Historical urban street darning method based on genetic algorithm

Abstract

The invention provides a historical urban street darning method based on a genetic algorithm. Comprising the following steps: step 1, acquiring and processing map and street axis data; 2, street solution set optimization based on a genetic algorithm; and step 3, evaluating and visualizing the optimized street. According to the method, optimization decision making and iterative evaluation are carried out on original street axis data by applying the genetic algorithm, it is ensured that the generated street axis meets the traffic demand and spatial layout of a region, and the method adapts to complex structure characteristics of historical urban areas. And finally, an evaluation index is formed to provide quantitative data support for a designer, so that the street darning design of a historical urban area is guided more scientifically and systematically.

Description

Technical Field

[0001] The invention relates to the fields of computer-aided design, geographic information system, street patching technology and urban renewal, and in particular to a method for patching streets and lanes in a historical urban area based on a genetic algorithm. Background Art

[0002] Urban development has shifted from extensive incremental expansion to stock renewal and optimization. During the period of rapid growth, the focus of urban development tends to be on new cities, and it is easy to ignore the contradiction between the high population concentration and insufficient transportation carrying capacity in the old city.

[0003] The historical urban area is an area with highly concentrated functions and a dense street system in the old urban area. However, due to the protection of historical and cultural elements and the boundaries of property rights, the existing streets and lanes in the historical urban area are mostly slow-moving streets and lanes with narrow scales, which cannot meet the main commuting needs of residents. There are problems such as uneven density, poor continuity, and low traffic efficiency. Therefore, in-depth exploration and improvement of the street system in the historical urban area, and through scientific and reasonable planning and design, enhancing the connectivity and density of streets and lanes and improving traffic efficiency have become important issues that need to be urgently addressed in urban renewal practice.

[0004] In the past practice of historical urban renewal, designers need to conduct on-site surveys and data collection to obtain basic site information, identify street networks with traffic potential, and then analyze them through traffic simulation and environmental assessment, and finally rely on experience to complete the street patching design. However, traditional methods have problems such as inconsistent evaluation standards, low efficiency of design process, and high weight of subjective judgment. Traditional street generation methods, such as tensor field, L-System, and finite element grid, are mainly suitable for regular new development areas, but they show insufficient adaptability in complex built environments and are difficult to handle irregular plots and diverse spatial needs. In addition, these methods lack a systematic optimization mechanism based on multiple evaluation indicators, and cannot effectively balance the density, uniformity, and functionality of the street network, thereby limiting the ability to fine-tune the street structure. Summary of the invention

[0005] The purpose of the present invention is to provide a method for patching streets and lanes in historical urban areas based on genetic algorithms. The present invention uses genetic algorithms to optimize decisions and iteratively evaluate the original street and lane axis data to ensure that the generated street and lane axes meet the regional traffic needs and spatial layout and adapt to the complex structural characteristics of historical urban areas. The evaluation indicators are finally formed to provide quantitative data support for designers, thereby guiding the street and lane patching design of historical urban areas in a more scientific and systematic way.

[0006] In order to achieve the above technical objectives, the present invention adopts the following technical solutions:

[0007] A method for mending streets and alleys in historical urban areas based on genetic algorithm, comprising the following steps:

[0008] A. Obtain the map data and street and alley axis data of the historical urban area, and import the integrated DXF file into the interactive platform;

[0009] B. Use the GeneticSharp framework to implement the genetic algorithm based on the function of mending streets and alleys, add the map data and street and alley axis data obtained in step A to the genetic algorithm, and perform iterative evaluation to obtain the optimal street and alley axis;

[0010] The genetic algorithm specifically includes the following sub-steps:

[0011] B1. Initialization;

[0012] B2. The genetic algorithm generates an axis topology network according to the results of each iteration for fitness evaluation, sets a fitness function F, which is based on multiple street and alley evaluation indicators. The larger the value of the target fitness function F, the more the street and alley meets the requirements in these street and alley evaluation indicators;

[0013] F = SR atio + S Density + S Efficiency + S Evenness + S MeanDepth , where

[0014] S Ratio is the proportion of the length of the street and alley in evaluation index one; S Density is the street and alley density in evaluation index two;

[0015] S Efficiency is the street and alley efficiency in evaluation index three; S Evenness is the plot evenness in evaluation index four; S MeanDepth is the topological depth in evaluation index five;

[0016] C. Visualize the street and alley plan in the interactive platform for the optimal street and alley axis obtained in step B, and visually display the dynamic change trend of the optimization process through the visualization of the fitness evolution curve.

[0017] Beneficial effects: Aiming at the mending of streets and alleys in urban renewal, the present invention simulates various street and alley configuration schemes in historical urban areas through genetic algorithm, and searches for the best solution through continuous optimization and iterative cycle, avoiding information loss caused by subjective tendency of manual judgment in complex influencing factors, and improving the scientificity and accuracy of the results. By introducing multiple types of street and alley evaluation indicators and clear evaluation criteria, it is ensured that the optimized street and alley mending results are superior to those before optimization in all indicators, providing a scientific basis for the quantitative evaluation of street and alley mending results, and improving the accuracy and practicability of street and alley mending methods.

[0018] In an alternative embodiment, step A specifically includes the following sub-steps:

[0019] A1. Obtain map data from an open-source platform, determine the range of alleyway patching, classify the layers of the selected DXF file of the historical urban texture, extract information on plot boundaries, building outlines, road axes, and alleyway axes to be optimized, and convert the above information into the Map file format;

[0020] A2. Extract intersection points, split axes, and remove duplicates from the obtained road axis data to enable reasonable connection with the existing road network. The finally generated axis contains information on the axis start point, axis end point, whether the axis is connected to the original road network, and a complete topological graph structure, laying a data foundation for subsequent optimization calculations.

[0021] Beneficial effects: The present invention obtains map data from an open-source platform through big data, integrates it with the extracted alleyway axis data, and performs data classification processing. It solves the problem of large workload in collecting data by traditional manual methods and improves efficiency.

[0022] In an alternative embodiment, in step B, the iterative evaluation includes the following two evaluation criteria:

[0023] Evaluation criterion 1: Alleyway density reaches 8.0 km / km², the proportion of alleyway length is 0.5 - 0.7, and the plot evenness is 1:2 - 2:3;

[0024] Evaluation criterion 2: The five indicators of alleyway density, the proportion of alleyway length, plot evenness, alleyway efficiency, and topological depth have all improved compared to the evaluation before optimization.

[0025] Beneficial effects: The present invention combines rigid and elastic evaluation conditions to ensure that the optimization results meet the planning standards while allowing for adaptive adjustment. Setting thresholds can reduce the amount of calculation, improve the optimization efficiency, accelerate convergence, and ensure the rationality and adaptability of alleyway network optimization.

[0026] In an alternative embodiment, the initialization of step B1 is specifically as follows: The user customizes and sets the maximum number maxPopulationSize and the minimum number minPopulationSize of the population; set the initial population, and each chromosome RoadNetChromosome in the initial population represents a possible alleyway configuration scheme; the gene coding on each individual represents a minimum generation block, and the minimum generation block is a basic spatial unit divided based on the block zoning theory by arterial road - arterial road, arterial road - branch road, and branch road - branch road, and has a set of spatial attributes for the genetic algorithm. The content stored in the gene is the axis result of the minimum generation block;

[0027] The above content is re - coded and implemented based on the GeneticSharp framework. The key objects and their related formulas in the coding structure are as follows:

[0028] a) Chromosome representation: The coding form of each chromosome RoadNetChromosome is:

[0029] RoadNetChromosome = {G1, G2, …, G n}

[0030] Where, Gi represents the i - th gene, corresponding to a minimum generated block, and n is the number of minimum generated blocks in the population;

[0031] b) The expression formula of each gene coding Gi is:

[0032] Gi = (Li, Fi, Bi), Li = {Li1, Li2, …, Lim}, where,

[0033] Li: represents the axis state in the minimum generated block: where, Lij ∈ {0, 1}, representing the state of the j - th axis in the i - th generated block. Lij = 1 indicates that the axis is enabled, and Lij = 0 indicates that the axis is not enabled. m is the number of axes included in this generated block;

[0034] Fi: Fitness score, used for fitness evaluation;

[0035] Bi: Boolean flag, used to determine whether this minimum generated block unit participates in the optimization calculation. Bi = 1 indicates that this minimum generated block unit participates in the calculation, and Bi = 0 indicates that this minimum generated block unit does not participate in the calculation.

[0036] Beneficial effects: The present invention enables the generation of street and alley networks to be efficient and adaptable through user - defined population parameters and intelligent optimization in combination with the GeneticSharp framework. The chromosome coding method is used to express the street and alley configuration scheme, and combined with the topological connection relationship, a flexible optimization search space is realized. Through Boolean flag control, the operability and constraint consistency during the street and alley optimization process are ensured, and the calculation efficiency is improved.

[0037] In an alternative embodiment, in step B2, the plot uniformity is used to measure the area ratio relationship of the plots formed by the enclosure of streets and alleys, and is defined as the aspect ratio of the sub - blocks formed by the division to ensure the coordination and uniformity of the street and alley blocks;

[0038] The topological depth is the average number of steps or distance calculated from any street and alley node to other nodes in the whole network based on the network topological structure, and is used as an index to measure the depth and connectivity of the spatial distribution of the street and alley network.

[0039] Beneficial effects: The present invention uses the evaluation index of plot uniformity to ensure the coordination and uniformity of street blocks, and uses the evaluation index of topological depth as an index to measure the spatial distribution depth and connectivity of the street network. Through clear quantitative indicators for optimization decision-making, the design result is more scientific and logical.

[0040] In an alternative embodiment, in step B2, the calculation rule of the scoring function S for each evaluation index is as follows, where V is the measured index value and B is the reference index value:

[0041] a) For the three indexes of street density, proportion of street length and plot uniformity, if the measured index value V is not within the threshold range preset by Evaluation Criterion 1 in Claim 3: S = -1;

[0042] b) For the street efficiency index, the higher the value of V, the better the street performance: S = (V - B) / B * 10;

[0043] c) For the topological depth index, the lower the value of V, the better the street performance: S = (B - V) / B * 10.

[0044] Beneficial effects: The present invention makes the optimization of the street network quantitatively evaluable by setting the calculation rules of the evaluation indexes. By comparing and calculating the reference index and the measured value, score normalization is realized, making different evaluation indexes comparable. Based on the threshold setting and the scoring function, the areas to be optimized can be quickly identified, improving the optimization efficiency.

[0045] In an alternative embodiment, it further includes step B3 to optimize the calculation unit of the system. Specifically: when the system runs, it first initializes each calculation unit, assigns a chromosome and its corresponding map data copy to each unit to ensure that each unit can independently execute the evaluation task. The calculation units achieve data synchronization through a lock mechanism to avoid the occurrence of race conditions and ensure the consistency and accuracy of fitness calculation. The relevant formulas are as follows:

[0046] The system processes multiple calculation units in parallel:

[0047] {F1, F2,..., FN} = {Fitness(C1, D1), Fitness(C2, D2),..., Fitness(CN, DN)}

[0048] where F i is the fitness score of the i-th chromosome, C i is the i-th chromosome, D i is the map data corresponding to C i and it is ensured that That is, it is ensured that there will be no repeated assignment or conflicting solutions in the calculation units during the optimization process.

[0049] Beneficial effects: By initializing the calculation unit and performing parallel calculations, the present invention improves the optimization efficiency and the stability of the solution. By matching chromosomes with map data and combining a constraint mechanism to prevent conflicts, the consistency, accuracy, and operability of the optimization scheme are ensured, and the reliability of the street and alley network optimization is enhanced.

[0050] In an alternative embodiment, it further includes step C of outputting the optimization result, specifically: outputting the street and alley configuration scheme with the highest fitness according to the default number of iterations or the user-defined number of iterations, and visually displaying the street and alley layout and street and alley evaluation indicators before and after optimization through the Unity engine, where:

[0051] a) The expression formula for the optimal configuration scheme is:

[0052] BestChromosome = argmaxF(RoadNetChomosomei)

[0053] b) Visualization of the optimization content of the street and alley evaluation indicators: Quantify the change in the indicator values before and after optimization according to the topological depth, plot uniformity, street and alley density, road network connectivity, and the proportion of street and alley length indicators, and display the improvement rate of each indicator through a chart:

[0054]

[0055] where I k is the k-th indicator, and the improvement rate ΔI k is used to display the change in fitness for each iteration during the optimization process through a curve, providing dynamic visualization of the optimization process.

[0056] Beneficial effects: By quantifying the changes in street and alley indicators before and after optimization, the present invention provides an intuitive performance evaluation. Combining with the fitness evolution curve, the optimization process is visualized, enhancing the transparency and operability of the scheme. At the same time, it supports dynamic adjustment of the optimization strategy, improving the scientificity and efficiency of the street and alley configuration scheme.

[0057] In an alternative embodiment, the interactive platform uses the Unity platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 is a flowchart of the method of the present invention;

[0059] Figure 2 Schematic diagram of genetic algorithm encoding;

[0060] Figure 3 Ideal model - superposition street and alley axis map of basic conditions;

[0061] Figure 4Ideal model - optimized generated map;

[0062] Figure 5 Current situation model - basic condition overlay street axis map;

[0063] Figure 6 Current situation model - optimized generated map. Specific implementation manners

[0064] The technical solution of the present invention will be further described below in conjunction with the accompanying drawings of the specification and specific implementation cases, so that those skilled in the art can better understand the present invention and be able to implement it, but the implemented cases cited do not limit the present invention.

[0065] A method for mending streets and alleys in historical urban areas based on genetic algorithms of the present invention, the specific method flow is as Figure 1 shown. First, map and street axis data are obtained and the data is processed; then the genetic algorithm is used to optimize the street layout and evaluated through multiple indicators such as node depth, street and alley uniformity, etc. Finally, it is determined whether adjustment is needed according to the evaluation result, and the final street layout is output. The schematic diagram of the genetic algorithm coding used in this process is as Figure 2 shown, and this process includes the following steps:

[0066] 1. Obtain data related to the urban map and street axis, identify basic elements and establish a block street and alley file.

[0067] 1.1. First, import geographical data according to the map import function, identify and classify the street and alley grades and current situation elements within the block, such as arterial roads, branch roads, streets and alleys, ordinary buildings, historical buildings, water systems, closed areas, etc. These elements will be used as the basic data for street and alley generation.

[0068] 1.2. Import DXF format files on the Unity program platform, perform overlay analysis on each data layer, and obtain a complete MAP format block street and alley map, as Figure 3 shown, and save it for future use. When opening the same map for the second time, you can click the load map option.

[0069] 1.3. In Map Management - Basic Information, you can view the detailed geometric information of existing streets and alleys, buildings, areas, etc. The positions of the basic elements in the Hexian map can be viewed with reference to the legend.

[0070] 2. Add the preliminarily generated street axis to the genetic algorithm for iterative evaluation.

[0071] 2.1. Click Optimization - Create Genetic Algorithm Unit, and check the streets and alleys that need to be optimized. Use the genetic algorithm to optimize the preliminarily generated street axis. The algorithm will iterate according to the set optimization goal to find a better solution.

[0072] In this embodiment, the specific steps of the genetic algorithm are as follows:

[0073] B1. Initialization: The user customizes and sets the maximum population size maxPopulationSize and the minimum population size minPopulationSize; an initial population is set. Each chromosome RoadNetChromosome in the initial population represents a possible street configuration plan; the gene coding on each individual represents a minimum generation block. The minimum generation block is a basic spatial unit divided based on the block partition theory, formed by main road - main road, main road - branch road, and branch road - branch road, and has a set of spatial attributes for the genetic algorithm. The content stored in the gene is the axis result of this minimum generation block.

[0074] The above content is re - coded and implemented based on the GeneticSharp framework. The relevant formulas for each step are as follows:

[0075] g) Population initialization:

[0076] PopulationSize = min(maxPopulationSize, max(minPopulationSize, N))

[0077] where N is the recommended number of the initial population estimated based on the region size or complexity;

[0078] h) Chromosome representation: The coding form of each chromosome RoadNetChromosome is:

[0079] RoadNetChromosome = {G1, G2, …, Gn}

[0080] where Gi represents the i - th gene, corresponding to a minimum generation block, and n is the number of minimum generation blocks in the population;

[0081] i) The expression formula for each gene coding Gi is:

[0082] Gi = (Li, Fi, Bi), Li = {Li1, Li2,..., Lim}, where

[0083] 7) Li: represents the axis state in the minimum generation block: where Lij ∈ {0, 1}, representing the state of the j - th axis in the o - th generation block. Lij = 1 indicates that the axis is enabled, and Lij = 0 indicates that the axis is not enabled. m is the number of axes included in this generation block;

[0084] 8) Fi: fitness score, used for fitness evaluation;

[0085] 9) Bi: A Boolean flag used to determine whether the minimum generated block unit participates in the optimization calculation. Bi = 1 indicates that the minimum generated block unit participates in the calculation, and Bi = 0 indicates that the minimum generated block unit does not participate in the calculation;

[0086] B2. The genetic algorithm generates an axis topology network according to the results of each iteration for fitness evaluation. A fitness function RoadNetFitness(F) is set, and this function is based on multiple street and lane evaluation indicators, including: Indicator 1, the proportion of street and lane length S Ratio ; Indicator 2, street and lane density S Density ; Indicator 3, street and lane efficiency S Efficiency ; Indicator 4, plot evenness S Evenness ; Indicator 5, topological depth S MeanDepth ;

[0087] Convert the above indicators into a scoring function, and the calculation rules of the scoring function S for each indicator are as follows:

[0088] a) The basic score calculation rule is: S = (V - B) / B * 10, where V is the measured index value and B is the reference index value;

[0089] b) If the measured value V is not within the threshold range: S = -1;

[0090] c) If the reference score value is 1: S = V * 10;

[0091] d) If the lower the measured index, the better the street and lane performance: S = (B - V) / B * 10;

[0092] Accumulate each indicator S i to obtain the fitness function:

[0093] F = S Ratio + S Density + S Efficiency + S Evenness + S MeanDepth ,

[0094] The larger the value of the objective function F, the more the street and lane meet the requirements in these evaluation indicators;

[0095] B3. Optimize the computing units of the system: The system improves the computing efficiency through the parallel execution of multiple computing units. Each computing unit independently undertakes the task of calculating the chromosome fitness, that is, evaluating each map scheme, so as to achieve efficient multi-threaded processing. When the system runs, it first initializes each computing unit, allocates a chromosome and its corresponding copy of map data to each unit to ensure that each unit can independently execute the evaluation task. The computing units achieve data synchronization through a lock mechanism to avoid the occurrence of race conditions and ensure the consistency and accuracy of fitness calculation.

[0096] B4. Output of optimization results: Finally, according to the default number of iterations or the number of iterations defined by the user, the street configuration plan BestChromosome with the highest fitness is output, and the street layout and street evaluation indicators before and after optimization are visually displayed through the Unity engine, where:

[0097] a) The expression formula for the optimal configuration plan is:

[0098] BestChromosome = argmax F(RoadNetChromosomei)

[0099] b) Visualization of the optimization content of street evaluation indicators: According to indicators such as topological depth, plot evenness, street density, road network connectivity, and the proportion of street length, quantify the changes in indicator values before and after optimization, and display the improvement rate of each indicator through a chart:

[0100]

[0101] where I k is the k-th indicator, and the improvement rate ΔI k is used to show the change of fitness in each iteration during the optimization process through a curve, providing dynamic visualization of the optimization process, so that designers can clearly understand the convergence situation and improvement efficiency of the objective function;

[0102] 2.2. The optimized objective function is based on the elastic optimization conditions of streets, including indicators such as street length and ratio, street density, street efficiency, plot evenness, and average topological depth. Users can view the change trends of each indicator in real time through the interface.

[0103] 2.3. Users can control the generation process of streets by adjusting the number of iterations or setting the maximum time threshold for optimization. The generated streets will be displayed through the visualization interface in the program, and users can intuitively view and evaluate the results.

[0104] 2.4. Users can modify and set the display style, global parameters, and evaluation weights in the style and settings to obtain more reasonable street results.

[0105] 2.5. After completing the optimization process, the system will output the finally optimized street and alley layout, obtaining an ideal model - optimized generation map as shown in Figure 4 . The final street and alley results can be exported in DXF format for actual urban planning and architectural design.

[0106] The following is a further illustration of the method of this example in combination with Figures 5 - 6 , taking a certain historical block area in He County, Ma'anshan City, Anhui Province as an example:

[0107] 1. Obtain relevant data of the city map, identify the basic elements and establish a street and alley file for the block. The function of this area is defined as a historical block. Import the vector data of the reference case prototype floor plan into the AutoCAD software platform, classify the data according to layers, convert it into polylines and store it in DXF format. Import the DXF into the Unity platform, identify the street and alley levels and current elements and establish a street and alley MAP file for the block, obtaining a current model - basic condition superposed street and alley axis as shown in Figure 5 .

[0108] 2. Add the initially determined street and alley axis to the genetic algorithm for iterative evaluation. Check all plots, set the population threshold to 100 - 150, and click Start to perform the genetic algorithm operation. After setting the iteration time to 240 minutes, the highest score is obtained in the 25th generation. Click to display the best result, obtaining a current model - optimized generation map as shown in Figure 6 .

[0109] 3. After the genetic algorithm completes the iteration, comprehensively evaluate the generated optimal street and alley axis scheme through a series of street and alley evaluation indicators. These indicators include topological depth, plot uniformity, street and alley density, road network connectivity, and the proportion of street and alley length. Each indicator is used as a flexible condition to measure the optimization effect of the road network and determine the degree of realization of the optimization goal.

[0110] In the Unity platform, through the visualization tool, dynamically present the increase and decrease change curves of each indicator during each iteration update process. Designers can view the optimization trend of the indicators in real time and clarify the key improvement directions during the optimization process. At the same time, by comparing the final iteration result with the road network indicators before optimization, calculate the improvement rate of each indicator to intuitively quantify the optimization effect of the road network.

Claims

1. A method for mending the streets and alleys in the historical urban area based on genetic algorithm, characterized in that, It includes the following steps: A. Obtain the map data and street axis data of the historical urban area, and import the integrated DXF file into the interactive platform; B. Use the GeneticSharp framework to implement a genetic algorithm for street weaving function, add the map data and street axis data obtained in step A to the genetic algorithm, and perform iterative evaluation to obtain the optimal street axis; The genetic algorithm specifically includes the following sub-steps: B1. Initialization, set the hyperparameters of the genetic algorithm; B2. The genetic algorithm generates an axis topology network according to the results of each iteration for fitness evaluation, set the target fitness function F, which is based on multiple street evaluation indicators. The larger the value of the target fitness function F, the more compliant the street is with the requirements on these street evaluation indicators; F = S Ratio + S Density + S Efficiency + S Evenness + S MeanDepth , where, S Ratio is the proportion of the length of streets and alleys in Evaluation Index 1; S Density is the density of streets and alleys in Evaluation Index 2; S Efficiency is the evaluation index three - street - lane efficiency; S Evenness is the evaluation index four - land - parcel uniformity; S MeanDepth is the evaluation index five - topological depth; C. Visualize the street plan of the optimal street axis obtained in step B in the interactive platform, and visually display the dynamic change trend of the optimization process through the visualization of the fitness evolution curve.

2. The method for mending and weaving historical urban streets and alleys based on genetic algorithm according to claim 1, wherein Step A specifically includes the following sub-steps: A1. Obtain map data from the open source platform, determine the street weaving range, classify the layers of the selected DXF file of the historical urban area texture, extract the plot boundary, building outline, road axis and the street axis information to be optimized, and convert the above information into the Map file format; A2. Extract intersections, split axes and remove duplicates from the obtained road axis data, so that it can be reasonably connected to the existing road network. The finally generated axis includes the axis start point, axis end point, information on whether the axis is connected to the original road network and the complete topological graph structure, laying a data foundation for subsequent optimization calculations.

3. The method for patching streets and alleys in historical urban areas based on genetic algorithm according to claim 1, characterized in that In step B, the iterative evaluation includes the following two evaluation criteria: Evaluation criterion 1: The street density reaches 8.0 km / km², the proportion of street length is 0.5 - 0.7, and the plot evenness is 1:2 - 2:3; Evaluation criterion 2: The five indicators of street density, proportion of street length, plot evenness, street efficiency and topological depth have all improved compared with the evaluation before optimization.

4. The method for patching streets and alleys in historical urban areas based on genetic algorithm according to claim 1, wherein The initialization of step B1 is specifically: the user customizes and sets the maximum number maxPopulationSize and the minimum number minPopulationSize of the population; set the initial population, and each chromosome RoadNetChromosome in the initial population represents a possible street configuration plan; the gene encoding on each individual represents a minimum generation block, and the minimum generation block is a basic spatial unit divided based on the block partition theory by arterial road - arterial road, arterial road - branch road, and branch road - branch road, with a set of spatial attributes for the genetic algorithm, and the content stored in the gene is the axis result of the minimum generation block; The above content is re - coded and implemented based on the GeneticSharp framework, and the key objects and their related formulas in the coding structure are as follows: a) Chromosome representation: The coding form of each chromosome RoadNetChromosome is: RoadNetChromosome = {G1, G2, …, Gn} Among them, Gi represents the o-th gene, corresponding to a minimum spanning block, and n is the number of minimum spanning blocks in the population; b) The expression formula for each gene encoding Gi is: Gi = (Li, Fi, Bi), Li = {Li1, Li2,..., Lim}, where, Li: represents the axis state in the minimum spanning block: among them, Lij ∈ {0, 1}, representing the state of the j-th axis in the i-th spanning block, Lij = 1 indicates that the axis is enabled, and Lij = 0 indicates that the axis is not enabled, and m is the number of axes included in this spanning block; Fi: fitness score, used for fitness evaluation; Bi: boolean flag, used to determine whether this minimum spanning block unit participates in the optimization calculation, Bi = 1 indicates that this minimum spanning block unit participates in the calculation, and Bi = 0 indicates that this minimum spanning block unit does not participate in the calculation.

5. The method for patching streets and alleys in historical urban areas based on genetic algorithm according to claim 1, characterized in that In step B2, the evaluation index four - plot uniformity is used to measure the area ratio relationship of the plots formed by the enclosure of streets and alleys, defined as the aspect ratio of the sub - blocks formed by the division, to ensure the coordination and uniformity of the street and alley blocks; The evaluation index five - topological depth is based on the network topology structure to calculate the average number of steps or distance from any street and alley node to other nodes in the whole network, as an index to measure the spatial distribution depth and connectivity of the street and alley network.

6. The method for patching historical urban streets and alleys based on genetic algorithm according to claim 1, wherein In step B2, the calculation rule of the scoring function S for each evaluation index is as follows, where V is the measured index value and B is the reference index value: a) For the three indexes of street and alley density, street and alley length ratio, and plot uniformity, if the measured index value V is not within the threshold range preset in evaluation criterion 1 of claim 3: S = - 1; b) For the street efficiency index, the higher the value of V, the better the performance of the street and alley: S = (V - B) / B * 10; c) For the topological depth index, the lower the value of V, the better the performance of the street and alley: S = (B - V) / B * 10.

7. The method for mending and weaving the streets and alleys in the historical urban area based on the genetic algorithm according to claim 1, wherein It also includes the calculation unit for optimizing the system in step B3, specifically: when the system runs, it first initializes each calculation unit, allocates a chromosome and its corresponding map data copy to each unit to ensure that each unit can independently execute the evaluation task. The calculation units achieve data synchronization through a lock mechanism to avoid the occurrence of race conditions and ensure the consistency and accuracy of fitness calculation. The relevant formula is as follows: The system processes multiple calculation units in parallel: {F1, F2,..., FN} = {Fitness(C1, D1), Fitness(C2, D2),..., Fitness(CN, DN)} Among them, F i is the fitness score of the i-th chromosome, C i is the i-th chromosome, D i is the map data corresponding to C i , and it is ensured that this formula is a constraint function, ensuring that there are no conflicts in the allocation between computing units during the optimization process, so as to avoid duplicate or unreasonable solutions during the optimization process.

8. The method for patching streets and alleys in historical urban areas based on genetic algorithm according to claim 6, wherein It also includes the output of the optimization result in step B4, specifically: according to the default number of iterations or the user - defined number of iterations, output the street and alley configuration plan with the highest fitness, and visually display the street and alley layout and street and alley evaluation indexes before and after optimization through an interactive platform, where: a) The expression formula for the optimal configuration plan is: BestChromosome = argmax F(RoadNetChromosome i ) b) Visualization of the optimization content of the street and alley evaluation indexes: According to the indexes of topological depth, plot uniformity, street and alley density, road network connectivity, and street and alley length ratio, quantify the change of the index values before and after optimization, and display the improvement rate of each index through a chart: Among them, I k is the k-th index, and the improved rate ΔI after optimization k is used to show the fitness change at each iteration during the optimization process through a curve, providing dynamic visualization of the optimization process.

9. The method for mending and weaving historical urban streets and alleys based on genetic algorithm according to claim 1, wherein In step C, the interactive platform uses the Unity platform.

Images