A line reconstruction optimization method, medium and device for a local strong constraint section

Abstract

The present invention relates to the technical field of railway reconstruction, and discloses a line reconstruction optimization method, medium and equipment for a local strong constraint section, the method comprising: S1: importing an initial plane plan; S2: performing constraint detection and generating a strong constraint section; S3: optimizing the initial plane plan; S4: using a strong constraint oscillation iteration method to detect the point-line consistency and constraint conditions of a reconstructed plane plan. The present invention generates a relatively optimal initial reconstruction plan for the plane line position of an existing railway based on the coordinates of the measuring points, using a dynamic threshold method and a MADS algorithm, and proposes a line default quantitative assessment model to formulate assessment standards and methods for two different types of constraints, which can accurately and comprehensively measure their default situations. The plane linear reconstruction is achieved through a "difficult first, easy later" fitting strategy, and finally a strong constraint oscillation iteration method is proposed to optimize the linear parameters, thereby ensuring the accuracy of local optimization.

Description

Technical Field

[0001] The present invention relates to the technical field of railway reconstruction, and in particular to a line reconstruction optimization method, medium and equipment for a local strong constraint section. Background Art

[0002] With the increasing demand for railway transportation, the plane alignment of existing railways will inevitably change in the long-term operation due to the impact between wheels and rails, foundation settlement deformation and many other factors. The smoothness of the line is difficult to maintain, which in turn affects the smoothness and safety of train operation and the comfort of passengers. At this time, in order to ensure the safe operation of trains and the comfort of passengers, the railway maintenance department needs to regularly measure the continuous measurement points of the existing track centerline, and then reconstruct the standard railway plane alignment based on these measurement points.

[0003] When faced with line reconstruction involving a large number of locally strongly constrained sections, existing methods have certain limitations: since most of them use the "intersection method" for optimization model design, when adjusting the intersection parameters, all line elements within adjacent intersections are affected, resulting in a large chain reaction that can easily cause lines that originally meet the constraints to become non-compliant; usually, all intersections are treated "equally", and there is a lack of sufficient attention to the sections that violate the constraints, making it impossible to directly lock and optimize these sections, thus affecting the overall efficiency of the reconstruction; in the search process, the constraints are handled too simply, and are often directly abandoned once a violation occurs, or a penalty function is added. Under strong constraint conditions, only an initial solution that fails to meet all constraints can be generated.

[0004] Therefore, in view of the shortcomings of existing linear reconstruction methods under strong constraints, it is urgent to establish an optimization model with a smaller influence range and suitable for the reconstruction of local strong constraint sections, so as to perform intelligent optimization search on the existing linear parameters under strong constraints and improve the quality and efficiency of reconstruction. Summary of the invention

[0005] The purpose of the present invention is to provide a line reconstruction optimization method, medium and device for a local strong constraint section, and its specific technical solution is as follows:

[0006] A line reconstruction optimization method for a local strong constraint section comprises the following steps:

[0007] S1: Import the main line element set, connecting line element set and initial plane plan of the railway line;

[0008] The main line element set includes a straight line measurement point set and circular curve point set The connecting line element set includes the front transition curve measurement point set The post-transition curve measurement point set ;

[0009] S2: perform constraint detection on the line elements in the initial plane solution and generate strong constraint segments;

[0010] S3: Optimizing the initial plane plan, including optimizing and fixing the strongly constrained segments and fitting and optimizing the remaining segments based on the optimized and fixed strongly constrained segments, to obtain a reconstructed plane plan;

[0011] S4: Use the strongly constrained oscillation iteration method to check the point-line consistency and constraint conditions of the reconstructed plane plan until the entire line no longer forms a constrained section, and the final plane plan is obtained.

[0012] Preferably, the step S2 specifically includes:

[0013] S2-1: traverse all main line element sets and connection line element sets, check the constraints of each line element, and mark it as a constrained line element if it violates the constraint conditions to form a set;

[0014] S2-2: For each constrained main line element, analyze whether its adjacent main line elements also violate the constraint;

[0015] If the adjacent main line elements do not violate the constraints, the constrained main line elements will be divided into a constraint section separately. If the adjacent main line elements also violate the constraints, these adjacent constrained main line elements and the connecting line elements between them will be integrated into a constraint section.

[0016] S2-3: The connecting line element violating the constraint and its adjacent main line elements before and after are divided into a constraint section;

[0017] S2-4: Check whether there are adjacent or overlapping constraint segments, and if so, merge them into a single larger constraint segment;

[0018] S2-5: A quantitative assessment model for line default is proposed. Through the difference model, the constraint section is scored from two aspects: geometric constraints and dialing constraints. If the constraint section exceeds any preset threshold, it is identified as a strong constraint section.

[0019] Preferably, the line default quantitative assessment model specifically includes:

[0020] Geometric default assessment:

[0021] The difference model is used for quantitative evaluation, and the difference between the allowable value and the actual value of each violated geometric constraint is used as the geometric violation quantitative score. , and considering the weight factor, the formula is as follows:

[0022] ;

[0023] in, is the weight factor of the geometric constraint; is the function for calculating the difference of geometric constraints;

[0024] Said The following four constraints are included:

[0025] ①、The minimum length of the straight line between two curves:

[0026] ;

[0027] in, To this end, the length of the straight line element in the constrained segment; The minimum straight line length required by the specification;

[0028] ②Minimum curve radius:

[0029] ;

[0030] in, To this end, the radius of the circular curve element in the constrained segment; is the radius of the minimum circular curve required by the specification;

[0031] ③. Minimum curve length:

[0032] ;

[0033] in, To this end, the length of the circular curve element in the segment is constrained; is the minimum curve length required by the specification;

[0034] ④. Minimum length of transition curve:

[0035] ;

[0036] in, To this end, the length of the transition curve line element in the segment is constrained; is the minimum transition curve length required by the specification;

[0037] Assessment of the degree of default of the allocation amount:

[0038] The difference model is used to evaluate the degree of default in the allocation volume. The difference between the actual allocation volume at the default measurement point and the specified allocation volume is calculated as the quantitative score of the allocation volume default. To quantify the degree of default, the formula is as follows:

[0039] ;

[0040] in, To constrain the dialing of the measuring points in the section, the left side of the line's forward direction is negative, and the right side is positive; is the allowable dial value; and Evaluation to determine the constraint section Is it a strongly constrained section?

[0041] ;

[0042] Among them, 1 represents the attribute is a strong constraint section; 0 represents the attribute is a normal constraint section; is the geometric constraint threshold; is the dialing amount constraint threshold; Represents an OR operation.

[0043] Preferably, the step S3 optimizes the initial plane scheme by using an improved composite ABC algorithm, specifically including:

[0044] S3-1: Traverse all constraint sections , list the strongly constrained section as the primary reconstruction optimization target; form the initial honey source and calculate the maximum amount of measurement points in the strongly constrained section And allowable amount , the larger value of the two is taken as the initial feasible bandwidth of the line , forming a search feasible domain;

[0045] S3-2: The initial search parameters are formed based on the feasible domain. The number of honey sources is , the spatial dimension of each nectar source is defined as , each initial honey source contains 4 dimensions, namely, the line element intercept , the slope of the line 、Circular curve center position and the radius of the circular curve ; Each honey source sets trial to indicate the number of attempts. Set to , which indicates the maximum number of attempts to search for each honey source; the maximum number of iterations is set to ; Generate the initial honey source from the plane initial plan , and then generate the honey source by the following random equation :

[0046] ;

[0047] in, Indicates The first Dimension; and Respectively represent The search lower limit and search upper limit of the dimension; is a random number in the range [0,1];

[0048] S3-3: Entering the hired bee stage, each hired bee will be dispatched to its corresponding nectar source to search for potential new solutions near it, and use the improved search equation to obtain the mutated new nectar source ;

[0049] S3-4: The three candidate solutions found by the employed bees are strictly screened through the modified feasibility criteria, and the optimal solution is retained as the newly discovered nectar source for the employed bees;

[0050] S3-5: The new nectar source and the old nectar source corresponding to the employed bees are adjusted according to the line default quantitative assessment model. The constraint method is used for comparison. If the new nectar source is better, the employed bee will abandon the old nectar source and choose the new nectar source. If the old nectar source is better, the employed bee will retain the old nectar source and the trial parameter of this nectar source will be increased by 1.

[0051] S3-6: Use the three-level sorting method to sort all nectar sources after being updated in the employed bee stage;

[0052] In the first layer, individuals are sorted according to their geometric violation degree, and individuals that meet the geometric constraint are arranged in front of individuals that do not meet the geometric constraint; in the second layer, the two individual groups formed in the previous step are further sorted, and individuals that meet the allocation constraint are arranged in front of individuals that do not meet the allocation constraint; in the third layer, based on the four individual groups that meet both geometric and allocation constraints, only meet geometric constraints, only meet allocation constraints, and do not meet both, they are sorted in order from small to large according to the objective function value, and then the fitness is calculated to judge the quality of the nectar source;

[0053] S3-7: Enter the follower bee stage. Each follower bee selects the nectar source to follow according to the roulette method and searches near the nectar source. The new nectar source found is also based on the adjusted The constraint method compares the new nectar source with the original one; if the new nectar source is better, replace the old one; otherwise, keep the old one and add 1 to its trial parameter;

[0054] S3-8: Enter the scout bee stage, traverse all nectar sources, and check whether there is a nectar source whose trial parameter has reached If the value exists, the nectar source is abandoned and a new nectar source is randomly generated in the solution space to explore potential better areas;

[0055] S3-9: Go to the hired bee stage and continue the loop. If the search loop reaches the set maximum number of iterations, When , the search is terminated, and the plane solution corresponding to the best nectar source is obtained.

[0056] Preferably, the improved search equation in S3-3 includes:

[0057] Search equation 1:

[0058] ;

[0059] in, express A random number between Indicates the neighboring nectar source Honey source, for A random number between is the vibration frequency;

[0060] Search equation two and search equation three:

[0061] ;

[0062] in, Respectively represent 3 different and not equal initial nectar sources, Indicates The first Dimensions, Indicates The first Dimensions, Indicates The first Dimension; is a random number in the range [-0.75, 0.75], As the number of generations increases, The individual with the minimum dialing amount and the individual with the minimum constraint violation degree are selected to form search equation 2 and search equation 3 respectively.

[0063] Preferably, the modified feasibility criteria described in S3-4 include:

[0064] Rule 1: When both individuals are feasible solutions, that is, they satisfy all constraints, the individual with the smallest allocation value is considered better;

[0065] Rule 2: If one individual is a feasible solution and the other is an invalid solution that does not satisfy certain constraints, then the feasible solution will be considered better;

[0066] Rule 3: When both solutions are infeasible, further analysis is required; if one of the solutions has no geometric default and the other has a default, the solution with no geometric default is considered better; if both have geometric defaults, the solution with a smaller degree of geometric default is considered better; if the degree of geometric default is the same, the solution with a smaller degree of volume default is considered better.

[0067] Preferably, the adjustment The binding laws specifically include:

[0068] Set as Geometric constraint relaxation parameters, is the relaxation parameter of the dial constraint, where is a calculation function of the constraint violation degree, using the line default quantitative assessment model described in step S3;

[0069] like and are all feasible solutions, that is, and , and , then select , if and only if ;

[0070] like is a feasible solution, is an invalid solution, that is and , or , then select ;

[0071] like and If all are invalid solutions, judge according to Rule 3 in the feasibility criteria.

[0072] Preferably, the step S4 specifically includes:

[0073] S4-1: Re-dividing the measuring point set according to the new attribution in the reconstructed plane alignment scheme, the measuring point set includes a straight line measuring point set, a curve measuring point set and a transition curve measuring point set;

[0074] S4-2: Check whether the line element ownership of all measuring points has changed before and after reconstruction. If it is found that the line element ownership of any measuring point has changed during the reconstruction process, the measuring point set is re-divided according to the projection of the measuring point on the line at this time and the initial plan of the plane is adjusted synchronously. The method of S2-S3 is repeated to reconstruct again. If the ownership of the measuring point line element no longer changes, it is considered that the point and line are consistent;

[0075] S4-3: Check whether there are constrained sections in the reconstructed line plane. If so, record them and return to the method described in S3 for optimization again. If some constrained sections still cannot be eliminated after multiple iterations, define them as strong constrained sections and restore them first. If new strong constrained sections continue to appear during the iteration process, this indicates that the division of the strong constrained sections affects the reconstruction of the remaining sections. At this time, expand the scope of all strong constrained sections and include the main line elements before and after. Finally, when no more constrained sections are formed in the entire line, the iteration ends and the final plane plan is obtained.

[0076] The application of the technical solution of the present invention has the following beneficial effects:

[0077] (1) The present invention provides a line reconstruction optimization method for a local strong constraint section. According to the coordinates of the measuring points, a dynamic threshold method and a MADS algorithm are used to generate a better initial reconstruction scheme for the existing railway plane line position. A line default quantitative evaluation model is proposed. According to two different types of constraint conditions, corresponding evaluation standards and methods are formulated respectively, which can accurately and comprehensively measure their default situations.

[0078] (2) In view of the problem that existing methods lack effective treatment of strongly constrained sections, this paper proposes to adopt a "difficult first, easy later" fitting strategy to achieve plane linear reconstruction. The strongly constrained sections where the constraints are difficult to meet are first identified and reconstructed, and then other sections are reconstructed. Finally, a strongly constrained oscillation iteration method is proposed to optimize the linear parameters, ensuring the accuracy of local optimization.

[0079] (3) Aiming at the complex constraints faced by the optimization model and the "one-size-fits-all" problem of the current heuristic algorithm directly discarding infeasible solutions when performing line reconstruction optimization under constraints, the present invention uses a composite ABC algorithm to solve the reconstruction problem. During the search process, the algorithm intelligently retains infeasible solutions near the feasible solution area, thereby guiding the algorithm from the infeasible area to the feasible area, with a faster convergence speed. In addition, the algorithm has fewer control parameters and is relatively simple to implement in programming.

[0080] (4) The present invention adopts the existing composite ABC algorithm and considers the particularity of the plane linear reconstruction problem by introducing the MR equation, adjusting the feasibility criterion and adopting the dual The constraints are specifically improved to find the optimal combination of line element parameters that meet the constraints, which improves the restoration of plane line shape under strong constraints and ultimately achieves accurate reconstruction and optimization of any strongly constrained section and the remaining sections of the railway line.

[0081] 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 line reconstruction optimization method for the local strong constraint section mentioned above is implemented.

[0082] 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 line reconstruction optimization method for the local strongly constrained section by executing the executable instructions.

[0083] In addition to the above-described objects, features and advantages, the present invention has other objects, features and advantages. The present invention will be further described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] 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.

[0085] Figure 1 is a flow chart of a line reconstruction optimization method for a local strong constraint section of an embodiment;

[0086] Figure 2 It is a schematic diagram of the measurement point set and the initial plan of the plane;

[0087] Figure 3 It is a schematic diagram of strongly constrained segment identification;

[0088] Figure 4 It is a flowchart of the improved composite ABC algorithm;

[0089] Figure 5 It is a schematic diagram of the consistency of the inspection points and lines and the changes in the constraint sections. DETAILED DESCRIPTION

[0090] The embodiments of the present invention are described in detail below with reference to the accompanying drawings. Example

[0091] like Figure 1 As shown, a line reconstruction optimization method for a local strong constraint section includes the following steps:

[0092] S1: Import the main line element set, connecting line element set and initial plane plan of the railway line;

[0093] See also Figure 2 , the main line element set includes a straight line measurement point set and circular curve point set The connecting line element set includes the front transition curve measurement point set The post-transition curve measurement point set ;

[0094] S2: Constraint detection is performed on the line elements in the initial plan, and the constrained line elements are identified. Constrained segments are synthesized by classifying the constrained line elements, and a line default quantitative assessment model is used. Figure 3 , identify the strongly constrained sections, including:

[0095] S2-1: traverse all main line element sets and connection line element sets, check the constraints of each line element, and mark it as a constrained line element if it violates the constraint conditions to form a set;

[0096] S2-2: For each constrained main line element, analyze whether its adjacent main line elements also violate the constraint;

[0097] If the adjacent main line elements do not violate the constraints, the constrained main line elements will be divided into a constraint section separately. If the adjacent main line elements also violate the constraints, these adjacent constrained main line elements and the connecting line elements between them will be integrated into a constraint section.

[0098] S2-3: The connecting line element violating the constraint and its adjacent main line elements before and after are divided into a constraint section;

[0099] S2-4: Check whether there are adjacent or overlapping constraint segments, and if so, merge them into a single larger constraint segment;

[0100] S2-5: A quantitative assessment model for line default is proposed. Through the difference model, the constraint section is scored from two aspects: geometric constraints and dialing constraints. If the constraint section exceeds any preset threshold, it is identified as a strong constraint section.

[0101] The line default quantitative assessment model specifically includes:

[0102] Geometric default assessment:

[0103] The difference model is used for quantitative evaluation, and the difference between the allowable value and the actual value of each violated geometric constraint is used as the geometric violation quantitative score. , and considering the weight factor, the formula is as follows:

[0104] ;

[0105] in, For the The weight factor of each geometric constraint is set to 1 by default and can be set according to the specific conditions of the line. For example, the weight factor of the minimum curve radius may be increased for areas with strict terrain restrictions; To calculate the A function of the difference of geometric constraints;

[0106] Said There are four constraints:

[0107] ①、The minimum length of the straight line between two curves:

[0108] ;

[0109] in, For this constraint section The length of the line element; The minimum straight line length required by the specification;

[0110] ②Minimum curve radius:

[0111] ;

[0112] in, For this constraint section The radius of the circular curve element; is the radius of the minimum circular curve required by the specification;

[0113] ③. Minimum curve length:

[0114] ;

[0115] in, For this constraint section The length of the circular curve element; is the minimum curve length required by the specification;

[0116] ④. Minimum length of transition curve:

[0117] ;

[0118] in, For this constraint section The length of the transition curve element; is the minimum transition curve length required by the specification;

[0119] In this embodiment, the constraint conditions are specifically:

[0120] (1) The length of the straight line before and after the curve should not be less than the minimum straight line length ;

[0121] ;

[0122] in, For the The length of the line element; The minimum straight line length required by the specification; is the number of straight line elements. The calculation formula is as follows:

[0123] ;

[0124] in, For the The coordinates of the point ZH on the line element; For the The coordinates of the point HZ on the straight line element.

[0125] (2) The radius of the circular curve shall not be less than the minimum circular curve radius ;

[0126] ;

[0127] in, For the The radius of the circular curve element It is the minimum circular curve radius required by the specification.

[0128] (3) The length of the circular curve shall not be less than the minimum circular curve length ;

[0129] ;

[0130] in, For the The arc corresponding to the circular curve element; For the The radius of the circular curve element;

[0131] ;

[0132] in, For the The tangent azimuth of the point HY on the circular curve element; For the The tangent azimuth of the point YH on the circular curve element.

[0133] (4) The length of the transition curve shall not be less than the minimum transition curve length ;

[0134] ;

[0135] in For the The length of the front transition curve of each circular curve element; For the The length of the transition curve after the circular curve element.

[0136] (5) The specified dialing amount shall not exceed the allowable dialing amount.

[0137] Assume that the point set of major structures and control points is , with the left side of the line's forward direction as negative and the right side as positive. To limit the measuring point The negative maximum allowable dial value, To limit the measuring point The maximum allowable positive displacement value is:

[0138] .

[0139] Assessment of the degree of default of the allocation amount:

[0140] The difference model is used to evaluate the degree of default in the allocation volume. The difference between the actual allocation volume at the default measurement point and the specified allocation volume is calculated as the quantitative score of the allocation volume default. To quantify the degree of default, the formula is as follows:

[0141] ;

[0142] in, For this constraint section The dial value of each measuring point is negative for the left side of the line direction and positive for the right side. is the allowable dial value; and Evaluation to determine the constraint section Is it a strongly constrained section?

[0143] ;

[0144] Among them, 1 represents the attribute is a strong constraint section; 0 represents the attribute is a normal constraint section; is the geometric constraint threshold; is the dialing amount constraint threshold; Represents an OR operation.

[0145] S3: The composite ABC (Artificial Bee Colony) algorithm based on the line default quantitative assessment model is improved to apply dual constraints. Based on the strategy of "difficult first, easy later", the strongly constrained section is solved first, and then the optimized strongly constrained section is fixed, and the remaining sections are optimized based on it. The main line element of the remaining section and the nearest main line element of the adjacent strongly constrained section are included in the improved composite artificial bee colony algorithm for solution. During the solution process, the parameters of the main line element of the strongly constrained section are not changed.

[0146] See also Figure 4 The improved composite ABC algorithm optimizes the initial plane scheme specifically including:

[0147] S3-1: Traverse all constraint sections , list the strongly constrained section as the primary reconstruction optimization target; form the initial honey source and calculate the maximum amount of measurement points in the strongly constrained section The larger value of the two is taken as the initial feasible bandwidth of the line. , forming a search feasible domain;

[0148] S3-2: The initial search parameters are formed based on the feasible domain. The number of honey sources is , the spatial dimension of each nectar source is defined as , each initial honey source contains 4 dimensions, namely, the line element intercept , the slope of the line 、Circular curve center position and the radius of the circular curve ; Each honey source sets trial to indicate the number of attempts. Set to , which indicates the maximum number of attempts to search for each honey source; the maximum number of iterations is set to ; Generate the initial honey source from the plane initial plan , and then generate the honey source by the following random equation :

[0149] ;

[0150] in, Indicates The first Dimension; and Respectively represent The search lower limit and search upper limit of the dimension; is a random number in the range [0,1];

[0151] S3-3: Entering the hired bee stage, each hired bee will be dispatched to its corresponding nectar source to search for potential new solutions near it, and use the improved search equation to obtain the mutated new nectar source ;

[0152] The improved search equation includes:

[0153] Search equation 1:

[0154] ;

[0155] in, express A random number between Indicates the first Honey source, for A random number between is the vibration frequency, take 0.4;

[0156] Search equation two and search equation three:

[0157] ;

[0158] in, Respectively represent 3 different and not equal initial nectar sources, Indicates The first Dimensions, Indicates The first Dimensions, Indicates The first Dimension; is a random number in the range [-0.75, 0.75], As the number of generations increases, The individual with the minimum dialing amount and the individual with the minimum constraint violation degree are selected to form search equation 2 and search equation 3 respectively.

[0159] S3-4: The three candidate solutions found by the employed bees are strictly screened through the modified feasibility criteria, and the optimal solution is retained as the newly discovered nectar source for the employed bees;

[0160] The revised feasibility criteria include:

[0161] Rule 1: When both individuals are feasible solutions, that is, they satisfy all constraints, the individual with the smallest allocation value is considered better;

[0162] Rule 2: If one individual is a feasible solution and the other is an invalid solution that does not satisfy certain constraints, then the feasible solution will be considered better;

[0163] Rule 3: When both solutions are infeasible, further analysis is required; if one of the solutions has no geometric default and the other has a default, the solution with no geometric default is considered better; if both have geometric defaults, the solution with a smaller degree of geometric default is considered better; if the degree of geometric default is the same, the solution with a smaller degree of volume default is considered better.

[0164] S3-5: The new nectar source and the old nectar source corresponding to the employed bees are adjusted according to the line default quantitative assessment model. The constraint method is used for comparison. If the new nectar source is better, the employed bee will abandon the old nectar source and choose the new nectar source. If the old nectar source is better, the employed bee will retain the old nectar source and the trial parameter of this nectar source will be increased by 1.

[0165] The adjusted The binding laws specifically include:

[0166] Set as Geometric constraint relaxation parameters, is the relaxation parameter of the dial constraint, where is a calculation function of the constraint violation degree, using the line default quantitative assessment model described in step S3;

[0167] like and are all feasible solutions, that is, and , and , then select , if and only if ;

[0168] like is a feasible solution, is an invalid solution, that is and , or , then select ;

[0169] like and If all are invalid solutions, judge according to Rule 3 in the feasibility criteria.

[0170] S3-6: Use the three-level sorting method to sort all nectar sources after being updated in the employed bee stage;

[0171] In the first layer, individuals are sorted according to their geometric violation degree, and individuals that meet the geometric constraint are arranged in front of individuals that do not meet the geometric constraint; in the second layer, the two individual groups formed in the previous step are further sorted, and individuals that meet the allocation constraint are arranged in front of individuals that do not meet the allocation constraint; in the third layer, based on the four individual groups that meet both geometric and allocation constraints, only meet geometric constraints, only meet allocation constraints, and do not meet both, they are sorted in order from small to large according to the objective function value, and then the fitness is calculated to judge the quality of the nectar source;

[0172] S3-7: Enter the follower bee stage. Each follower bee selects the nectar source to follow according to the roulette method and searches near the nectar source. The new nectar source found is also based on the adjusted The constraint method compares the new nectar source with the original one; if the new nectar source is better, replace the old one; otherwise, keep the old one and add 1 to its trial parameter;

[0173] S3-8: Enter the scout bee stage, traverse all nectar sources, and check whether there is a nectar source whose trial parameter has reached If the value exists, the nectar source is abandoned and a new nectar source is randomly generated in the solution space to explore potential better areas;

[0174] S3-9: Go to the hired bee stage and continue the loop. If the search loop reaches the set maximum number of iterations, When , the search is terminated, and the plane solution corresponding to the best nectar source is obtained.

[0175] S4: See Figure 5 ,The strongly constrained oscillation iterative method is used to detect the point and line consistency and the constraint conditions of the ,reconstructed plane scheme once until the entire line no longer forms a ,constrained section, and the final plane scheme is obtained.

[0176] The step S4 specifically includes:

[0177] S4-1: Re-dividing the measuring point set according to the new attribution in the reconstructed plane alignment scheme, the measuring point set includes a straight line measuring point set, a curve measuring point set and a transition curve measuring point set;

[0178] S4-2: Check whether the line element ownership of all measuring points has changed before and after reconstruction. If it is found that the line element ownership of any measuring point has changed during the reconstruction process, the measuring point set is re-divided according to the projection of the measuring point on the line at this time and the initial plan of the plane is adjusted synchronously. The method of S2-S3 is repeated to reconstruct again. If the ownership of the measuring point line element no longer changes, it is considered that the point and line are consistent;

[0179] S4-3: Check whether there are constrained sections in the reconstructed line plane. If so, record them and return to the method described in S3 for optimization again. If some constrained sections still cannot be eliminated after multiple iterations, define them as strong constrained sections and restore them first. If new strong constrained sections continue to appear during the iteration process, this indicates that the division of the strong constrained sections affects the reconstruction of the remaining sections. At this time, expand the scope of all strong constrained sections and include the main line elements before and after. Finally, when no more constrained sections are formed in the entire line, the iteration ends and the final plane plan is obtained.

[0180] This embodiment further discloses a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the line reconstruction optimization method for the local strong constraint section described above in this embodiment is implemented.

[0181] This embodiment also discloses an electronic device, including: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to execute the line reconstruction optimization method of the local strongly constrained section as described above in this embodiment by executing the executable instructions.

[0182] 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 line reconstruction optimization method for a local strong constraint section, characterized in that: The steps include: S1: Import the main line element set, connecting line element set and initial plane plan of the railway line; The main line element set includes a straight line measuring point set and circular curve point set The connecting line element set includes the front transition curve measurement point set The post-transition curve measurement point set S2: perform constraint detection on the line elements in the initial plane solution and generate strong constraint segments; S3: Optimizing the initial plane plan, including optimizing and fixing the strongly constrained segments and fitting and optimizing the remaining segments based on the optimized and fixed strongly constrained segments, to obtain a reconstructed plane plan; S4: Use the strong-constrained oscillation iteration method to check the point-line consistency and constraint conditions of the reconstructed plane plan until the entire line no longer forms a constrained section, and obtain the final plane plan; The step S2 specifically includes: S2-1: traverse all main line element sets and connection line element sets, check the constraints of each line element, and mark it as a constrained line element if it violates the constraint conditions to form a set; S2-2: For each constrained main line element, analyze whether its adjacent main line elements also violate the constraint; If the adjacent main line elements do not violate the constraints, the constrained main line elements will be divided into a constraint section separately. If the adjacent main line elements also violate the constraints, these adjacent constrained main line elements and the connecting line elements between them will be integrated into a constraint section. S2-3: The connecting line element violating the constraint and its adjacent main line elements before and after are divided into a constraint section; S2-4: Check whether there are adjacent or overlapping constraint segments, and if so, merge them into a single larger constraint segment; S2-5: A quantitative assessment model for line default is proposed. Through the difference model, the constraint section is scored from two aspects: geometric constraints and dialing constraints. If the constraint section exceeds any preset threshold, it is identified as a strong constraint section.

2. The line reconstruction optimization method for a local strong constraint section according to claim 1 is characterized in that: The line default quantitative assessment model specifically includes: Geometric default assessment: The difference model is used for quantitative evaluation, and the difference between the allowable value and the actual value of each violated geometric constraint is used as the geometric violation quantitative score Score c , and considering the weight factor, the formula is as follows: Score c =∑W i ·f(C i ); Among them, W i is the weight factor of the geometric constraint; f(C i ) is a function for calculating the difference of geometric constraints; The f(C i ) includes the following four constraints: ①、The minimum length of the straight line between two curves: f(C1)=∑max(0,L Tmin -L Tj ); Among them, L Tj To constrain the length of the straight line element in this segment; L Tmin The minimum straight line length required by the specification; ②Minimum curve radius: f(C2)=∑max(0,R cmin -R cj ); Among them, R cj The radius of the circular curve element in this constrained segment; R cmin is the radius of the minimum circular curve required by the specification; ③. Minimum curve length: f(C3)=∑max(0,L Cmin -L Cj ); Among them, L Cj To constrain the length of the circular curve element in the segment; L Cmin is the minimum curve length required by the specification; ④. Minimum length of transition curve: f(C4)=∑max(0,l min -l j ); Among them, l j To constrain the length of the transition curve line element in the segment; l min is the minimum transition curve length required by the specification; Assessment of the degree of default of the allocation amount: The difference model is used to evaluate the degree of default in the allocation volume. The difference between the actual allocation volume at the default measurement point and the specified allocation volume is calculated as the quantitative score of the allocation volume default. d To quantify the degree of default, the formula is as follows: Score d =∑max(0,|d k |-d limit ); Among them, d k To constrain the dialing of the measuring points in the section, the left side of the line is negative and the right side is positive; d limit is the allowable dial value; by c and Score d The constraint section L is determined by evaluating con Is it a strongly constrained section? Among them, 1 represents the attribute is a strong constraint section; 0 represents the attribute is a normal constraint section; S climit is the geometric constraint threshold; S dlimit is the dialing constraint threshold; || represents the OR operation.

3. The line reconstruction optimization method for a local strong constraint section according to claim 2 is characterized in that: In step S3, the initial plane scheme is optimized by using an improved composite ABC algorithm, which specifically includes: S3-1: Traverse all constraint segments L con , list the strongly constrained section as the primary reconstruction optimization target; form the initial honey source and calculate the maximum dialing amount d of the measurement point in the strongly constrained section max and the allowable dialing amount, the larger value of the two is taken as the initial feasible bandwidth w of the line to form the search feasible domain; S3-2: The initial search parameters are formed according to the feasible domain. The number of honey sources is SN. The spatial dimension of each honey source is defined as D. Each initial honey source contains 4 dimensions, namely the line element intercept b, the line slope k, the circular curve center position (X, Y) and the circular curve radius R; each honey source is set to trial to indicate the number of attempts, and limit is set to SN×D, indicating the maximum number of attempts for each honey source; the maximum number of iterations is set to N max ; Generate the initial honey source x0 from the plane initial scheme, and then generate the honey source x by the following random equation ij : x ij =x minj +rand·(x maxj -x minj ); Among them, x ij represents the jth dimension of the i-th nectar source; x minj and x maxj They represent the lower and upper limits of the search for the jth dimension respectively; rand is a random number in the range of [0,1]; S3-3: Entering the hired bee stage, each hired bee will be dispatched to its corresponding nectar source to search for potential new solutions near it, and use the improved search equation to obtain the mutated new nectar source v ij ; S3-4: The three candidate solutions found by the employed bees are strictly screened through the modified feasibility criteria, and the optimal solution is retained as the newly discovered nectar source for the employed bees; S3-5: Compare the new nectar source with the old nectar source corresponding to the employed bee itself using the ε constraint method adjusted according to the line default quantitative evaluation model. If the new nectar source is better, the employed bee abandons the old nectar source and chooses the new nectar source; if the old nectar source is better, the employed bee retains the old nectar source, and the trial parameter of this nectar source is increased by 1; S3-6: Use the three-level sorting method to sort all nectar sources after being updated in the employed bee stage; In the first layer, individuals are sorted according to their geometric violation degree, and individuals that meet the geometric constraint are arranged in front of individuals that do not meet the geometric constraint; in the second layer, the two individual groups formed in the previous step are further sorted, and individuals that meet the allocation constraint are arranged in front of individuals that do not meet the allocation constraint; in the third layer, based on the four individual groups that meet both geometric and allocation constraints, only meet geometric constraints, only meet allocation constraints, and do not meet both, they are sorted in order from small to large according to the objective function value, and then the fitness is calculated to judge the quality of the nectar source; S3-7: Entering the follower bee stage, each follower bee selects the nectar source to follow according to the roulette method and searches near the nectar source; the new nectar source found is also compared with the original nectar source based on the adjusted ε constraint method; if the new nectar source is better, the old nectar source is replaced; otherwise, the old nectar source is retained and its trial parameter is increased by 1; S3-8: Enter the scout bee stage, traverse all nectar sources, and check whether the trial parameter of any nectar source has reached the limit value. If so, abandon the nectar source and randomly generate a new nectar source in the solution space to explore potential better areas; S3-9: Go to the hired bee stage and continue the loop. If the search loop reaches the set maximum number of iterations N, max When , the search is terminated, and the plane solution corresponding to the best nectar source is obtained.

4. The line reconstruction optimization method for a local strong constraint section according to claim 3 is characterized in that: The improved search equation described in S3-3 includes: Search equation 1: in, Represents a random number between [-1,1], x kj represents the kth nectar source of the adjacent nectar source, R ij is a random number between (-1,1), MR is the vibration frequency; Search equation two and search equation three: Among them, r1, r2, and r3 represent three different and unequal initial nectar sources. represents the jth dimension of the r1th nectar source, represents the jth dimension of the r2th nectar source, represents the jth dimension of the r3th nectar source; φ ij is a random number between -0.75, 0.

75. F decreases as the number of generations increases. best,j The individual with the minimum dialing amount and the individual with the minimum constraint violation degree are selected to form search equation 2 and search equation 3 respectively.

5. The line reconstruction optimization method for a local strong constraint section according to claim 4 is characterized in that: The revised feasibility criteria described in S3-4 include: Rule 1: When both individuals are feasible solutions, that is, they satisfy all constraints, the individual with the smallest allocation value is considered better; Rule 2: If one individual is a feasible solution and the other is an invalid solution that does not satisfy certain constraints, then the feasible solution will be considered better; Rule 3: When both solutions are infeasible, further analysis is required; if one of the solutions has no geometric default and the other has a default, the solution with no geometric default is considered better; if both have geometric defaults, the solution with a smaller degree of geometric default is considered better; if the degree of geometric default is the same, the solution with a smaller degree of volume default is considered better.

6. The line reconstruction optimization method for a local strong constraint section according to claim 5 is characterized in that: The adjusted ε constraint method specifically includes: Let ε J Geometric constraint relaxation parameter, ε B is the relaxation parameter of the dialing constraint, wherein CV is a calculation function of the constraint violation degree, and the line default quantitative assessment model described in step S3 is adopted; If A and B are both feasible solutions, that is, CV(A)≤ε B And CV(A)≤ε J 、CV(B)≤ε B And CV(B)≤ε J , then choose A if and only if f(A)<f(B); If A is a feasible solution and B is an invalid solution, that is, CV(A)≤ε B And CV(A)≤ε J , CV(B)>ε B Or CV(B)>ε J , then choose A; If both A and B are invalid solutions, then make the judgment according to Rule 3 in the feasibility criteria.

7. The line reconstruction optimization method for a local strong constraint section according to claim 6 is characterized in that: The step S4 specifically includes: S4-1: Re-dividing the measuring point set according to the new attribution in the reconstructed plane alignment scheme, the measuring point set includes a straight line measuring point set, a curve measuring point set and a transition curve measuring point set; S4-2: Check whether the line element ownership of all measuring points has changed before and after reconstruction. If it is found that the line element ownership of any measuring point has changed during the reconstruction process, the measuring point set is re-divided according to the projection of the measuring point on the line at this time and the initial plan of the plane is adjusted synchronously. The method of S2-S3 is repeated to reconstruct again. If the ownership of the measuring point line element no longer changes, it is considered that the point and line are consistent; S4-3: Check whether there are constrained sections in the reconstructed line plane. If so, record them and return to the method described in S3 for optimization again. If some constrained sections still cannot be eliminated after multiple iterations, define them as strong constrained sections and restore them first. If new strong constrained sections continue to appear during the iteration process, this indicates that the division of the strong constrained sections affects the reconstruction of the remaining sections. At this time, expand the scope of all strong constrained sections and include the main line elements before and after. Finally, when no more constrained sections are formed in the entire line, the iteration ends and the final plane plan is obtained.

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 line reconstruction optimization method for a local strong constraint section 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; The processor is configured to execute the line reconstruction optimization method for the local strongly constrained section according to any one of claims 1 to 7 by executing the executable instructions.

Images