch settlement. A proxy model of the arch settlement G(x) is established, that is, G(x) = SVM(x), which can reflect the safety of tunnel excavation support.

2. A multi-objective intelligent optimization method for soft rock support structure of a water diversion tunnel according to claim 1, characterized in that: The method further comprises: The support material cost model J(x) is established to evaluate and improve the economic efficiency of tunnel support, where: J(x)=J1(x)+J2(x)+J3(x) In the formula: J1(x) represents the cost of steel usage; J2(x) represents the cost of concrete usage; J3(x) represents the cost of other materials usage; x represents different support parameters.

3. The multi-objective intelligent optimization method for soft rock support structure of a water diversion tunnel according to claim 1 is characterized by: The method further comprises: A dual-objective optimization model for the soft rock support structure of the water diversion tunnel is established. This model is a multi-objective optimization model that is very helpful in improving the safety and economy of tunnel support at the same time. The specific dual-objective optimization model and support parameter value range of the soft rock support structure of the water diversion tunnel are as follows: minF(x)=[G(x),J(x)] T ,x=[s a ,s b ,s c ,s d ] T ∈R 4 Where G(x) is the established vault settlement proxy model, J(x) is the cost of support materials, s a is the longitudinal spacing of anchor rods (m), s b is the circumferential spacing of anchor rods (m), s c is the steel support spacing (m), s d is the thickness of the concrete spray layer (cm).

4. The multi-objective intelligent optimization method for soft rock support structure of a water diversion tunnel according to claim 1 is characterized by: The method adopts the multi-objective multiverse algorithm MOMVO to solve the established multi-objective optimization model that takes into account the vault settlement G(x) and the support material cost J(x), and obtains the Pareto solution set; the entropy weight method is used to determine the weight of each objective, and then the optimal support scheme is finally decided based on the TOPSIS comprehensive ranking comparison.

Images