Intelligent optimization method for urban construction space layout under multiple constraints

Abstract

This invention belongs to the field of urban construction planning technology, specifically involving an intelligent optimization method for urban construction spatial layout under multiple constraints. The method includes: dividing the land into grid parcels; determining the grid's economic output and ecological loss based on terrain sensitivity and slope angle; initializing a particle swarm to obtain an initial layout scheme; calculating the repulsion flux based on parcel differences and standardizing it as a traction factor; statistically analyzing surrounding consensus land use types; applying spatial assimilation attraction pointing towards surrounding consensus land use types with the traction factor as weight to update the location; calculating the fitness value by combining output, loss, macro-ratio penalty term, and spatial conflict penalty term; and iteratively outputting the optimal urban construction spatial layout scheme. This invention can solve the problems of scattered and interwoven spatial layouts and disorderly land expansion, achieving an optimization effect that balances economic and ecological considerations, maintains a balanced ratio, and achieves high spatial agglomeration.

Description

Technical Field

[0001] This invention relates to the field of urban construction planning technology. More specifically, this invention relates to an intelligent optimization method for urban construction spatial layout under multiple constraints. Background Technology

[0002] Urban spatial layout planning is the core of coordinating regional economic development and ecological protection. Currently, the traditional particle swarm optimization algorithm is widely used in this field. This technology usually divides the area to be planned into a grid, and the position vector of each particle corresponds to a spatial layout scheme. In the iterative search, the algorithm constructs an objective function based on land use economic benefits and construction cost parameters, evaluates the merits of the scheme by calculating the fitness value, and drives the particles to iteratively update based on the individual and global optimal positions, continuously moving towards high fitness areas.

[0003] However, in real-world urban planning scenarios, this search mechanism based on a purely numerical objective function has significant technical limitations: traditional particle swarm optimization treats each grid parcel as an independent optimization dimension, severely neglecting the physical connections and developmental synergies between adjacent parcels in geographic space. During iterative updates, particles, in pursuit of a unidirectional increase in global fitness values, are prone to generating high-frequency abrupt changes in land use types in local areas. Furthermore, this purely numerically driven mechanism easily leads to the unlimited expansion of high-yield land use types, which in turn can disorderly engulf other basic functional areas.

[0004] The aforementioned defects in optimization lead to a severe fragmented and intertwined distribution of land of different natures in the final layout scheme, or cause an overall imbalance. It completely lacks spatial agglomeration and ecological balance that conform to the laws of urban development. This fragmented layout not only significantly increases the construction cost of infrastructure networks, but also easily causes environmental conflicts between industrial pollution and residential areas, making the planning scheme difficult to implement in practice and failing to meet the physical requirements of urban construction for spatial contiguous development and high functional synergy. Summary of the Invention

[0005] To address the technical problems of scattered and interwoven spatial layouts and global disproportion caused by the purely numerical driving of traditional particle swarm optimization algorithms, this invention provides an intelligent optimization method for urban construction spatial layout under multiple constraints, including: Based on the elevation information of the area to be planned, grid plots are divided. Combining the topographic sensitivity of various land uses and the slope angle of each grid plot, the grid economic output and grid ecological loss of each grid plot under different land use types are determined. The particle swarm is initialized, and the particle position vectors are mapped to an initial urban construction spatial layout scheme including various land uses. Based on the differences in land use types and elevation differences between each grid plot and its neighboring grid plots, the repulsion flux of each grid plot is calculated. The repulsion flux is globally standardized to obtain the traction factor, and the surrounding consensus land use types of each grid plot are statistically analyzed. Using the traction factor and the preset spatial adjustment benchmark coefficient as weights, a spatial assimilation attraction pointing to the surrounding consensus land use type is applied in the particle velocity update to update the position of each particle. Based on the grid economic output, grid ecological loss, macro-ratio penalty term between the quantity of various land uses and the ideal ratio corresponding to the updated position of each particle, and the spatial conflict penalty term constituted by the repulsion flux, the fitness value of each particle is calculated. The global optimal position is iteratively updated according to the fitness value, and the optimal urban construction spatial layout scheme is output.

[0006] This invention combines the topographic sensitivity of various land uses with the slope angle of each grid plot to determine the grid's economic output and ecological loss under different land use types, reflecting the physical constraints of topography on urban development potential. Based on the differences in land use types and elevation differences between each grid plot and its neighboring grid plots, this invention calculates the repulsion flux, extracts the spatial topological connections and topographic buffering effects between plots, and obtains a traction factor after globally standardizing this repulsion flux. This factor, along with a preset spatial adjustment benchmark coefficient, is used as a weight in particle velocity updates, applying a spatial assimilation attraction pointing towards the surrounding consensus land use type. This overcomes the blind optimization problem caused by traditional swarm intelligence algorithms being driven only by global numerical targets, enabling severely repulsive isolated grid plots to actively assimilate to surrounding mainstream plots and dynamically correct [the repulsion]. This invention addresses the anomaly of the independent distribution of grid plots. When calculating the fitness value of each particle, it integrates the macro-level ratio penalty term between the quantity of various land uses and the ideal ratio, as well as the spatial conflict penalty term composed of exclusion flux, based on the grid's economic output and ecological loss corresponding to the updated position. At the underlying logic level, it simultaneously cuts off two abnormal evolutionary paths: the unlimited expansion of high-yield land due to economic supremacy and the spatial fragmentation caused by the lack of evaluation guidance. It guides the particle swarm to evolve towards the direction of optimal comprehensive evaluation, balanced functional proportions, and high spatial agglomeration. This solves the dual core problems of scattered and intertwined functional areas and global disproportion in urban construction spatial layout, and outputs an optimal urban construction spatial layout scheme that combines economic and ecological benefits, harmonious proportions, and high agglomeration of large blocks, thereby improving the practical feasibility of the planning scheme.

[0007] Preferably, determining the grid economic base output and grid ecological base loss of each grid plot under different land use types includes: extracting the slope angle of each grid plot to obtain economic and ecological topographic sensitivity factors for various land uses; constructing an exponential decay function using the economic topographic sensitivity factor and the tangent of the slope angle to calculate the economic correction coefficient; constructing a polynomial penalty function using the ecological topographic sensitivity factor and the square of the slope angle tangent to calculate the ecological correction coefficient; obtaining the standard economic output per unit area and the standard ecological base loss per unit area for various land uses under standard flat terrain; multiplying the standard economic output per unit area by the economic correction coefficient to obtain the grid economic base output, and multiplying the standard ecological base loss per unit area by the ecological correction coefficient to obtain the grid ecological base loss.

[0008] This invention extracts slope angle and two types of terrain-sensitive factors, and uses tangent values ​​to construct economic attenuation functions and ecological penalty functions. These are then used to correct standard terrain output and loss to obtain the basic output and loss of the grid. Since different types of land use have significantly different degrees of dependence on terrain flatness, the tangent function amplifies terrain resistance when grid plots are steeper. This causes the economic correction coefficient to decrease exponentially and the ecological correction coefficient to increase quadratically. This actively suppresses the optimization score of land use with high flatness requirements in steep areas, effectively avoiding the high costs of peak shaving and valley filling and the risk of soil erosion faced when planning schemes are actually implemented. This ensures that the fitness assessment can truly reflect the physical constraints of topography on the potential for urban construction.

[0009] Preferably, the economic correction coefficient satisfies the expression: ,in, Indicates the first The land use type of each grid plot The economic correction coefficient below; Represents an exponential function with the natural constant as its base; Indicate land use type Economic topography sensitive factors; Represents the tangent function; Indicates the first The slope angle of each grid plot; the ecological correction coefficient satisfies the expression: ,in, Indicates the first The land use type of each grid plot The ecological correction coefficient below; Indicate land use type Ecological and topographical sensitive factors; Represents the tangent function; Indicates the first The slope angle of each grid plot.

[0010] Preferably, the initialization of the particle swarm, which maps the particle position vectors to an initial urban construction spatial layout scheme containing various land uses, includes: randomly generating the initial position vector and initial velocity vector of each particle in the initial solution space; rounding down the continuous values ​​in the initial position vector of each particle and mapping them to the corresponding discrete index values ​​in the preset set of land use types to form the initial urban construction spatial layout scheme.

[0011] Preferably, the repulsion flux satisfies the expression:

[0012] in, Indicates the first In the layout scheme corresponding to the nth particle, the nth... Expulsion flux of each grid parcel; Indicates the first The set of neighboring grids for each grid parcel; Represents the neighborhood grid set Adjacent grid parcels in the grid; Indicates the first Each grid plot and adjacent grid plots The type criterion factor, when the first Each grid plot and adjacent grid plots In the When the land use type mapped in the layout scheme of each particle is the same, If it is 1, otherwise, =0; Indicates the first Each grid plot and adjacent grid plots The absolute value of the elevation difference between them; Indicates the side length of the grid plot; This represents an exponential function with the natural constant as its base.

[0013] This invention combines the type determination factor of adjacent grid plots with the absolute value of elevation difference to calculate the repulsion flux. Since different attribute lands are adjacent in urban construction and lack natural elevation undulations as barriers, the repulsion effect caused by functional conflict is significantly enhanced in space. When the elevation difference is smaller, the cumulative repulsion value after the exponential function mapping is significantly amplified. It can accurately distinguish the intervention intensity under multi-dimensional spatial structure, objectively quantify and capture potential spatial conflict hotspots caused by attribute differences and lack of physical barriers between plots and the surrounding environment, and provide an accurate numerical measurement benchmark for subsequent elimination of regional fragmentation.

[0014] Preferably, the step of performing global normalization processing on the repulsion flux to obtain the traction factor includes: obtaining the repulsion flux of all grid plots in the current particle layout scheme, and filtering out the largest repulsion flux in the global range; dividing the repulsion flux of the target grid plot by the sum of the largest repulsion flux and a preset anti-zero smoothing constant to obtain the traction factor of the target grid plot.

[0015] Preferably, the step of statistically analyzing the surrounding consensus land use types of each grid plot includes: statistically analyzing the frequency of each land use type appearing in the neighboring grid set of the target grid plot, and selecting the land use type with the highest frequency as the surrounding consensus land use type of the target grid plot.

[0016] Preferably, the particle velocity update expression is:

[0017] in, , The first The particle in the first sequence In the nth iteration, regarding the th The speed of each grid plot; Inertial weights; For individual learning factors; The first random number; For the first The particle about the first The individual historical best location of each grid plot; For the first The particle in the first In the nth iteration, regarding the th The location of each grid plot; As a global learning factor; The second random number; For particle swarm optimization regarding the first The global optimal location of each grid parcel; For the first The particle in the first During the nth iteration The traction factor of each grid plot; This serves as the spatial adjustment reference coefficient. The third random number; For the first The discrete index value corresponding to the surrounding consensus land use type of each grid plot.

[0018] In the particle velocity update iteration, this invention combines the traction factor with the spatial adjustment benchmark coefficient to apply a spatial assimilation attraction pointing towards the surrounding consensus land use type. When the traction factor of the target grid plot is larger, it indicates that its spatial difference in the global layout is higher. The resulting assimilation pull forcibly guides the update speed of the plot to accelerate towards the surrounding consensus, breaking the blindness of the conventional optimization process. This prompts highly conflicting isolated plots to quickly change their attributes in subsequent iterations to maintain assimilation with the surrounding mainstream configuration. From a micro-dynamic level, this actively smooths out local conflicts and ensures the smoothness and orderliness of the functional area boundary evolution.

[0019] Preferably, the fitness value of each particle satisfies the expression:

[0020] in, For the first The particle in the first Fitness value in the next iteration; , , , These are respectively the economic driving force weight, the ecological resistance weight, the macro-allocation penalty weight, and the spatial conflict penalty weight; This is the total set of all grid parcels; , The first The particle about the first The normalized output and normalized loss of each grid plot corresponding to the land use type currently mapped; A set of land use types; For the first The particle in the first In the next iteration, it is mapped to land use type. The number of grid parcels; This represents the total number of grid parcels. Indicate land use type The ideal ratio; For the first The particle in the first During the nth iteration Expulsion flux of each grid parcel.

[0021] This invention is based on normalized economic output and normalized ecological loss, and introduces a macro-level ratio penalty term based on the quantity of various types of land use and the ideal ratio, as well as a spatial conflict penalty term based on exclusion flux, to jointly construct a fitness value. It incentivizes both economic and ecological excellence through positive benefit scores, uses quadratic macro-level ratio penalties to provide moderate compromise and spatial margin flexibility, and uses spatial conflict penalties to suppress scattered and intersecting boundaries. This forces the particle swarm to simultaneously consider the improvement of comprehensive benefits, adhere to the bottom line of the overall planning ratio, and eliminate fragmentation during iteration, thus achieving precise evolutionary guidance towards the direction of optimal comprehensive evaluation, balanced functional proportions, and high agglomeration of large blocks.

[0022] Preferably, the method for obtaining the normalized output is as follows: obtain the maximum output of all grid plots under all possible land use types, denoted as the global maximum output, and divide the grid economic base output by the global maximum output to obtain the normalized output; the method for obtaining the normalized loss is as follows: obtain the maximum loss of all grid plots under all possible land use types in advance, denoted as the global maximum loss, and divide the grid ecological base loss by the global maximum loss to obtain the normalized loss.

[0023] The beneficial effects of this invention are as follows: By extracting the elevation and type differences of grid plots to construct a repulsion flux, this invention explores the spatial topological connections and terrain buffering effects between plots. This physical conflict characteristic is then transformed into a traction factor that intervenes in the velocity update calculation of the particle swarm. During particle updates, a spatial assimilation gravity constraint explicitly pointing to the surrounding consensus land use type is applied, causing isolated grid plots with severe repulsion to actively assimilate to the surrounding mainstream plots. This overcomes the problem of blind optimization caused by traditional swarm intelligence algorithms being driven solely by global numerical objectives. When calculating fitness values, this invention integrates macro-level allocation penalty terms and spatial conflict penalty terms based on the grid's economic base output and ecological base loss, which truly reflect terrain constraints. At the underlying logic level, it simultaneously cuts off two abnormal evolutionary paths: the unlimited expansion of high-yield land due to economic exclusivity and the spatial fragmentation caused by a lack of evaluation guidance. This solves the dual core problems of scattered and overlapping functional areas and global proportional imbalance in urban spatial layout. Ultimately, it outputs an optimal urban spatial layout scheme that combines economic and ecological benefits, harmonious proportions, and high concentration of large-scale areas, improving the practical feasibility of the planning scheme. Attached Figure Description

[0024] Figure 1 This is a flowchart illustrating the intelligent optimization method for urban construction spatial layout under multiple constraints in this invention; Figure 2 This is a schematic diagram illustrating the change in the absolute value of the average velocity of the particle swarm. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0027] This invention discloses an intelligent optimization method for urban construction spatial layout under multiple constraints, referring to... Figure 1 This includes steps S1 to S5: S1. Divide the area to be planned into grid plots based on the elevation information. Combine the topographic sensitivity of various land uses with the slope angle of each grid plot to determine the grid economic base output and grid ecological base loss of each grid plot under different land use types. Initialize the particle swarm and map the particle position vectors to an initial urban construction spatial layout scheme that includes various land uses.

[0028] Specifically, the elevation information of the area to be planned is obtained through digital elevation model data, and the area is divided into grid plots of fixed size according to a preset side length. The center coordinates and average elevation value of each grid plot are recorded. In this embodiment, the preset side length is set to 100 meters. In other embodiments, the implementer can set the preset side length according to the total area of ​​the area to be planned.

[0029] A set of predefined land use types is established, which includes residential land, commercial land, industrial land, and green space. Based on the requirements of the overall urban plan, the ideal ratio of each land use type in the set is obtained. Statistical yearbook data of the area to be planned is obtained as historical statistical data, and the standard economic output per unit area and standard ecological loss per unit area of ​​each land use type in the set under standard flat terrain are extracted from the historical statistical data.

[0030] It should be noted that different types of land use have significantly different degrees of dependence on terrain flatness. For example, industrial areas require extremely high flatness to deploy large equipment, while green areas can fully adapt to or even require mountainous terrain. If a uniform output and loss index under standard flat terrain is used for adaptability assessment, the planning scheme will face high peak-shaving and valley-filling costs and serious soil erosion risks when implemented in actual mountainous or hilly areas. Therefore, this invention constructs a terrain assessment model to transform the slope angle of grid plots into economic attenuation and ecological penalties for different land use types, so as to truly reflect the physical constraints of topography on the potential for urban construction.

[0031] Specifically, no. The land use type of each grid plot Economic adjustment coefficient below Satisfies the expression of the economic decay function:

[0032] in, Indicates the first The land use type of each grid plot The economic correction coefficient below; Represents an exponential function with the natural constant as its base; Indicate land use type Economic topography sensitive factors; Represents the tangent function; Indicates the first The slope angle of each grid plot.

[0033] No. The land use type of each grid plot Ecological correction coefficient Satisfying the ecological penalty function expression:

[0034] in, Indicates the first The land use type of each grid plot The ecological correction coefficient below; Indicate land use type Ecological and topographical sensitive factors; Represents the tangent function; Indicates the first The slope angle of each grid plot.

[0035] Because industrial land requires large-scale leveling of factory buildings, it is most sensitive to terrain slope; commercial land is next; residential land can adapt to certain slopes through stepped construction; while green land is not limited by slope and can stabilize soil and retain water. Therefore, implementers can obtain the economic and ecological terrain sensitivity factors for each land type based on the typical construction engineering quantity standards. Among them, the economic and ecological terrain sensitivity factors for industrial land are the largest, for example, 5 and 4 respectively; the economic and ecological terrain sensitivity factors for commercial land are next, for example, 3 and 2.5 respectively; the economic and ecological terrain sensitivity factors for residential land are relatively small, for example, 1.5 and 1 respectively; and the economic and ecological terrain sensitivity factors for green land are close to zero, for example, 0.1 and 0.1 respectively.

[0036] When the Slope angle of each grid plot The larger the value, the steeper the terrain. The difficulty and damage of non-greening construction on this plot of land increase dramatically and non-linearly. At this point, the value after tangent function mapping becomes larger, and the calculated economic correction coefficient... It shrinks exponentially, while the ecological correction coefficient... This results in a quadratic amplification, which in turn actively suppresses the scores of planning industrial or commercial zones in steep areas at the underlying level.

[0037] Furthermore, the first The land use type of each grid plot Economic adjustment coefficient below With land use type The product of the corresponding standard economic output per unit area is used as the first... The land use type of each grid plot The grid-based economic output; the first The land use type of each grid plot Ecological correction coefficient With land use type The product of the corresponding standard ecological loss per unit area is used as the first... The land use type of each grid plot The amount of loss of the underlying ecological infrastructure of the grid.

[0038] Furthermore, the size and maximum number of iterations of the particle swarm are defined. The initial position vector and initial velocity vector of each particle are randomly generated in the initial solution space. The continuous values ​​in the initial position vector of each particle are rounded down and mapped to the corresponding discrete index values ​​in the land use type set, so that the initial position vector of each particle represents an initial urban construction spatial layout scheme.

[0039] It should be noted that a particle swarm size that is too small will result in insufficient search space coverage, and a maximum number of iterations that is too small will result in the algorithm failing to converge. Conversely, both excessively large particle swarm sizes will significantly increase computation time. Therefore, considering the high-dimensional solution space complexity caused by the large number of grids in urban construction planning and the reasonable computation time limit for a single planning task, this invention sets the particle swarm size to 100 and the maximum number of iterations to 500. In other embodiments, implementers can set the particle swarm size and the maximum number of iterations according to computational resources and solution space complexity.

[0040] S2. Based on the differences in land use type and elevation between each grid plot and its neighboring grid plots, calculate the repulsion flux of each grid plot.

[0041] It should be noted that functional conflicts can occur when different land use types are adjacent in urban construction, and the lack of natural elevation changes as barriers in flat terrain leads to a significant increase in the repulsion effect of such conflicts in space. Traditional methods do not consider this spatial repulsion effect, resulting in a fragmented layout scheme. Therefore, this invention integrates the differences in land parcel attributes and terrain undulation characteristics to calculate the repulsion flux and capture potential spatial conflict hotspots in the layout scheme.

[0042] Specifically, no. In the layout scheme corresponding to the nth particle, the nth... Repulsion flux of each grid parcel Satisfying the expression:

[0043] in, Indicates the first In the layout scheme corresponding to the nth particle, the first... Expulsion flux of each grid parcel; Indicates the first The set of neighboring grids for each grid parcel. In order to be with the first A set consists of all adjacent grid parcels that share a common boundary; Represents the neighborhood grid set Adjacent grid parcels in the grid; Indicates the first Each grid plot and adjacent grid plots The type criterion factor, when both are in the first When the land use type mapped in the layout scheme of each particle is the same, The value is 1, when the mapped land use types are different. =0; Indicates the first Each grid plot and adjacent grid plots The absolute value of the elevation difference between them; It represents the side length of the grid plot and is used to standardize the absolute value of the elevation difference according to the spatial scale of the grid, so as to reflect the relative severity of terrain undulation at a specific grid resolution. This represents an exponential function with the natural constant as its base.

[0044] When the Each grid plot and adjacent grid plots When the land use types are different, the difference item The value is 1, at which point the superposition calculation of the repulsion effect is triggered: when the absolute value of the elevation difference is 1. The smaller the ratio The smaller, the more exponential the function The larger the value obtained after mapping, the greater the repulsive flux calculated by the final summation. The larger the value, the more severe the conflict between the grid plot and the surrounding environment, and the higher the degree of fragmentation of the area.

[0045] S3. Perform global normalization on the repulsion flux to obtain the traction factor.

[0046] It should be noted that, since the absolute values ​​of the repulsion flux differ in different regions, directly using it for particle velocity updates would cause the intervention intensity received by each grid plot to lose a unified benchmark, leading to divergence in the optimization process. Therefore, this invention standardizes the repulsion flux by dividing it by the global maximum value to obtain the traction factor.

[0047] Specifically, no. In the layout scheme corresponding to the nth particle, the first... Traction factor of individual grid plots Satisfying the expression:

[0048] in, Indicates the first In the layout scheme corresponding to the nth particle, the first... The traction factor of each grid plot; Indicates the first In the layout scheme corresponding to the nth particle, the first... Expulsion flux of each grid parcel; This represents the total number of grid parcels. Represents the set of totals Any grid parcel in the grid; Indicates the first In the layout scheme corresponding to the nth particle, the first... The repulsion flux of each grid parcel, when the repulsion flux of the target grid parcel... When it is larger, the traction factor The larger the value, the greater the degree of difference of the grid plot in the overall urban layout, and the greater the need for traction intervention. This represents the maximum value function, used to filter out the th... The largest repulsive flux of a single particle over the global scope; This represents the zero-prevention smoothing constant, used to prevent calculation errors caused by a denominator of 0. In this embodiment, it is... Setting it to 0.001, in other embodiments, the implementer can adjust the anti-zero smoothing constant according to the computer's processing precision. The settings.

[0049] S4. Using the traction factor and the preset spatial adjustment benchmark coefficient as weights, apply a spatial assimilation gravity pointing towards the surrounding consensus land use type during particle velocity update to update the position of each particle.

[0050] It should be noted that since the traction factor only represents the relative degree of difference between plots and lacks control over the balance between the overall spatial agglomeration of urban construction and the anti-assimilation boundary, if no adjustment constraint is introduced, the particles will excessively pursue the absolute homogenization of local land use in the iteration and obliterate the normal functional area boundary. Therefore, this invention obtains the spatial adjustment benchmark coefficient to control the spatial intervention weight when updating the speed.

[0051] Specifically, a spatial adjustment benchmark coefficient is set: When the spatial adjustment benchmark coefficient approaches 0, the particle velocity update is completely dominated by the global optimum and the individual optimum, resulting in the failure of spatial traction and the inability to correct the abnormal independent distribution of grid plots; while when the spatial adjustment benchmark coefficient is too large, the particles excessively pursue the absolute uniformity of local land use in the iteration, causing the particle update to stagnate and the system to fall into local extrema prematurely. Therefore, combined with the urban spatial agglomeration law, the empirical range of the spatial adjustment benchmark coefficient is derived to be 0.15 to 0.2. In this embodiment, the spatial adjustment benchmark coefficient is set to 0.15. In other embodiments, the implementers can set the spatial adjustment benchmark coefficient according to the severity of the terrain undulation of the area to be planned.

[0052] It should be noted that, since the original particle position update mechanism is driven solely by numerical objectives, it leads to disordered jump changes in particle positions. If there is a high degree of conflict between the grid plot and its surrounding environment, the traditional mechanism cannot actively eliminate the differences. Therefore, this invention combines the traction factor and the spatial adjustment benchmark coefficient into the particle velocity equation, and applies an attractive constraint to the particle velocity equation to move towards the dominant land use type in the neighborhood, dynamically guiding the high-conflict grid plot to change its attributes so that it assimilates with the surrounding configuration.

[0053] Specifically, before the speed update, the statistics for the first... In the layout scheme corresponding to the nth particle, the first... The set of neighboring grids for each grid parcel The frequency of each land use type within the area is used to determine the most frequent land use type as the consensus land use type for the surrounding area, and the corresponding discrete index value is used as the [number of] [land use types]. The surrounding consensus index of each grid parcel is denoted as... . No. The particle in the first In the nth iteration, regarding the th The speed of each grid plot Satisfying the expression:

[0054] in, Indicates the first The particle in the first In the nth iteration, regarding the th The speed of each grid plot; Indicates inertia weight; Indicates the first The particle in the first In the nth iteration, regarding the th The speed of each grid plot; Represents an individual's learning factor; Represents the first random number; Indicates the first The particle about the first The individual historical best location of each grid plot; Indicates the first The particle in the first In the nth iteration, regarding the th The location of each grid plot; Represents the global learning factor; Indicates the second random number; The particle swarm represents the first... The global optimal location of each grid parcel; For the first The particle in the first The layout scheme corresponding to the iteration is the first one. The traction factor of each grid plot; Indicates the spatial adjustment reference coefficient; Indicates the third random number; Indicates the first The surrounding consensus index of the grid parcel, i.e. the first grid parcel The discrete index value corresponding to the surrounding consensus land use type of each grid parcel. When the... Traction factor of individual grid plots When it is larger, The larger the value, the stronger the spatial assimilation pull, which forcibly guides the velocity of the grid parcel. Index to surrounding consensus The accelerated direction of this process prompted it to quickly change to a land use type consistent with most surrounding plots in subsequent iterations, thereby eliminating spatial conflicts.

[0055] Due to inertial weight The individual learning factor determines the tendency for particles to maintain their current velocity. The global learning factor determines the tendency of particles to approach their historical best position. The first random number determines the tendency of particles to move closer to the global optimal position. Second random number With the third random number Used to increase the exploratory nature of iterations. , and This represents a random value between 0 and 1. To maintain the balance between global blind search and local optimization capabilities, as well as the final convergence stability of the swarm intelligence algorithm, inertia weights are used. The value ranges from 0.4 to 0.9, representing the individual learning factor. With global learning factor The value range is from 1.5 to 2.5. In this embodiment, the inertia weight is... Setting it to 0.8 will adjust the individual learning factor. Set to 2 to adjust the global learning factor. In other embodiments, the inertia weight and learning factor can be set to 2 according to the convergence speed requirements.

[0056] For example, Figure 2 This is a schematic diagram illustrating the change in the absolute value of the average velocity of the particle swarm. Figure 2 It can be seen that in the early stage of iteration, the absolute value of the average velocity of the particle swarm is relatively high, indicating that the system is in the global exploration stage of the wide-area space. As the number of iterations increases, the absolute value of the average velocity generally shows an oscillating downward trend, indicating that the algorithm gradually transitions from global search to local fine optimization, and stabilizes at a low velocity level in the middle and late stages to achieve smooth convergence. In addition, the absolute value of the average velocity curve shows a sudden rebound jump during the iteration process, such as the rebound of the absolute value of the average velocity after 200 iterations and 360 iterations. This reflects that the particle swarm triggers a perturbation or mutation mechanism to escape the local extreme value trap during the convergence process, effectively maintaining the optimization vitality and diversity of the population.

[0057] Furthermore, according to speed Update # The particle about the first The location of each grid plot.

[0058] S5. Based on the grid economic base output, grid ecological base loss, macro-ratio penalty term between the quantity of various types of land and the ideal ratio corresponding to the updated position of each particle, and the spatial conflict penalty term constituted by the expulsion flux, calculate the fitness value of each particle, iteratively update the global optimal position according to the fitness value, and output the optimal urban construction spatial layout scheme.

[0059] It should be noted that relying solely on traction factors for local spatial assimilation and purely economic and ecological numerical drives in urban planning can lead to the unlimited expansion of high-yield single land use types. Furthermore, if the global fitness evaluation system lacks a penalty mechanism for spatial fragmentation, even if local velocity updates promote land parcel assimilation, the parcels will be discarded by the algorithm due to the lack of global fitness improvement, ultimately resulting in a fragmented and scattered urban layout. Therefore, this invention normalizes economic output and ecological losses, introduces an ideal ratio to construct a macro-level ratio penalty term, and transforms the repulsion flux into a spatial conflict penalty term. This is used to calculate fitness values ​​that guide the particle swarm towards a direction with optimal comprehensive evaluation, balanced functional proportions, and high spatial agglomeration.

[0060] Specifically, the updated particle positions are remapped to discrete land use types, and the corresponding grid economic output and ecological loss for that land use type at a specific grid plot are obtained. The maximum output of all grid plots across all possible land use types is pre-determined and denoted as the global maximum output. ;Pre-obtain the maximum loss of all grid parcels under all possible land use types, and record it as the global maximum loss. Divide the grid-based economic output by the global maximum output. The normalized output is obtained and denoted as Divide the amount of loss in the grid's ecological foundation by the global maximum loss. The normalized loss is obtained and denoted as .

[0061] Calculate the first The particle in the first Fitness value in the next iteration :

[0062] in, Indicates the first The particle in the first Fitness value in the next iteration; Indicates the economic driving force weight; This represents the total number of grid parcels. Represents the set of totals Any grid parcel in the grid; Indicates the first The particle about the first The normalized output corresponding to the land use type currently mapped to each grid parcel; Indicates the weight of ecological resistance; Indicates the first The particle about the first The normalized loss amount corresponding to the land use type currently mapped to each grid parcel; Indicates the macro-level allocation penalty weight; Represents a set of land use types; Represents the set of land use types Any of the land use types listed; Indicates the first The particle in the first In the next iteration's layout scheme, it is mapped to land use type. The number of grid parcels; Indicates the total number of grid parcels; Indicate land use type The ideal ratio; Indicates the weight of the space conflict penalty; Indicates the first The particle in the first In the layout scheme of the next iteration, the first Expulsion flux of each grid parcel.

[0063] Indicates economic benefits; Represents ecological loss; This indicates a macro-level allocation penalty. Indicates the spatial conflict penalty term; when normalized output Larger and normalized loss The smaller the value, the higher the positive return score; when the macro-allocation penalty item... The smaller the value, the more the current plan's land use ratios align with the ideal ratios of the city's overall plan; simultaneously, when spatial conflict penalty items... The smaller the sum, the fewer the boundary conflicts in the current scheme, the higher the degree of contiguousness of large plates, and the better the fitness value obtained in the final calculation. The larger the value, the higher the overall evaluation of the scheme. This indicates that the spatial arrangement of the particles has achieved a high degree of spatial aggregation while improving economic efficiency, taking into account ecological protection and adhering to the ratio red line.

[0064] Since urban development needs to balance economic development and ecological protection, if the economic driving force is... Too large an area will lead to overexploitation; if the ecological resistance weight is too high... If it is too large, it will severely restrict the city's economic development, therefore and The empirical range is 0.3 to 0.7, and the sum of the two is always 1. In this embodiment, Set to 0.6, Set to 0.4. Macro-level allocation penalty weight. To balance overall economic benefits and urban functional diversity, setting an extremely small lower limit for penalty weights leads to particles being completely dominated by economic output, resulting in the unlimited expansion of high-yield land. Conversely, setting an excessively large upper limit for penalty weights locks particles into a rigid proportional framework, losing the flexibility to find optimal clusters in complex terrain, thus severely damaging economic benefits. Therefore, to maintain the bottom line of macro-planning while giving the algorithm appropriate compromises to smooth spatial edges, a macro-proportional penalty weight is derived and selected based on quadratic penalty characteristics. The value range is 1 to 2. In this embodiment, Set to 1.5. Space conflict penalty weight. To control the degree of fragmentation elimination, when a very small lower limit for the penalty weight is set, particle fitness evaluation is mainly dominated by economics and proportion, leading to the failure of spatial clustering constraints and causing severe scattered and intertwined layouts. When an excessively large upper limit for the penalty weight is set, particles become overly afraid of boundary repulsion during iteration, causing various land uses to ignore terrain features and forcibly cluster in corners, seriously deviating from the economic and ecological optimal solution. Therefore, based on spatial smoothing experience, a spatial conflict penalty weight is derived and adopted. The value range is 3 to 5. In this embodiment, In other embodiments, the parameter can be set to 4, depending on the stringency of the urban contiguous development.

[0065] Furthermore, the individual historical best position of each particle and the global best position of the particle swarm are updated based on the fitness value.

[0066] Repeat the particle swarm algorithm iteration process from step S2 to step S5 until the maximum number of iterations is reached. Then, take the mapping result corresponding to the global optimal position at this time as the optimal urban construction spatial layout scheme and output it.

Claims

1. A method for intelligent optimization of urban construction spatial layout under multiple constraints, characterized in that, include: Based on the elevation information of the area to be planned, grid plots are divided. Combining the topographic sensitivity of various land uses with the slope angle of each grid plot, the grid economic output and grid ecological loss of each grid plot under different land use types are determined. Initialize the particle swarm and map the particle position vectors to an initial urban construction spatial layout scheme that includes various land uses; based on the differences in land use types and elevation differences between each grid plot and its neighboring grid plots, calculate the repulsion flux of each grid plot, perform global standardization on the repulsion flux to obtain the traction factor, and statistically analyze the surrounding consensus land use types of each grid plot. Using the traction factor and the preset spatial adjustment benchmark coefficient as weights, a spatial assimilation gravity pointing towards the surrounding consensus land use type is applied in the particle velocity update to update the position of each particle. Based on the grid economic output, grid ecological loss, macro-ratio penalty term between the quantity of various types of land and the ideal ratio, and spatial conflict penalty term formed by the repulsion flux corresponding to the updated position of each particle, the fitness value of each particle is calculated; the global optimal position is iteratively updated according to the fitness value, and the optimal urban construction spatial layout scheme is output.

2. The intelligent optimization method for urban construction spatial layout under multiple constraints as described in claim 1, characterized in that, The determination of the grid's basic economic output and ecological basic loss under different land use types for each grid plot includes: Extract the slope angle of each grid plot to obtain the economic and ecological topographic sensitivity factors for various land uses; construct an exponential decay function using the economic topographic sensitivity factor and the tangent of the slope angle to calculate the economic correction coefficient; construct a polynomial penalty function using the ecological topographic sensitivity factor and the square of the slope angle tangent to calculate the ecological correction coefficient; obtain the standard economic output per unit area and the standard ecological loss per unit area for various land uses under standard flat terrain; multiply the standard economic output per unit area by the economic correction coefficient to obtain the grid's basic economic output, and multiply the standard ecological loss per unit area by the ecological correction coefficient to obtain the grid's basic ecological loss.

3. The intelligent optimization method for urban construction spatial layout under multiple constraints as described in claim 2, characterized in that, The economic correction coefficient satisfies the expression: ,in, Indicates the first The land use type of each grid plot The economic correction coefficient below; Represents an exponential function with the natural constant as its base; Indicate land use type Economic topography sensitive factors; Represents the tangent function; Indicates the first The slope angle of each grid plot; The ecological correction coefficient satisfies the expression: ,in, Indicates the first The land use type of each grid plot The ecological correction coefficient below; Indicate land use type Ecological and topographical sensitive factors; Represents the tangent function; Indicates the first The slope angle of each grid plot.

4. The intelligent optimization method for urban construction spatial layout under multiple constraints as described in claim 1, characterized in that, The initialization of the particle swarm maps particle position vectors to an initial urban construction spatial layout scheme containing various land uses, including: In the initial solution space, the initial position vector and initial velocity vector of each particle are randomly generated; the continuous values ​​in the initial position vector of each particle are rounded down and mapped to the corresponding discrete index values ​​in the preset land use type set to form the initial urban construction spatial layout scheme.

5. The intelligent optimization method for urban construction spatial layout under multiple constraints as described in claim 1, characterized in that, The repulsion flux satisfies the expression: in, Indicates the first In the layout scheme corresponding to the nth particle, the first... Expulsion flux of each grid parcel; Indicates the first The set of neighboring grids for each grid parcel; Represents the neighborhood grid set Adjacent grid parcels in the grid; Indicates the first Each grid plot and adjacent grid plots The type criterion factor, when the first Each grid plot and adjacent grid plots In the When the land use type mapped in the layout scheme of each particle is the same, If it is 1, otherwise, =0; Indicates the first Each grid plot and adjacent grid plots The absolute value of the elevation difference between them; Indicates the side length of the grid plot; This represents an exponential function with the natural constant as its base.

6. The intelligent optimization method for urban construction spatial layout under multiple constraints as described in claim 1, characterized in that, The process of globally standardizing the repulsion flux to obtain the traction factor includes: Obtain the repulsion flux of all grid blocks in the current particle layout scheme, and filter out the largest repulsion flux in the global range; divide the repulsion flux of the target grid block by the sum of the largest repulsion flux and the preset anti-zero smoothing constant to obtain the traction factor of the target grid block.

7. The intelligent optimization method for urban construction spatial layout under multiple constraints as described in claim 1, characterized in that, The statistical analysis of the surrounding consensus land use types for each grid plot includes: The frequency of each land use type in the neighboring grid set of the target grid plot is counted, and the land use type with the highest frequency is selected as the consensus land use type in the surrounding area of ​​the target grid plot.

8. The intelligent optimization method for urban construction spatial layout under multiple constraints as described in claim 7, characterized in that, The particle velocity update expression is: in, , The first The particle in the first sequence In the nth iteration, regarding the th The speed of each grid plot; Inertial weights; For individual learning factors; The first random number; For the first The particle about the first The individual historical best location of each grid plot; For the first The particle in the first In the nth iteration, regarding the th The location of each grid plot; As a global learning factor; The second random number; For particle swarm optimization regarding the first The global optimal location of each grid parcel; For the first The particle in the first During the nth iteration The traction factor of each grid plot; This serves as the spatial adjustment reference coefficient. The third random number; For the first The discrete index value corresponding to the surrounding consensus land use type of each grid plot.

9. The intelligent optimization method for urban construction spatial layout under multiple constraints as described in claim 1, characterized in that, The fitness values ​​of each particle satisfy the expression: in, For the first The particle in the first Fitness value in the next iteration; , , , These are respectively the economic driving force weight, the ecological resistance weight, the macro-allocation penalty weight, and the spatial conflict penalty weight; This is the total set of all grid parcels; , The first The particle about the first The normalized output and normalized loss of each grid plot corresponding to the land use type currently mapped; A set of land use types; For the first The particle in the first In the next iteration, it is mapped to land use type. The number of grid parcels; This represents the total number of grid parcels. Indicate land use type The ideal ratio; For the first The particle in the first During the nth iteration Expulsion flux of each grid parcel.

10. The intelligent optimization method for urban construction spatial layout under multiple constraints as described in claim 9, characterized in that, The method for obtaining the normalized output is as follows: obtain the maximum output of all grid plots under all possible land use types, and record it as the global maximum output. Divide the grid economic base output by the global maximum output to obtain the normalized output. The method for obtaining the normalized loss is as follows: the maximum loss of all grid plots under all possible land use types is obtained in advance and recorded as the global maximum loss. The grid ecological foundation loss is divided by the global maximum loss to obtain the normalized loss.

Images