A railway line selection method, medium and equipment considering the risk of active fault zone

Abstract

The present invention relates to the technical field of railway line selection and design, and discloses a railway line selection method, medium and equipment that considers the risk of active fault zones. The method comprises: S1: establishing a comprehensive geographic information model of the study area; S2: establishing a mathematical optimization model based on the comprehensive geographic information model, which is used to find design variables that meet the constraints and make the objective function optimal; the mathematical optimization model comprises a basic optimization model and a distributed Bruce rod optimization model; S3: solving the mathematical optimization model using a distributed Bruce rod particle swarm optimization algorithm, and outputting the optimal line plan. The present invention considers the uncertainty risk of active fault zones, constructs an intelligent line selection optimization model that considers the risk of active fault zones; and generates the final line plan by solving the distributed Bruce rod particle swarm optimization algorithm, which can provide auxiliary design for railway line selection in strong earthquake areas, and has the characteristics of high automation, strong practicality and high operating efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of railway route selection and design, and in particular to a railway route selection method, medium and equipment taking into account the risk of an active fault zone. Background Art

[0002] Railway line selection and design plays a key role in railway construction and maintenance. In traditional line design, line selection mainly relies on manual experience. Especially under complex geological and topographic conditions, engineers often have to consider many control factors and other constraints. The whole process requires a lot of human and material resources, and some valuable solutions may be missed. In order to solve the above problems encountered in manual line selection, with the development of computer technology, intelligent line selection technology has emerged. The existing intelligent line selection technology collects topographic and geological data, establishes a comprehensive geographic information model, and then constructs a mathematical optimization model that describes the actual line selection engineering problem. Then, the distance transformation algorithm, particle swarm algorithm, genetic algorithm and other methods are used to solve the optimization model and intelligently search for line solutions. Compared with manual line selection, intelligent line selection can greatly improve the efficiency and quality of line selection.

[0003] However, although the existing intelligent line selection technology can generate optimal line plans under flat terrain conditions, when selecting railway lines in earthquake-affected areas, there is still a gap in the research on intelligent line selection that considers the risks of active fault zones caused by earthquakes. Especially in complex and dangerous mountainous areas, fault zones are often numerous, densely distributed, and extremely long, and line plans usually cannot completely avoid all active fault zones. Therefore, once the line crosses a specific active fault zone, the risks of railway construction and operation caused by the fault zone misalignment must be considered. In addition, since railway line selection is an iterative and refined dynamic design process, and there are huge uncertainties in the misalignment position and displacement of the fault zone, the research on intelligent line selection that considers the risks of fault zones in earthquake-affected areas is very complex and faces a lot of challenges. In this regard, the existing research results on intelligent line selection are often powerless.

[0004] Therefore, there is an urgent need for a railway route selection method that takes into account the risks of active fault zones to solve the above-mentioned problems existing in railway route selection in earthquake-affected areas. Summary of the invention

[0005] The purpose of the present invention is to provide a railway line selection method considering the risk of active fault zones, and its specific technical scheme is as follows:

[0006] A railway line selection method considering the risk of active fault zones comprises the following steps:

[0007] S1: Establish a comprehensive geographic information model of the study area;

[0008] S2: A mathematical optimization model is established based on the comprehensive geographic information model to find the design variables that optimize the objective function under the constraints;

[0009] The mathematical optimization model includes a basic optimization model and a robust optimization model. The establishment of the robust optimization model includes selecting uncertain variables, establishing a robust optimization objective function, and proposing a regret opportunity constraint.

[0010] S3: using the distributed robust particle swarm optimization algorithm to solve the mathematical optimization model and output the optimal route plan; the distributed robust particle swarm optimization algorithm includes three stages: route plan generation, route plan evaluation and route plan iterative evolution.

[0011] Preferably, the establishment of the basic optimization model in S2 comprises the following steps:

[0012] Select design variables: Describe the railway line through the plane intersection HPI and the slope change point VPI, and use the circular curve radius of each plane intersection The intersection coordinates (X, Y), as well as the pile mileage P and design elevation E of each slope change point constitute the design variables for determining the three-dimensional line shape;

[0013] Determine basic constraints including geometric constraints, structural constraints and area constraints;

[0014] Establish annual construction build costs The basic objective function is as follows:

[0015] ;

[0016] in, is the investment benefit coefficient; is the construction cost of the bridge, is the construction cost of the roadbed section, is the construction cost of the tunnel, For land acquisition costs, is the cost of the structure related to its length.

[0017] Preferably, the uncertain variables selected in S2 are:

[0018] Consider the influence of the main shock fault zone, branch fault zones and secondary fault zones existing at the time of the earthquake, assuming that the fault zones farther from the main shock fault zone than the secondary fault zones are inactive during the earthquake;

[0019] The fracture at the intersection of the fault zone and the line is simplified as a rigid fracture;

[0020] Based on simplification and through line reconstruction analysis, the uncertain deformation range is converted into the line shape automatic reconstruction part;

[0021] Get the uncertainty variable, which includes the number of the main shock fault zone , the impact range of the main shock fault zone , the impact range of branch fault zones , the impact range of secondary fault zones , the parallel displacement of each active fault zone , the normal displacement of each active fault zone and the longitudinal displacement of each active fault zone ;

[0022] Representing uncertain variables as sets of variables .

[0023] Preferably, establishing the robust optimization objective function in S2 comprises the following steps:

[0024] For a given railway line, if The first fault zone is the main shock fault zone. , then Expected losses due to fault zones The calculation formula is as follows:

[0025] ;

[0026] in: represents the expected value, Indicates the number of potential destruction scenarios; To rebuild the Unit cost of a structure; It is The length of the reconstruction segment of each structure; is the expected average daily revenue of the railway; It is The repair time of each structure; Indicates The total number of structures in the fault zone; Indicates The fault zone Number of structures;

[0027] Through Monte Carlo simulation, a series of damage scenarios are simulated according to the range of uncertain variables to calculate the expected risk loss. The maximum expected risk loss is selected from the active fault zones, and combined with the basic objective function, the robust optimization objective function is obtained. for:

[0028] .

[0029] Preferably, the regret opportunity constraint in S2 includes:

[0030] Assume that the minimum cost solution obtained by the solution determined by the basic optimization model is , then other alternatives generated by the distributed robust optimization model ,and , the chance constraint should be satisfied It is expressed as follows:

[0031] ;

[0032] in, represents the confidence level set by the designer; Calculate the regret between the two options, the expression is as follows:

[0033] ;

[0034] ;

[0035] ;

[0036] in, is used to control the conservative level of the distributed robust optimization model index; and The solutions are and solutions The basic objective function of

[0037] By changing and The value of adjusts the conservativeness of the distributed robust optimization model. The decrease in value and As the value of increases, the distributed robust optimization model becomes more and more conservative; if =0 and =100%, the robust optimization model degenerates into the basic optimization model.

[0038] Preferably, the generation of the line plan includes the following steps:

[0039] Abstract each circuit into particles described by design variables;

[0040] According to the route selection and design habits in actual projects, the step-by-step strategy of "first plane-then longitudinal-then integration" is adopted to determine the plane and longitudinal routes under the premise of meeting all design constraints;

[0041] Repeat the above steps to generate a particle swarm consisting of a group of line individuals.

[0042] Preferably, the evaluation of the line scheme comprises the following steps:

[0043] Step 1: Number all fault zones that intersect the railway line in order. ,set up , ;

[0044] Step 2: If , the program is terminated; otherwise, assuming that The fault zone is taken as the main earthquake fault zone, and it is taken as the center line of the main earthquake fault zone area to obtain the empirical maximum value of the influence range of the main earthquake fault zone, branch fault zone and secondary fault zone, and random sampling is carried out within the empirical maximum value range;

[0045] Step 3: Randomly generate the three-dimensional displacement of the main shock fault zone, branch fault zone and secondary fault zone, denoted as , and , the expression is as follows:

[0046] ;

[0047] ;

[0048] ;

[0049] in, , and are respectively the maximum possible displacements corresponding to different fault zones obtained through geological surveys; the three-dimensional displacements include horizontal, normal and vertical displacements;

[0050] Step 4: Reconstruct the local 3D line section damaged by the deformation of the fault zone, including three scenarios:

[0051] The first type: when the intersection of the fault zone and the line is in a straight line segment, configure an S-curved line segment, as follows;

[0052] The straight line between two adjacent curves is set to the minimum threshold , curve radius and deflection angle It is the main variable to determine the S-bend line, the shortest reconstruction length The calculation formula is as follows:

[0053] ;

[0054] in, is the length of the circular curve, not less than the minimum circular curve length ; is the length of the transition curve, given by Sure; Not less than the minimum curve radius ;

[0055] The following equation is derived:

[0056] ;

[0057] ;

[0058] ;

[0059] ;

[0060] ;

[0061] in, is the displacement of the fault zone;

[0062] By traversing and The possible value of obtains the shortest reconstruction length ;

[0063] The second method: when the intersection of the fault zone and the line is in a circular curve segment, adjust the radius of the circular curve and the length of the transition curve to reconnect the two tangent lines before and after the circular curve; or redesign the clip line to connect the two adjacent plane circular curves at the intersection of the fault zone and the curve to determine the shortest curve segment reconstruction length;

[0064] The third method: when the intersection of the fault zone and the line is in the transition curve section, the transition curve is reconstructed similarly to the reconstruction method of the two circular curves in the second method; or the length of the transition curve is extended or shortened to achieve the reconstruction transition between the straight line and the circular curve;

[0065] After the plane alignment is reconstructed, its corresponding longitudinal section is checked and realized by adding a new slope section at the location of the rigid fracture; the final line reconstruction length is the maximum value between the plane and longitudinal line reconstruction lengths;

[0066] Step 5: Consider the constraints, calculate the basic cost, and calculate the reconstruction risk cost based on the reconstructed line;

[0067] Step 6: Command ,if , then The fracture zone Monte Carlo simulations have been completed; find The expected value of the risk cost of the second simulation reconstruction is When the fault zone is the main earthquake fault zone, the reconstruction risk cost of the line plan is , , execute steps 2 to 5, and then The Monte Carlo simulation is performed when the fault zone is the main shock fault zone; otherwise, The fracture zone If the Monte Carlo simulation has not been completed, proceed to the next Monte Carlo simulation and directly execute steps 2 to 5.

[0068] Preferably, the iterative evolution of the line scheme comprises the following steps:

[0069] Build an external archive to store the base optimization model and the distributed robust optimization model Particle, basic optimization model The particle is recorded as , the distribution of the robust optimization model The particle is recorded as ;

[0070] Iteration is based on the following two functions:

[0071] ;

[0072] ;

[0073] in, and They are The displacement and velocity vectors of each individual, , is the size of the group; Indicates the current iteration number; is the inertia weight; and They are cognitive and social learning factors; and is a parameter randomly distributed between 0 and 1; represents the best particle solution found among all individuals in the population during the iteration process; Indicates the iteration period The best solution recorded by each particle;

[0074] After each iteration, the regret chance constraint is compared and ,if If this constraint is satisfied, then for Otherwise, eliminate it and take for ; Use regret opportunity constraint to select each individual ;

[0075] When checking the regret opportunity constraint, The value of changes as the search progresses, expressed as:

[0076] ;

[0077] in, is the current iteration number of the particle swarm, is the maximum number of iterations of the particle swarm; and They are The maximum and minimum values ​​of value to find a more robust solution and explore the search space more thoroughly; as the number of PSO iterations increases, the algorithm gradually converges. The value gradually decreases, focusing on improving the quality of the solution.

[0078] The application of the technical solution of the present invention has the following beneficial effects:

[0079] A railway line selection method considering the risk of active fault zones comprises the following steps: S1: establishing a comprehensive geographic information model of a study area; S2: establishing a mathematical optimization model based on the comprehensive geographic information model, for finding design variables that optimize an objective function under constraints; the mathematical optimization model comprises a basic optimization model and a robust optimization model, the establishment of which comprises selecting uncertain variables, establishing a robust optimization objective function, and proposing a regret opportunity constraint; S3: solving the mathematical optimization model using a robust particle swarm optimization algorithm, and outputting an optimal line plan; the robust particle swarm optimization algorithm comprises three stages: generation of a line plan, evaluation of a line plan, and iterative evolution of a line plan. The present invention aims at the risk problem caused by active fault zones in the process of railway line selection in earthquake-affected areas, takes into account the uncertainty risk of active fault zones, and then constructs an intelligent line selection optimization model taking into account the risk of active fault zones; generates an initial line plan based on a particle swarm algorithm, performs Monte Carlo simulation on each line plan in turn, selects individual optimal particles and group optimal particles through regret opportunity constraints, and updates the group according to the particle swarm algorithm rules; finally, iterates the group until the group converges or reaches the maximum number of iterations, and generates a final line plan. The method can provide auxiliary design for artificial intelligence, solves the problem of intelligent line selection in active fault zones in earthquake-affected areas, and has the characteristics of high automation, strong practicality, and high operating efficiency.

[0080] The present invention also provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method for selecting a railway line in consideration of the risk of an active fault zone is implemented.

[0081] The present invention also provides an electronic device, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to execute the aforementioned railway line selection method considering the risk of active fault zones by executing the executable instructions.

[0082] In addition to the above-described purposes, features and advantages, the present invention has other purposes, features and advantages. The present invention will be further described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0084] Figure 1 A schematic flow chart of a railway line selection method considering the risk of active fault zones according to the present invention;

[0085] Figure 2 This is a schematic diagram of a "rigid fracture" assumption at the intersection of the fracture zone and the line in the embodiment;

[0086] Figure 3 A schematic diagram of an intelligent line selection process based on a PSO algorithm in an embodiment;

[0087] Figure 4 It is a schematic diagram of the reconstruction of the local line plan after the fault zone is misaligned in the embodiment. DETAILED DESCRIPTION

[0088] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0089] See also Figure 1 In one embodiment, a railway line selection method considering the risk of active fault zones includes the following steps:

[0090] S1: Establish a comprehensive geographic information model of the study area; this step is to collect information such as topography and geology of the study area based on the existing intelligent line selection technology to establish a comprehensive geographic information model.

[0091] S2: A mathematical optimization model is established based on the comprehensive geographic information model to find the design variables that optimize the objective function under the constraints. The three elements of the model are design variables, objective functions and constraints. Considering the risks caused by active fault zones during railway line selection in earthquake-affected areas, this embodiment will establish a basic optimization model and a distributed robust optimization model.

[0092] The establishment of the distributed robust optimization model includes selecting uncertain variables, establishing a robust optimization objective function, and proposing regret opportunity constraints;

[0093] S2.1: The establishment of the basic optimization model includes the following steps:

[0094] S2.1.1: Selection of design variables: Describe the railway line through the plane intersection HPI and the slope change point VPI, using the circular curve radius of each plane intersection The intersection coordinates (X, Y), as well as the pile mileage P and design elevation E of each slope change point constitute the design variables for determining the three-dimensional line shape;

[0095] S2.1.2: Determine the basic constraints including geometric constraints, structural constraints and area constraints; various geometric constraints include minimum curve radius, maximum slope, etc.; various structural constraints include maximum bridge height and maximum tunnel length, etc.; various area constraints include restricted areas and environmentally sensitive areas, etc.

[0096] S2.1.3: Establish annual construction costs The basic objective function is as follows:

[0097] ;

[0098] in, is the investment benefit coefficient, usually taken as 0.065; is the construction cost of the bridge, is the construction cost of the roadbed section, is the construction cost of the tunnel, For land acquisition costs, It is the cost of length-related structures (such as tracks and retaining walls). The specific calculation method of the above engineering costs is prior art.

[0099] S2.2: Establish a distributionally robust optimization (DRO) model.

[0100] S2.2.1: Select the uncertain variables, specifically:

[0101] In addition to the design variables in S2.1, several types of uncertainties should be considered to assess the risk of active fault zones in railway line selection design: When an earthquake occurs, there are generally three types of active fault zones, namely, the main earthquake fault zone (main earthquake fault zone) that is closest to the epicenter and has the largest displacement, the branch fault zone (branch fault zone) that may be affected by the main earthquake fault zone, and the secondary fault zone (secondary fault zone) that is farther away from the main earthquake fault zone and has the smallest displacement. Considering the influence of the main earthquake fault zone, branch fault zone and secondary fault zone that exist when an earthquake occurs, other fault zones that are farther away from the main earthquake fault zone than the secondary fault zone are assumed to be inactive during the earthquake.

[0102] Generally speaking, when an earthquake occurs, the influence range of the main earthquake fault zone, branch fault zone and secondary fault zone is uncertain, and the fault displacement values ​​of the main earthquake fault zone, branch fault zone and secondary fault zone are also uncertain. In addition, when the railway line passes through the active fault zone, the deformation of the railway line at the intersection of the line and the fault zone is continuous and irregular, and is related to many indicators such as earthquake intensity and topography, which is difficult to calculate accurately. Therefore, this embodiment simplifies the fracture at the intersection of the fault zone and the line into a rigid fracture (see Figure 2 ), based on this simplification, the uncertain deformation range can be converted into the automatic reconstruction part of the line shape through line reconstruction analysis.

[0103] In summary, the uncertainty variables are obtained, which include the number of the main shock fault zone , the impact range of the main shock fault zone , the impact range of branch fault zones , the impact range of secondary fault zones , the parallel displacement of each active fault zone , the normal displacement of each active fault zone and the longitudinal displacement of each active fault zone ;

[0104] Representing uncertain variables as sets of variables , which will be searched in the robust optimization process S3.2 in combination with the design variables in S2.1.1.

[0105] S2.2.2: Establish a robust optimization objective function, including the following steps:

[0106] In order to solve the uncertainty problem mentioned in S2.2.1, this embodiment adopts a data-driven DRO objective function. For a given railway line, since each fault zone intersecting with it may become a main shock fault zone, if The first fault zone is the main shock fault zone. , then Expected losses due to fault zones The calculation formula is as follows:

[0107] ;

[0108] in, represents the expected value, Indicates the number of potential destruction scenarios; To rebuild the Unit cost of a structure; It is The length of the reconstruction segment of each structure; is the expected average daily revenue of the railway; It is The repair time of each structure; Indicates The total number of structures in the fault zone; Indicates The fault zone Number of structures;

[0109] Through Monte Carlo simulation, a series of damage scenarios are simulated according to the range of uncertain variables to calculate the expected risk loss. Select the maximum expected risk loss from the active fault zones, and combine it with the basic objective function in S2.1 to obtain the robust optimization objective function for:

[0110] .

[0111] S2.2.3: Propose degree of regret (DoR) opportunity constraints, including:

[0112] Due to the natural conservatism of the distributed robust optimization model, the designer may regret rejecting the solution determined by the basic optimization model and choosing the distributed robust optimization model solution. Assume that the minimum cost solution obtained by the solution determined by the basic optimization model is , then other alternatives generated by the distributed robust optimization model ,and , the chance constraint should be satisfied It is expressed as follows:

[0113] ;

[0114] in, represents the confidence level set by the designer, which is 75% in this embodiment; Calculate the regret between the two options, the expression is as follows:

[0115] ;

[0116] ;

[0117] ;

[0118] in, is used to control the conservative level of the distributed robust optimization model index; and The solutions are and solutions The basic objective function of

[0119] By changing and The value of adjusts the conservativeness of the distributed robust optimization model. The decrease in value and As the value of increases, the distributed robust optimization model becomes more and more conservative; if =0 and =100%, the robust optimization model degenerates into the basic optimization model.

[0120] In order to solve the optimization model proposed in S2, this embodiment adds multiple uncertain optimization modules based on the existing particle swarm optimization (PSO) algorithm and proposes a distributed robust particle swarm optimization (DRO-PSO) algorithm.

[0121] S3: Solve the mathematical optimization model using the distributed robust particle swarm optimization algorithm and output the optimal route plan; the distributed robust particle swarm optimization algorithm includes three stages: route plan generation, route plan evaluation, and route plan iterative evolution (see Figure 3 ).

[0122] S3.1: Generation of the line plan. At this stage, the distributed robust particle swarm optimization algorithm basically adopts the existing PSO algorithm in the field of intelligent line selection, including the following steps:

[0123] Abstract each circuit into particles described by design variables;

[0124] According to the route selection and design habits in actual projects, the step-by-step strategy of "first plane-then longitudinal-then integration" is adopted to determine the plane and longitudinal routes under the premise of meeting all design constraints;

[0125] Repeat the above steps to generate a particle swarm consisting of a group of line individuals.

[0126] S3.2: Evaluation of the line scheme. The line scheme evaluation needs to consider the various uncertainties discussed in S2.2. For this purpose, a line reconstruction method based on Monte Carlo simulation (MCS) is designed, which includes the following steps:

[0127] Step 1: Number all fault zones that intersect the railway line in order. ,set up , ;

[0128] Step 2: If , the program is terminated; otherwise, assuming that The fault zone is taken as the main earthquake fault zone, and it is taken as the center line of the main earthquake fault zone area to obtain the empirical maximum value of the influence range of the main earthquake fault zone, branch fault zone and secondary fault zone, and random sampling is carried out within the empirical maximum value range;

[0129] Step 3: Randomly generate the three-dimensional displacement of the main shock fault zone, branch fault zone and secondary fault zone, denoted as , and , the expression is as follows:

[0130] ;

[0131] ;

[0132] ;

[0133] in, , and are respectively the maximum possible displacements corresponding to different fault zones obtained through geological surveys; the three-dimensional displacements include horizontal, normal and vertical displacements;

[0134] Step 4: Reconstruct the local three-dimensional line section damaged by the deformation of the fault zone. Under the assumptions in S2.2.1, there are three types of reconstructed plane lines (see Figure 4 ), specifically:

[0135] The first type: when the intersection of the fault zone and the line is in a straight line segment, see Figure 4 In part (a), configure the S-curved line segment as follows;

[0136] The straight line between two adjacent curves is set to the minimum threshold , curve radius and deflection angle It is the main variable to determine the S-bend line, the shortest reconstruction length The calculation formula is as follows:

[0137] ;

[0138] in, is the length of the circular curve, not less than the minimum circular curve length ; is the length of the transition curve, given by Sure; Not less than the minimum curve radius ;

[0139] pass Figure 4 From part (a), we can derive the following equation:

[0140] ;

[0141] ;

[0142] ;

[0143] ;

[0144] ;

[0145] in, is the displacement of the fault zone;

[0146] Combined with the above formula, by traversing and The possible value of obtains the shortest reconstruction length ;

[0147] The second type: When the intersection of the fault zone and the line is in a circular curve segment, see Figure 4 As shown in part (b), the radius of the circular curve and the length of the transition curve can be adjusted to reconnect the two tangent lines before and after the circular curve; or the clip line can be redesigned to connect the two adjacent plane circular curves at the intersection of the fault zone and the curve, see Figure 4 Part (c) is used to determine the shortest curve segment reconstruction length; the principles of these two methods are relatively simple, so their detailed geometric calculations are omitted here. In the optimization process, both methods will be tried to determine the shortest curve segment reconstruction length;

[0148] The third method: when the intersection of the fault zone and the line is in the transition curve section, the transition curve is reconstructed similarly to the reconstruction method of the two circular curves in the second method; or the length of the transition curve is extended or shortened to achieve the reconstruction transition between the straight line and the circular curve;

[0149] After the plane alignment is reconstructed, check its corresponding longitudinal section, such as Figure 4 As shown in part (d), this is achieved by adding a new slope section at the location of the rigid fracture; the final line reconstruction length is the maximum value between the plane and longitudinal section line reconstruction lengths;

[0150] Step 5: Consider the constraints, calculate the basic cost, and calculate the reconstruction risk cost based on the reconstructed line;

[0151] Step 6: Command ,if , then The fracture zone Monte Carlo simulations have been completed; find The expected value of the risk cost of the second simulation reconstruction is When the fault zone is the main earthquake fault zone, the reconstruction risk cost of the line plan is , , execute steps 2 to 5, and then The Monte Carlo simulation is performed when the fault zone is the main shock fault zone; otherwise, The fracture zone If the Monte Carlo simulation has not been completed, proceed to the next Monte Carlo simulation and directly execute steps 2 to 5.

[0152] In this embodiment, the number of fracture zones =13, number of Monte Carlo simulations =100.

[0153] S3.3: Iterative evolution of the line scheme includes the following steps:

[0154] Build an external archive to store the base optimization model and the distributed robust optimization model Particle, basic optimization model The particle is recorded as , the distribution of the robust optimization model The particle is recorded as ;

[0155] Iteration is based on the following two functions, both of which are mature PSO algorithm formulas:

[0156] ;

[0157] ;

[0158] in, and They are The displacement and velocity vectors of each individual, , is the size of the group; Indicates the current iteration number; is the inertia weight; and They are cognitive and social learning factors; and is a parameter randomly distributed between 0 and 1; represents the best particle solution found among all individuals in the population during the iteration process; Indicates the iteration period The best solution recorded by each particle;

[0159] After each iteration, the regret chance constraint is compared and ,if If this constraint is satisfied, then for Otherwise, eliminate it and take for ; Use regret opportunity constraint to select each individual ;

[0160] When checking the regret opportunity constraint, The value of changes as the search progresses, expressed as:

[0161] ;

[0162] in, is the current iteration number of the particle swarm; is the maximum number of iterations of the particle swarm, which is 200 in this embodiment; and They are The maximum and minimum values ​​of are 1.0 and 0.6 respectively in this embodiment; in the initial stage of the optimization algorithm, a larger value to find a more robust solution and explore the search space more thoroughly; as the number of PSO iterations increases, the algorithm gradually converges. The value gradually decreases, focusing on improving the quality of the solution.

[0163] This embodiment further discloses a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the railway line selection method considering the risk of active fault zones described above in this embodiment is implemented.

[0164] This embodiment also discloses an electronic device, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to execute the railway route selection method considering the risk of active fault zones as described above in this embodiment by executing the executable instructions.

[0165] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A railway line selection method considering the risk of active fault zones, characterized in that: The steps include: S1: Establish a comprehensive geographic information model of the study area; S2: A mathematical optimization model is established based on the comprehensive geographic information model to find the design variables that optimize the objective function under the constraints; The mathematical optimization model includes a basic optimization model and a robust optimization model. The establishment of the robust optimization model includes selecting uncertain variables, establishing a robust optimization objective function, and proposing a regret opportunity constraint. S3: using the distributed blue bar particle swarm optimization algorithm to solve the mathematical optimization model and output the optimal line plan; the distributed blue bar particle swarm optimization algorithm includes three stages: line plan generation, line plan evaluation and line plan iterative evolution; The establishment of the robust optimization objective function in S2 comprises the following steps: For a given railway line, if the nth fault zone is the main shock fault zone, n = 1, ..., N, then the expected loss F caused by the nth fault zone is L,n The calculation formula is as follows: ； in, represents the expected value, K represents the number of potential damage scenarios; To reconstruct the mth n,k Unit cost of a structure; It is the mth n,k The length of the reconstruction segment of each structure; v R is the expected average daily revenue of the railway; It is the mth n,k The repair time of a structure; M n,k represents the total number of structures in the nth fault zone; m n,k The mth fault zone of the nth fault zone n,k Number of structures; Through Monte Carlo simulation, a series of damage scenarios are simulated according to the value range of uncertain variables to calculate the expected risk loss. The maximum expected risk loss is selected among N active fault zones, and combined with the basic objective function, the robust optimization objective function minF is obtained. DRO for: 。 2. A railway line selection method considering the risk of active fault zones according to claim 1, characterized in that: The establishment of the basic optimization model in S2 includes the following steps: Select design variables: Describe the railway line through the plane intersection HPI and the slope change point VPI, and use the circular curve radius R of each plane intersection H The intersection coordinates (X, Y), as well as the pile mileage P and design elevation E of each slope change point constitute the design variables for determining the three-dimensional line shape; Determine basic constraints including geometric constraints, structural constraints and area constraints; Establish annual construction cost F R The basic objective function is as follows: minF R (HPI,VPI)=min[Δ·(C B +C T +C E +C R +C L )]; Among them, Δ is the investment benefit coefficient; C B is the construction cost of the bridge, C T is the construction cost of the roadbed section, C E is the construction cost of the tunnel, C R is the land acquisition cost, C L is the cost of the structure related to its length.

3. A railway line selection method considering the risk of active fault zones according to claim 2, characterized in that: The uncertain variables selected in S2 are specifically: Consider the influence of the main shock fault zone, branch fault zones and secondary fault zones existing when the earthquake occurred, assuming that the fault zones farther from the main shock fault zone than the secondary fault zones are inactive during the earthquake; The fracture at the intersection of the fault zone and the line is simplified as a rigid fracture; Based on simplification and through line reconstruction analysis, the uncertain deformation range is converted into the line shape automatic reconstruction part; The uncertainty variables include the number N of the main shock fault zone, the influence range A of the main shock fault zone, and the M 、The influence range of branch fault zone A B 、The impact range of secondary fault zone A S , the parallel displacement D of each active fault zone P , the normal displacement D of each active fault zone N and the longitudinal displacement D of each active fault zone V ; Denote the uncertain variables as a variable set Ψ.

4. A railway line selection method considering the risk of active fault zones according to claim 3, characterized in that: The regret opportunity constraints in S2 include: Assuming that the minimum cost solution determined by the basic optimization model is α, the other alternative solutions β generated by the distributed robust optimization model, and β≠α, satisfying the chance constraint P(α,β) are expressed as follows: ； Where λ represents the confidence level set by the designer; f DoR (α, β) calculates the regret between the two options, and its expression is as follows: Where ε is the DoR index used to control the conservative level of the distributed robust optimization model; F R,α and F R,β are the basic objective functions of scheme α and scheme β respectively; The conservatism of the distributed robust optimization model is adjusted by changing the values ​​of λ and ε. As the value of λ decreases and the value of ε increases, the distributed robust optimization model becomes more and more conservative. If ε=0 and λ=100%, the distributed robust optimization model degenerates into the basic optimization model.

5. A railway line selection method considering the risk of active fault zones according to claim 4, characterized in that: The generation of the line plan includes the following steps: Abstract each circuit into particles described by design variables; According to the route selection and design habits in actual projects, and on the premise of meeting all design constraints, the step-by-step strategy of "first plane-then longitudinal-then integration" is adopted to determine the plane and longitudinal routes; Repeat the above steps to generate a particle swarm consisting of a group of line individuals.

6. A railway line selection method considering the risk of active fault zones according to claim 5, characterized in that: The evaluation of the line plan includes the following steps: Step 1: Number all fault zones intersecting the railway line from 1 to N in sequence, assuming n = 1, k = 1; Step 2: If n>N, terminate the program; otherwise, assume that the nth fault zone is the main shock fault zone, take it as the center line of the main shock fault zone area, obtain the empirical maximum value of the influence range of the main shock fault zone, branch fault zone and secondary fault zone, and perform random sampling within the empirical maximum value range; Step 3: Randomly generate the three-dimensional displacement of the main shock fault zone, branch fault zone and secondary fault zone, denoted as D MFZ , D BFZ and D SFZ , the expression is as follows: D MFZ =rand(0,D MFZmax ); D BFZ =rand[0,min(D BFZmax ,D MFZ )]; D SFZ =rand[0,min(D SFZmax ,D BFZ )]; Among them, D MFZmax , D BFZmax and D SFZmax are respectively the maximum possible displacements corresponding to different fault zones obtained through geological surveys; the three-dimensional displacements include horizontal, normal and vertical displacements; Step 4: Reconstruct the local 3D line section damaged by the deformation of the fault zone, including three scenarios: The first type: when the intersection of the fault zone and the line is in a straight line segment, configure an S-curved line segment, as follows; The straight line between two adjacent curves is set to the minimum threshold L Tmin , curve radius R Tmin and the deflection angle δ are the main variables for determining the S-bend line, and the shortest reconstruction length L R The calculation formula is as follows: minL R (R Tmin ,δ)=2·(l c +2·l0)+L Tmin ; Among them, l c is the length of the circular curve, not less than the minimum circular curve length L Cmin ; l0 is the length of the transition curve, which is given by R Tmin OK; R Tmin Not less than the minimum curve radius R min ; The following equation is derived: l c =R Tmin ·d·π / 180°-l0; (2·T+L Tmin )·sinδ=D FZ ; m t ＝l0 / 2； Among them, D FZ is the displacement of the fault zone; By traversing R Tmin and the possible values ​​of δ to obtain the shortest reconstruction length L R ; The second method: when the intersection of the fault zone and the line is in a circular curve segment, adjust the radius of the circular curve and the length of the transition curve to reconnect the two tangent lines before and after the circular curve; or redesign the clip line to connect the two adjacent plane circular curves at the intersection of the fault zone and the curve to determine the shortest curve segment reconstruction length; The third method: when the intersection of the fault zone and the line is in the transition curve section, the transition curve is reconstructed similarly to the reconstruction method of the two circular curves in the second method; or the length of the transition curve is extended or shortened to achieve the reconstruction transition between the straight line and the circular curve; After the plane alignment is reconstructed, its corresponding longitudinal section is checked and realized by adding a new slope section at the location of the rigid fracture; the final line reconstruction length is the maximum value between the plane and longitudinal line reconstruction lengths; Step 5: Consider the constraints, calculate the basic cost, and calculate the reconstruction risk cost based on the reconstructed line; Step 6. Let k=k+1. If k>K, the K-times Monte Carlo simulation of the n-th fault zone has been completed. Calculate the expected value of the risk cost of the K-times simulation reconstruction, which is the reconstruction risk cost of the line plan when the n-th fault zone is the main shock fault zone. Let n=n+1, k=1, execute steps 2 to 5, and perform the Monte Carlo simulation when the n+1-th fault zone is the main shock fault zone. Otherwise, the K-times Monte Carlo simulation of the n-th fault zone has not been completed, and continue with the next Monte Carlo simulation, directly executing steps 2 to 5.

7. A railway line selection method considering the risk of active fault zones according to claim 6, characterized in that: The iterative evolution of the line scheme includes the following steps: Construct an external archive to store the gBest particles of the basic optimization model and the distributed robust optimization model. The gBest particles of the basic optimization model are denoted as gBest. raw , the gBest particle of the blue bar optimization model is denoted as gBest DRO ; Iteration is based on the following two functions: D m (t+1)=D m (t)+V m (t+1); V m (t+1)=ω·V m (t)+C1·r1·[gBest(t)-D m (t)]+C2·r2·[pBest m (t)-D m (t)]; Among them, D m and V m are the displacement and velocity vectors of the mth individual, 1≤m≤M, M is the size of the group; t represents the current iteration number; ω is the inertia weight; C1 and C2 are cognitive and social learning factors, respectively; r1 and r2 are parameters randomly distributed between 0 and 1; gBest represents the best particle solution found among all individuals in the group during the iteration process; pBest m represents the best solution recorded by the mth particle during the iteration; After each iteration, gBest is compared by the regret chance constraint raw and gBest DRO , if gBest DRO If this constraint is satisfied, then gBest is chosen. DRO is gBest, otherwise, it is eliminated and gBest is selected raw is gBest; pBest of each individual is selected using regret opportunity constraint m ; When checking the regret chance constraint, the value of ε changes as the search proceeds, expressed as: e=e max -(e max -e min )·g / G; Among them, g is the current iteration number of the particle swarm, G is the maximum iteration number of the particle swarm; ε max and ε min are the maximum and minimum values ​​of ε respectively; in the initial stage of the optimization algorithm, a larger ε value is used to find a more robust solution and explore the search space more thoroughly; as the number of PSO iterations increases, the algorithm gradually converges and the ε value gradually decreases, thus focusing on improving the quality of the solution.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the railway route selection method considering the risk of active fault zones as described in any one of claims 1 to 7 is implemented.

9. An electronic device, characterized in that: include: processor; and a memory for storing executable instructions for the processor; Wherein, the processor is configured to execute the railway route selection method considering the risk of active fault zones as described in any one of claims 1 to 7 by executing the executable instructions.

Images