Urban green space spatial layout optimization method, system and equipment for multi-function collaborative improvement

Abstract

The invention discloses an urban green space spatial layout optimization method, system and equipment oriented to multi-functional collaborative improvement, and the method comprises the steps: collecting geographical, environmental and social data of a target region, quantifying six dimension indexes of water, heat, carbon, pollution, growth and recreation, constructing a multi-dimensional ecological function evaluation index system, and dynamically allocating weight coefficients; dividing the green space of the target area into a plurality of sub-areas, classifying the sub-areas into corresponding functional units, and matching each functional unit with the green space type space; a multi-objective function model is constructed, and optimization calculation of green land type space configuration is carried out; visualization and decision making are carried out on all schemes in an optimization calculation result; and dynamically adjusting the weight coefficient, and circulating all the steps until a greenbelt space layout updating scheme is obtained. According to the method, the integration of core functions of water, heat, carbon, pollution, generation and recreation is realized, a scientific decision tool is provided for urban green land planning, and the method has great significance in relieving climate change and promoting harmonious symbiosis of people and nature.

Description

Technical Field

[0001] This invention belongs to the field of urban ecological planning and environmental engineering, and specifically relates to a method, system and equipment for optimizing the layout of urban green space for multi-functional synergistic improvement. Background Technology

[0002] With the acceleration of urbanization, various elements such as buildings, transportation, and people are converging in cities, forcing them to face complex ecological and environmental problems such as biodiversity depletion, exacerbated heat island effects, worsening environmental pollution, and low quality of living environments. This further disrupts the balance between ecosystem services such as water, heat, carbon, pollution, biodiversity, and recreation. As a core component of the urban ecosystem, green space undertakes multiple functions, including hydrological regulation, cooling, carbon reduction, pollution control, biodiversity protection, and providing recreational spaces. However, existing green space optimization strategies often rely on empirical layout or single optimization objectives (such as considering only landscape aesthetics), lacking a quantitative balance between the benefits of multiple objectives. This leads to inefficient resource allocation and even functional conflicts (for example, while ecological parks emphasize biodiversity protection, they often suffer from insufficient recreational space or decreased thermal comfort due to dense vegetation). This makes it difficult for urban green spaces to adapt to the long-term goal of "synergistic improvement of multiple functions" in the context of rapid expansion, thus restricting their comprehensive benefits in sustainable urban development. Summary of the Invention

[0003] To address the shortcomings of the existing technologies, this invention provides a method, system, and equipment for optimizing the layout of urban green spaces with multi-functional synergistic enhancement. By optimizing the layout of urban green spaces, it aims to achieve synergistic enhancement of multiple functions, including hydrological regulation, cooling, carbon reduction, pollution control, biodiversity protection, and recreation (abbreviated as: water-heat-carbon-pollution-biology-recreation).

[0004] To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution:

[0005] A method for optimizing the layout of urban green space for multi-functional and synergistic improvement includes:

[0006] Step 1) Collect geographic, environmental and social data of the target area to calculate quantitative indicators of six dimensions: water, heat, carbon, pollution, ecology and recreation. Then, construct a multi-dimensional ecological function evaluation indicator system and dynamically allocate the weight coefficients of each dimension indicator.

[0007] Step 2) Divide the green space of the target area into several sub-areas. Based on the multi-dimensional ecological function evaluation index system, divide each sub-area into a corresponding functional unit. On this basis, match each functional unit with the green space type space of the target area to obtain the correlation matrix between the functional unit and the green space type space of each sub-area.

[0008] Step 3) Based on the correlation matrix between the functional units of each sub-region and the green space type space, a multi-objective function model is constructed with quantitative indicators, economic costs and construction feasibility (including slope and soil conditions) as constraints. The model is then used to optimize the green space type space configuration and obtain the Pareto optimal solution set of the green space type space configuration scheme for each sub-region.

[0009] Step 4) Based on the Pareto optimal solution set of the green space type spatial configuration schemes for each sub-region obtained, make a visual decision on all the schemes.

[0010] Step 5) Use the random forest spatiotemporal clustering model to predict the changing trend of urban user demand, dynamically adjust the weight coefficients in Step 1, and then repeat Steps 2-4 until the green space layout update scheme is obtained, thus realizing the dynamic feedback strategy.

[0011] Furthermore, in step 1, the specific construction method of the multi-dimensional ecological function evaluation index system is as follows:

[0012] 1) Collect geographic data, including topographic data, soil data, and hydrological data, including PM2.5. 2.5 Environmental data, including concentration data, NO2 concentration data, O3 concentration data, and surface temperature data, as well as social data, including population density data, carbon sink data, and green space area data;

[0013] 2) Based on the InVEST model, the topographic, soil, and hydrological data of the geographic data are processed to quantify the hydrological regulation function index data;

[0014] 3) The PM of the environmental data 2.5 Using concentration, NO2 concentration, O3 concentration and surface temperature data as input boundaries, the spatial distribution of urban pollution concentration and temperature was simulated using the SWAT model (validated using Landsat 8 / 9 remote sensing data with a spatial resolution of 15m), and thermal and pollution environmental index data were calculated.

[0015] 4) Based on the urban carbon sink panel data of the aforementioned social data, the Kriging spatial interpolation method was used and multi-scale spatial heterogeneity analysis was performed through the ArcGIS Pro platform to obtain key carbon emission indicator data;

[0016] 5) Combining the hydrological regulation function indicators with the topographic, soil and hydrological data of the geographic data, the MaxEnt algorithm is used to assess the species richness maintenance capacity and obtain biodiversity indicator data;

[0017] 6) Based on the population density and green space area data of the aforementioned social data, the value of recreation services is measured using the landscape visual sensitivity index and accessibility analysis model, and recreation index data is extracted.

[0018] 7) After spatially superimposing the hydrological regulation function index data, the thermal and pollution environmental index data, the carbon emission key index data, the biodiversity index data, and the recreation index data, the entropy weight-TOPSIS method is used to construct a multi-dimensional ecological function evaluation index system for water, heat, carbon, pollution, life, and recreation.

[0019] Furthermore, in step 1, the dynamic allocation method for the weight coefficients of each dimension indicator is as follows:

[0020] Based on the quantitative index data of each dimension of the multi-dimensional ecological function evaluation index system, the analytic hierarchy process is used to determine the initial weights of each ecological function evaluation index (water, heat, carbon, pollution, ecology, and recreation). Combined with urban planning goals, the Dynamic Weight Optimization (DWO) algorithm is used to dynamically iterate the weight coefficients of each dimension index according to data characteristics and task priorities.

[0021] Furthermore, in step 2, the method for dividing the functional units is as follows:

[0022] The green space of the target area is evenly divided into several (p×q) sub-regions. Based on the constructed multi-dimensional ecological function evaluation index system (CEFI), the weighted quantitative index data of each dimension index are matched to the divided sub-regions. By calculating and comparing the dimensionless proportion of the index data of each sub-region, the green space of each sub-region is classified into multiple functional units including ecological buffer zone, recreational activity zone, hydrological regulation zone, and biological habitat zone.

[0023] Furthermore, in step 2, the method for matching the functional unit with the green space type of the target area is as follows:

[0024] Based on dividing all the sub-regions into corresponding functional units, the green space types of the target area are further divided into park green space, production green space, protective green space, auxiliary green space and other green space. The spatial configuration results of the green space types of the target area are matched to each sub-region. The distance between each functional unit point and all green space type spatial points is calculated. The association strength under different distance thresholds is calculated using the bivariate Ripley's K function. Based on the association strength, the association matrix between functional units and green space type spaces is constructed.

[0025] Furthermore, in step 3, the optimization calculation of the spatial configuration of green space types is based on the constructed multi-objective function model and applies the NSGA-II algorithm to generate a Pareto optimal solution set for the spatial configuration of green space types in p×q sub-regions. Specifically, this includes the following steps:

[0026] First, an initial population is randomly generated, and each individual in the population represents a possible spatial configuration scheme of green space type. The fitness of each individual is evaluated according to a multi-objective function.

[0027] Then, individuals in the population are divided into different frontiers according to non-dominance relationships. Individuals in the first frontier are not dominated by any other individuals. The crowding distance of each individual in each frontier is calculated to measure the density of individuals around it (individuals with larger crowding distances are more conducive to maintaining population diversity).

[0028] Secondly, a binary tournament is used to select parent individuals for crossover and mutation, generating new offspring individuals. The parent and offspring populations are then merged, non-dominated sorting is performed, and individuals with higher fitness are selected to form the next generation population, thus obtaining the Pareto optimal solution set for the green space type spatial configuration scheme.

[0029] Furthermore, in step 4, the visual decision-making includes two aspects: 3D visualization and multi-solution comparison. The specific methods are as follows:

[0030] 1) 3D Visualization: Based on the green space type and area parameters of the target area, a 3D digital model of the green space layout is generated using the Lindenmayer system (L-System). First, the branching structure and growth direction of the plants are defined according to the rules of the L-System. Then, the 3D structure of the plants is recursively generated according to the rules. Finally, the generated model is processed for lighting and texture. The spatial configuration schemes of all green space types, their functional units, quantitative indicators, economic costs, and construction feasibility included in the Pareto optimal solution set are visualized.

[0031] 2) Multi-scheme comparison: Based on the multi-objective function model, output a radar chart of the comprehensive effects of all schemes corresponding to the Pareto optimal solution set (covering quantitative indicators, economic costs and construction feasibility), providing decision-makers with visual data reference; decision-makers can weigh and evaluate the comprehensive effects of all schemes, prioritize the deployment of the green space type spatial configuration scheme with the best effect, and output a phased implementation plan.

[0032] Furthermore, in step 5, the calculation process of the random forest spatiotemporal clustering model is as follows:

[0033] First, collect and organize user demand-related data in the target area, including time factors and meteorological factors; use the ST-DBSCAN algorithm to perform spatiotemporal analysis on the user demand-related data in the target area to identify data clusters with similar demand patterns.

[0034] Secondly, a random forest model is trained using historical data to learn the changing patterns of user demand in the target area.

[0035] Furthermore, the trained random forest model is used to predict the future user demand in the target area. The accuracy of the random forest model is evaluated by cross-validation and by using mean absolute error (MAE) as the evaluation metric.

[0036] A system employing the aforementioned method for optimizing the spatial layout of urban green spaces for multi-functional collaborative improvement includes a preprocessing module, a scheme establishment module, and a visualization feedback module; wherein,

[0037] The preprocessing module is used to implement weight allocation and function adaptation functions. Specifically, its functions are as follows: First, it collects geographical, environmental, and social data of the target area to construct a multi-dimensional ecological function evaluation index system for water, heat, carbon, pollution, ecology, and recreation, and dynamically iterates the weight coefficients of each dimension index according to the characteristics of the index data and task priority. Second, it divides the green space in the target area into multiple functional units, including ecological buffer zones, recreational activity zones, hydrological regulation zones, and biological habitat zones, and matches the spatial configuration results of green space types (park green space, production green space, protective green space, ancillary green space, and other green space) in the target area to each functional unit to construct a correlation matrix between functional units and green space.

[0038] The scheme establishment module is used to optimize the spatial configuration scheme of green space types. Specifically, it first constructs a multi-objective function model by taking quantitative indicators, economic costs, and construction feasibility (including slope and soil conditions) as constraints; second, it generates Pareto optimal solution sets for the spatial configuration schemes of green space types in each functional unit based on the multi-objective function model.

[0039] The visualization feedback module is used to realize the visualization decision-making and dynamic feedback of the optimal solution set of green space type spatial configuration schemes. Specifically, its functions are as follows: First, using a three-dimensional digital model of green space layout, the Pareto optimal solution set includes the green space type spatial configuration schemes, their functional units, quantitative indicators, economic costs, and construction feasibility, which are then visualized. Decision-makers can weigh and evaluate the comprehensive effects of all schemes. Second, a random forest spatiotemporal clustering model is proposed to predict the changing trends of urban user demand, and the preprocessing module, scheme establishment module, and visualization feedback module are called in a loop until the updated green space spatial layout scheme is obtained.

[0040] An urban green space layout optimization device for multi-functional collaborative improvement includes: a processor, a memory, a communication interface, and a communication bus. The processor, the memory, and the communication interface communicate with each other through the communication bus. The memory stores at least one executable instruction, which causes the processor to perform the operation corresponding to the urban green space layout optimization method for multi-functional collaborative improvement as described in any one of claims 1-8.

[0041] This invention proposes a method, system, and equipment for optimizing the spatial layout of urban green spaces with multifunctional synergistic enhancement. It integrates the core functions of water, heat, carbon, pollution, ecology, and recreation, providing a scientific decision-making tool for urban green space planning. This is of great significance for mitigating climate change and promoting harmonious coexistence between humans and nature.

[0042] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0043] 1. This invention avoids the functional imbalance often caused by a single objective orientation (such as focusing only on aesthetics or recreation) in traditional green space planning schemes. Instead, this invention can achieve multi-dimensional functional optimization in green space layout through a six-dimensional collaborative model of "water-heat-carbon-pollution-life-recreation" and a dynamic weight allocation mechanism, which has universal applicability.

[0044] 2. The green space layout optimization method, system and equipment proposed in this invention can display the Pareto optimal solution set and its effect in real time through a three-dimensional visualization platform, and realize the dynamic update of the green space layout scheme based on the predicted trend of urban user demand changes. It solves the limitations of traditional urban green space layout methods that rely on human experience, static planning, long decision-making cycle, high error rate and difficulty in coping with rapid urban expansion.

[0045] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the invention and to implement it according to the contents of the specification, the preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings. Specific embodiments of the present invention are given in detail below with reference to the accompanying drawings. Attached Figure Description

[0046] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0047] Figure 1 This is a framework diagram of the urban green space layout optimization system for multi-functional collaborative improvement according to the present invention. Detailed Implementation

[0048] The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings to provide a clearer understanding of the invention's purpose, features, and advantages. It should be understood that the embodiments shown in the drawings are not intended to limit the scope of the invention, but are merely illustrative of the essential spirit of the invention's technical solution.

[0049] In the following description, certain specific details are set forth for the purpose of illustrating various disclosed embodiments in order to provide a thorough understanding of the various disclosed embodiments. However, those skilled in the art will recognize that embodiments may be practiced without one or more of these specific details. In other instances, well-known apparatuses, structures, and techniques associated with this application may not have been shown or described in detail to avoid unnecessarily obscuring the description of the embodiments.

[0050] Unless the context requires otherwise, throughout the specification and claims, the word “comprising” and its variations, such as “including” and “having”, shall be understood to have an open, inclusive meaning, that is, to be interpreted as “including, but not limited to”.

[0051] Throughout this specification, references to "an embodiment" or "an embodiment" indicate that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment. Therefore, the appearance of "in an embodiment" or "an embodiment" in various places throughout the specification does not necessarily refer to the same embodiment. Furthermore, a particular feature, structure, or characteristic may be combined in any manner in one or more embodiments.

[0052] The singular forms “a” and “the” used in this specification and the appended claims include plural references unless otherwise expressly stated herein. It should be noted that the term “or” is generally used to mean “and / or” unless otherwise expressly stated herein.

[0053] Furthermore, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0054] See Figure 1 As shown, an urban green space layout optimization system for multi-functional collaborative improvement includes a preprocessing module, a scheme establishment module, and a visualization feedback module.

[0055] The preprocessing module is used to implement weight allocation and function adaptation. Specifically, its functions are as follows: First, it collects geographical, environmental, and social data of the target area to construct a multi-dimensional ecological function evaluation index system encompassing water, heat, carbon, pollution, ecology, and recreation, and dynamically iterates the weight coefficients of each dimension index based on the characteristics of the index data and task priorities. Second, it divides the green space within the target area into multiple functional units, including ecological buffer zones, recreational activity zones, hydrological regulation zones, and biological habitat zones, and matches the spatial configuration results of green space types (park green space, production green space, protective green space, ancillary green space, and other green space) in the target area to each functional unit, constructing a correlation matrix between functional units and green space.

[0056] The scheme establishment module is used to optimize the spatial configuration schemes of green space types. Its specific functions are as follows: First, it constructs a multi-objective function model using quantitative indicators, economic costs, and construction feasibility (including slope and soil conditions) as constraints; second, it generates Pareto optimal solution sets for the spatial configuration schemes of green space types in each functional unit based on the multi-objective function model.

[0057] The visualization feedback module is used to realize the visualization decision-making and dynamic feedback of the optimal solution set of green space type spatial configuration schemes. Its specific functions are as follows: First, using a 3D digital model of green space layout, it visualizes the Pareto optimal solution set, including the green space type spatial configuration schemes, their functional units, quantitative indicators, economic costs, and construction feasibility. Decision-makers can weigh and evaluate the comprehensive effects of all schemes. Second, it proposes a random forest spatiotemporal clustering model to predict the changing trends of urban user demand, and iteratively calls the preprocessing module, scheme establishment module, and visualization feedback module until an updated green space spatial layout scheme is obtained.

[0058] See Figure 1 As shown, a method for optimizing the layout of urban green space for multi-functional collaborative improvement includes the following steps:

[0059] Step 1) Weight allocation:

[0060] Step 1.1) Data Acquisition:

[0061] Collect geographic data of the target area X soi (i = 1, 2, ..., j) (including topographic, soil, and hydrological data), environmental data X eni (i = 1, 2, ..., j) (including PM) 2.5 (Concentration, NO2 concentration, O3 concentration, and surface temperature data) and social data X sci (i = 1, 2, ..., j) (including population density, carbon sink and green space area data).

[0062] The geographic data collection method involves using UAV remote sensing technology to acquire high-precision topographic data of the target area, including information such as elevation, slope, and aspect, with a spatial resolution of 0.1m. Simultaneously, soil data such as soil type, soil texture, and soil thickness, as well as hydrogeological data, such as groundwater level and surface runoff direction, are collected.

[0063] The environmental data collection method involves deploying an IoT sensor network, evenly distributing multiple monitoring points within the target area, and monitoring PM2.5 in real time. 2.5 The concentrations of pollutants such as NO2 and O3, as well as environmental parameters such as surface temperature, are monitored once per hour.

[0064] The social data collection methods included: collecting information on residents' population density and recreational needs (such as types of recreational activities, frequency of recreation, and duration of recreation) through questionnaires. Simultaneously, mobile phone signaling data was used to extract recreational demand hotspots, and population flow patterns and recreational behavior patterns in different areas were analyzed.

[0065] Step 1.2) Data Preprocessing:

[0066] The preprocessing methods for geographic data are as follows: coordinate transformation and projection are performed on the collected geographic data to unify them to the same geographic coordinate system. Soil data and hydrogeological data are standardized to eliminate dimensional differences.

[0067] The preprocessing methods for environmental data are as follows: filtering the environmental data to remove outliers and noise interference; and using time series analysis methods to interpolate the monitoring data, supplementing missing data and ensuring data integrity and continuity.

[0068] The preprocessing methods for social data include: encoding and classifying the social data; statistically analyzing the questionnaire survey results to extract key information; and performing spatial cluster analysis on mobile phone signaling data to identify the spatial distribution characteristics of recreational demand hotspots.

[0069] Step 1.3) Construction of a quantitative indicator system:

[0070] 1. Hydrological regulation function indicators: Based on the InVEST model, topographic, soil and hydrological data from geographic data are used as inputs. The topographic, soil and hydrological data are processed to calculate indicators such as rainwater retention rate, groundwater recharge and flood control capacity of the target area, and to quantify the hydrological regulation function indicator data.

[0071] 2. Thermal environment index: This refers to the PM2.5 concentration in environmental data. 2.5Using data on the concentrations of pollutants, NO2, O3, and surface temperature as input boundaries, the spatial distribution of urban pollution concentrations and temperatures was simulated using the SWAT model. The results were validated using Landsat 8 / 9 remote sensing data (spatial resolution of 15m) to calculate thermal environment indicators such as the reduction in surface temperature and the mitigation rate of the urban heat island effect in the target area.

[0072] 3. Carbon sink function indicators: Based on urban carbon sink panel data obtained from social surveys, the Kriging spatial interpolation method is used and the ArcGIS Pro platform is used to perform multi-scale spatial heterogeneity analysis, and the annual carbon sequestration and carbon sink efficiency of the target area are converted into key carbon emission indicators.

[0073] 4. Pollution control function indicators: Extract PM2.5 levels for the target area from environmental data. 2.5 Information on the concentrations of pollutants such as NO2 and O3 is used to calculate environmental indicators such as pollutant removal efficiency.

[0074] 5. Biodiversity conservation function indicators: Combining hydrological regulation function indicators and geographical data, the MaxEnt algorithm is used to evaluate biodiversity indicators such as species richness maintenance capacity and habitat connectivity in the target area.

[0075] 6. Recreation and recreation function indicators: Based on population density and green space area data, the landscape visual sensitivity index and accessibility analysis model are used to measure the recreation service value of the target area and extract recreation and recreation indicator data such as accessibility, comfort and activity carrying capacity.

[0076] 7. By spatially overlaying the data of hydrological regulation function indicators, thermal environment indicators, pollution environment indicators, key carbon emission indicators, biodiversity indicators, and recreational indicators, the entropy weight-TOPSIS method is used to construct a multi-dimensional ecological function evaluation index system encompassing water, heat, carbon, pollution, ecology, and recreation.

[0077] The expression for the multi-dimensional ecological function evaluation index system encompassing water, heat, carbon, pollution, ecology, and recreation is as follows:

[0078]

[0079] In equation (1), n ​​is the indicator dimension (n = 1-6, representing six dimensions of indicators: water, heat, carbon, pollution, life, and recreation), and m is the number of indicators for a certain dimension.

[0080] The "water" indicators include rainwater retention rate, groundwater recharge, and flood control capacity;

[0081] The "heat" indicators include the decrease in surface temperature and the mitigation rate of the urban heat island effect;

[0082] The "carbon" indicators include annual carbon sequestration and carbon sink efficiency.

[0083] The "pollution" indicator includes pollutants (PM2.5). 2.5 Removal efficiency of (NO2, O3);

[0084] The "life" indicators include species richness and habitat connectivity;

[0085] The "recreation" index includes accessibility, comfort, and activity carrying capacity.

[0086] Step 1.4) Dynamic weight allocation:

[0087] Based on the quantitative index data of each dimension of the multi-dimensional ecological function evaluation index system, the analytic hierarchy process (AHP) is used to determine the initial weights ω of each ecological function evaluation index (water, heat, carbon, pollution, ecology, and recreation). nm Based on urban planning goals, the Dynamic Weight Optimization (DWO) algorithm is used to dynamically iterate the weight coefficients of each dimension indicator according to data characteristics and task priorities; its function expression is as follows:

[0088]

[0089] In equation (2), n represents the indicator dimension (n = 1-6, representing the six dimensions of water, heat, carbon, pollution, ecology, and recreation, respectively), m represents the number of indicators in a certain dimension, and ω nm ' represents the initial weights for each ecological function evaluation index: water, heat, carbon, pollution, ecology, and recreation, and α is the learning rate. In each iteration, the weights are calculated and updated until the model converges.

[0090] Step 2) Functional adaptation:

[0091] Based on the weight allocation, spatial adaptation is carried out for six types of ecological functions: water, heat, carbon, pollution, ecology, and recreation, covering two steps: functional unit division and function-space matching.

[0092] Step 2.1) Functional unit division:

[0093] 1. In the GIS platform, the green area of ​​the target region is evenly divided into p×q sub-regions (e.g., a 500m×500m grid). Based on the topographic, soil, and hydrological information in the geographic data, the hydrogeological conditions and ecological sensitivity of each sub-region are preliminarily determined.

[0094] 2. Analyze the thermal and pollution characteristics of each sub-region by combining pollutant concentrations and surface temperature distributions from environmental data. Determine the locations of potential recreational activity zones based on the distribution of recreational demand hotspots in social data.

[0095] 3. Based on the constructed multi-dimensional ecological function evaluation index system (CEFI), the quantitative index data of each dimension of dynamically weighted indicators are matched to the divided sub-regions, and the dimensionless proportion of the index data of each sub-region is calculated.

[0096] The formula for calculating the dimensionless specific gravity is as follows:

[0097]

[0098] In equation (3), CEFI pq This represents the quantitative indicator data divided into the p-th row and q-th column sub-region; CEFI pq * represents the dimensionless proportion of the indicator data in the p-th row and q-th column sub-region.

[0099] 4. By comparing the dimensionless proportions of the indicator data of each sub-region, the green space of each sub-region is divided into multiple functional units such as ecological buffer zone, recreational activity zone, hydrological regulation zone, and biological habitat zone.

[0100] Dimensionless specific gravity in the range of 0-0.2 corresponds to an ecological buffer zone (which has the functions of pollution interception, biological habitat, and carbon sink);

[0101] The dimensionless specific gravity range of 0.2-0.5 corresponds to the recreational activity zone (which combines tree-lined trails, rest facilities, and functions as a cooling and carbon sink).

[0102] The dimensionless specific gravity range of 0.5-0.7 corresponds to the hydrological regulation zone (which also has the functions of rain garden, permeable paving, flood control and groundwater recharge).

[0103] Dimensionless specific gravity in the range of 0.7-1.0 corresponds to biological habitats (which also have the function of biodiversity conservation).

[0104] Step 2.2) Function - Spatial Matching:

[0105] Based on the division of all sub-regions into corresponding functional units, the spatial configuration results of green space types (park green space, production green space, protective green space, ancillary green space, and other green spaces) in the target area are matched to each sub-region. The distance between each functional unit point and all green space type spatial points is calculated. The association strength under different distance thresholds is calculated using the bivariate Ripley's K function. Based on the association strength, an association matrix between functional units and green space types is constructed. Its expression is as follows:

[0106]

[0107] In equation (4), K 12 (d) represents the connection strength between functional units and green spaces within a distance d; n1 and n2 are the number of points in the functional units and green spaces, respectively; dij It is the distance between functional unit point i and green space point j; I(d) ij ≤d) represents an indicator function, when d ij β is 1 if ≤ d, otherwise 0; ij It is an edge correction factor used to correct for boundary effects.

[0108] Step 3) Optimize calculations:

[0109] Based on the results of the adaptation between functional units and green space, the optimization calculation of green space type spatial configuration is realized, including model construction and algorithm optimization.

[0110] Step 3.1) Model Building:

[0111] Based on the correlation matrix between the functional units and green space types of each sub-region, and with quantitative indicators, economic costs, and construction feasibility (including slope and soil conditions) as constraints, a multi-objective function model F[K] is constructed. 12 (d)]. The objective function, taking the maximization of quantitative indicator benefits, the minimization of economic costs, and the optimization of construction feasibility, is expressed as follows:

[0112]

[0113] In equation (5), MAX(CEFI) pq * represents maximizing the benefits of quantitative indicators; MIN(Cost) represents minimizing economic costs; IF[Cf(slope,soil)] represents construction feasibility (slope represents slope, soil represents soil conditions).

[0114] Step 3.2) Algorithm optimization:

[0115] The input parameters are set according to the city planning budget and actual needs, with a budget constraint of ≤50,000 yuan and a comprehensive benefit improvement target of ≥30%.

[0116] The NSGA-II algorithm is applied to solve the constructed multi-objective function model, which includes the following steps:

[0117] First, an initial population is randomly generated, and each individual in the population represents a possible spatial configuration scheme of green space type. The fitness of each individual is evaluated according to a multi-objective function.

[0118] Then, individuals in the population are divided into different frontiers according to their non-dominance relationship. Individuals in the first frontier are not dominated by any other individuals. The crowding distance of each individual within each frontier is calculated to measure the density of individuals around them (individuals with larger crowding distances are more conducive to maintaining population diversity).

[0119] Next, a binary tournament is used to select parent individuals for crossover and mutation, generating new offspring individuals. The parent and offspring populations are then merged, non-dominated sorting is performed, and individuals with higher fitness are selected to form the next generation population.

[0120] Ultimately, after multiple iterations, a Pareto optimal solution set for green space type spatial configuration schemes was obtained, including three different types of optimal schemes: high carbon sink type, high recreation type, and balanced type.

[0121] Step 4) Visualize the decision:

[0122] Based on the Pareto optimal solution set of the obtained green space type spatial configuration schemes, a visual decision is made on all schemes within the optimal solution set, including two parts: three-dimensional visualization display and multi-scheme comparison and decision.

[0123] Step 4.1) 3D visualization display:

[0124] 1. Based on the green space type and area parameters of the target area, a 3D digital model of the green space layout is constructed using the Unity engine, and the various solutions in the Pareto optimal solution set are visualized. The 3D digital model of the green space layout details the spatial layout, functional unit division, and spatial distribution characteristics of relevant quantitative indicators for different green space types.

[0125] 2. The branching structure and growth direction of plants are defined by the Lindenmayer (L-System) rules, and the three-dimensional structure of plants is generated recursively according to the rules. Then, the generated model is processed with lighting and texture to make the three-dimensional model more realistic and intuitive.

[0126] 3. Visualize all green space type spatial configuration schemes, their functional units, quantitative indicators, economic costs, and construction feasibility included in the Pareto optimal solution set;

[0127]

[0128] In equation (5), V is the character set. This represents the plant growth elements (branches, leaves); it represents the initial state (starting string); P is the production rule, defined as how characters are replaced with new strings.

[0129] The summer cooling effects of different schemes are compared. For example, Scheme A reduces the surface temperature by 2.5℃, while Scheme B reduces it by 1.8℃. The differences in cooling effects are visually displayed in the 3D model through color changes or temperature contour lines.

[0130] Step 4.2) Comparison and Decision-Making of Multiple Options:

[0131] 1. Based on the multi-objective function model, output a radar chart showing the comprehensive effects of all solutions corresponding to the Pareto optimal solution set, including quantitative indicators, economic costs, and construction feasibility. Decision-makers can intuitively compare the advantages and disadvantages of different solutions using the radar chart.

[0132] The calculation process of the random forest spatiotemporal clustering model is as follows:

[0133] First, data related to user demand in the target area are collected and organized, including time and meteorological factors. The ST-DBSCAN algorithm is then used to perform spatiotemporal analysis on this data, identifying data clusters with similar demand patterns. The objective function expression is as follows:

[0134]

[0135] In equation (6), K represents the number of clusters; C i Indicates the i-th cluster; x s and x t These represent the spatial and temporal characteristics of data point x, respectively. and Cluster C i The spatial and temporal centroids; λ is the spatial and temporal weighting coefficient, used to balance the importance of space and time;

[0136] Secondly, a random forest model is trained using historical data to learn the changing patterns of user demand in the target area; the prediction formula for the random forest is as follows:

[0137]

[0138] In equation (7), B is the number of decision trees; h b (x) is the prediction result of the b-th decision tree; I(h) b (x) = c) represents the indicator function, when h b The value is 1 when (x) = c, otherwise it is 0;

[0139] Furthermore, the trained random forest model is used to predict the future user demand in the target area. The accuracy of the random forest model is evaluated by cross-validation and by using mean absolute error (MAE) as the evaluation metric.

[0140] 2. By combining urban planning goals and actual needs, decision-makers can weigh and evaluate the overall effects of all options, prioritize the deployment of the green space type configuration scheme with the best effect, and output a phased implementation plan, which clarifies the implementation content, time arrangement and resource allocation of each phase.

[0141] Step 5) Dynamic Feedback:

[0142] The random forest spatiotemporal clustering model is used to predict the changing trend of urban user demand. The weight coefficients in step 1 are dynamically adjusted, and then steps 2-4 are executed cyclically until the green space layout update scheme is obtained, thus realizing a dynamic feedback strategy.

[0143] Step 5.1) Data Update and Simulation Analysis:

[0144] 1. Update environmental parameter data collected by the IoT sensor network quarterly, including PM2.5. 2.5 The data includes NO2 and O3 concentrations and surface temperature. Additionally, new social data, such as changes in population density and recreational demand, are collected regularly.

[0145] 2. Using a digital twin platform, simulate extreme rainfall scenarios, input updated data, and analyze the hydrological regulation function of green space layout under different rainfall intensities and durations to assess flood control capacity and rainwater retention effect. Based on the simulation analysis results, adjust vegetation configuration and drainage paths to optimize the hydrological regulation function of green spaces.

[0146] Step 5.1) Public feedback and optimization adjustments:

[0147] 1. Develop a public participation app to collect residents' feedback on the use of recreational facilities, including opinions and suggestions on aspects such as seating density, shade coverage, types of recreational activities, and the condition of the facilities.

[0148] 2. Based on public feedback, the recreational facilities in the recreational activity zone will be optimized and adjusted, such as increasing the number of seats, improving shade conditions, and adding new recreational activity facilities to enhance residents' recreational experience and satisfaction. Simultaneously, the layout of green spaces and vegetation configurations in other functional areas will be appropriately adjusted to better meet the needs of urban residents and adapt to changes in the urban environment.

[0149] The present invention also provides an urban green space layout optimization device for multi-functional collaborative improvement, comprising: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other through the communication bus, the memory is used to store at least one executable instruction, and the executable instruction causes the processor to perform the operation corresponding to the above-mentioned urban green space layout optimization method for multi-functional collaborative improvement.

[0150] The method, system, and equipment for optimizing the layout of urban green spaces with multi-functional synergistic enhancement proposed in this invention integrate the core functions of water, heat, carbon, pollution, ecology, and recreation, providing a scientific decision-making tool for urban green space planning. This is of great significance for mitigating climate change and promoting harmonious coexistence between humans and nature.

[0151] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for optimizing the spatial layout of urban green spaces for multi-functional and synergistic improvement, characterized in that: include: Step 1) Collect geographic, environmental and social data of the target area to calculate quantitative indicators of six dimensions: water, heat, carbon, pollution, ecology and recreation. Then, construct a multi-dimensional ecological function evaluation indicator system and dynamically allocate the weight coefficients of each dimension indicator. Step 2) Divide the green space of the target area into several sub-areas. Based on the multi-dimensional ecological function evaluation index system, divide each sub-area into a corresponding functional unit. On this basis, match each functional unit with the green space type space of the target area to obtain the correlation matrix between the functional unit and the green space type space of each sub-area. Step 3) Based on the correlation matrix between the functional units of each sub-region and the green space type space, and with quantitative indicators, economic costs and construction feasibility as constraints, construct a multi-objective function model, and use it to perform optimization calculations on the green space type space configuration to obtain the Pareto optimal solution set of the green space type space configuration scheme for each sub-region. Step 4) Based on the Pareto optimal solution set of the green space type spatial configuration schemes for each sub-region obtained, make a visual decision on all the schemes. Step 5) Use the random forest spatiotemporal clustering model to predict the changing trend of urban user demand, dynamically adjust the weight coefficients in Step 1, and then repeat Steps 2-4 until the green space layout update scheme is obtained, thus realizing the dynamic feedback strategy.

2. The method for optimizing the layout of urban green space for multi-functional collaborative improvement according to claim 1, characterized in that, In step 1, the specific construction method of the multi-dimensional ecological function evaluation index system is as follows: 1) Collect geographic data, including topographic data, soil data, and hydrological data, including PM2.

5. 2.5 Environmental data, including concentration data, NO2 concentration data, O3 concentration data, and surface temperature data, as well as social data, including population density data, carbon sink data, and green space area data; 2) Based on the InVEST model, the topographic, soil, and hydrological data of the geographic data are processed to quantify the hydrological regulation function index data; 3) The PM of the environmental data 2.5 Using data on concentrations of pollutants, NO2 concentrations, O3 concentrations, and surface temperature as input boundaries, the SWAT model is used to simulate the spatial distribution of urban pollution concentrations and temperatures, and to calculate thermal and pollution environmental indicators. 4) Based on the urban carbon sink panel data of the aforementioned social data, the Kriging spatial interpolation method was used and the ArcGIS Pro platform was used to perform multi-scale spatial heterogeneity analysis to obtain the key carbon emission indicator data. 5) Combining the hydrological regulation function indicators with the topographic, soil and hydrological data of the geographic data, the MaxEnt algorithm is used to assess the species richness maintenance capacity and obtain biodiversity indicator data; 6) Based on the population density and green space area data of the aforementioned social data, the value of recreation services is measured using the landscape visual sensitivity index and accessibility analysis model, and recreation index data is extracted. 7) After spatially superimposing the hydrological regulation function index data, the thermal and pollution environmental index data, the carbon emission key index data, the biodiversity index data, and the recreation index data, the entropy weight-TOPSIS method is used to construct a multi-dimensional ecological function evaluation index system for water, heat, carbon, pollution, life, and recreation.

3. The method for optimizing the layout of urban green space for multi-functional collaborative improvement according to claim 1, characterized in that, In step 1, the dynamic allocation method for the weight coefficients of each dimension indicator is as follows: Based on the quantitative index data of each dimension of the multi-dimensional ecological function evaluation index system, the analytic hierarchy process is used to determine the initial weights of each ecological function evaluation index (water, heat, carbon, pollution, ecology, and recreation). Combined with urban planning goals, a dynamic weight optimization algorithm is used to dynamically iterate the weight coefficients of each dimension index according to data characteristics and task priorities.

4. The method for optimizing the layout of urban green space for multi-functional collaborative improvement according to claim 1, characterized in that, In step 2, the method for dividing the functional units is as follows: The green space of the target area is evenly divided into several sub-regions. Based on the constructed multi-dimensional ecological function evaluation index system, the weighted quantitative index data of each dimension index are matched to the divided sub-regions. By calculating and comparing the dimensionless proportion of the index data of each sub-region, the green space of each sub-region is classified into multiple functional units including ecological buffer zone, recreational activity zone, hydrological regulation zone, and biological habitat zone.

5. The method for optimizing the layout of urban green space for multi-functional collaborative improvement according to claim 1, characterized in that, In step 2, the method for matching the functional unit with the green space type of the target area is as follows: Based on dividing all the sub-regions into corresponding functional units, the green space types of the target area are further divided into park green space, production green space, protective green space, auxiliary green space and other green space. The spatial configuration results of the green space types of the target area are matched to each sub-region. The distance between each functional unit point and all green space type spatial points is calculated. The association strength under different distance thresholds is calculated using the bivariate Ripley's K function. Based on the association strength, the association matrix between functional units and green space type spaces is constructed.

6. The method for optimizing the layout of urban green space for multi-functional collaborative improvement according to claim 1, characterized in that, In step 3, the optimization calculation of the spatial configuration of green space types is based on the constructed multi-objective function model and applies the NSGA-II algorithm to generate the Pareto optimal solution set of the spatial configuration of green space types in the sub-region. Specifically, it includes the following steps: First, an initial population is randomly generated, and each individual in the population represents a possible spatial configuration scheme of green space type. The fitness of each individual is evaluated according to a multi-objective function. Then, individuals in the population are divided into different frontiers according to non-dominance relationships. Individuals in the first frontier are not dominated by any other individuals. The crowding distance of each individual within each frontier is calculated to measure the density of individuals around it. Secondly, a binary tournament is used to select parent individuals for crossover and mutation, generating new offspring individuals. The parent and offspring populations are then merged, non-dominated sorting is performed, and individuals with higher fitness are selected to form the next generation population, thus obtaining the Pareto optimal solution set for the green space type spatial configuration scheme.

7. The method for optimizing the layout of urban green space for multi-functional collaborative improvement according to claim 1, characterized in that, Step 4, the visual decision-making process includes two parts: 3D visualization and comparison of multiple options. The specific methods are as follows: 1) 3D Visualization: Based on the green space type and area parameters of the target area, a 3D digital model of the green space layout is generated using the Lindenmayer system. First, the branching structure and growth direction of the plants are defined according to the rules of the Lindenmayer system. Then, the 3D structure of the plants is recursively generated according to the rules. Finally, the generated model is processed for lighting and texture. The spatial configuration schemes of all green space types, their functional units, quantitative indicators, economic costs, and construction feasibility included in the Pareto optimal solution set are visualized. 2) Multi-scheme comparison: Based on the multi-objective function model, output a radar chart of the overall effect of all schemes corresponding to the Pareto optimal solution set, providing decision-makers with visual data reference; decision-makers can weigh and evaluate the overall effect of all schemes, prioritize the deployment of the green space type spatial configuration scheme with the best effect, and output a phased implementation plan.

8. The method for optimizing the layout of urban green space for multi-functional collaborative improvement according to claim 1, characterized in that, In step 5, the calculation process of the random forest spatiotemporal clustering model is as follows: First, collect and organize user demand-related data in the target area, including time factors and meteorological factors; use the ST-DBSCAN algorithm to perform spatiotemporal analysis on the user demand-related data in the target area to identify data clusters with similar demand patterns. Secondly, a random forest model is trained using historical data to learn the changing patterns of user demand in the target area. Furthermore, the trained random forest model is used to predict the future user demand in the target area. The accuracy of the random forest model is evaluated by cross-validation and by using mean absolute error as the evaluation metric.

9. A system employing the urban green space layout optimization method for multi-functional collaborative enhancement as described in any one of claims 1-8, characterized in that, It includes a preprocessing module, a solution creation module, and a visualization feedback module; among which, The preprocessing module is used to implement weight allocation and function adaptation functions. Specifically, its functions are as follows: First, it collects geographical data, environmental data, and social data of the target area to construct a multi-dimensional ecological function evaluation index system of water, heat, carbon, pollution, ecology, and recreation, and dynamically iterates the weight coefficients of each dimension index according to the characteristics of the index data and task priority. Second, it divides the green space in the target area into multiple functional units, including ecological buffer zones, recreational activity zones, hydrological regulation zones, and biological habitat zones, and matches the spatial configuration results of green space types in the target area to each functional unit to construct a correlation matrix between functional units and green space. The scheme establishment module is used to optimize the calculation of green space type spatial configuration schemes. Specifically, it first constructs a multi-objective function model by taking quantitative indicators, economic costs, and construction feasibility as constraints; second, it generates Pareto optimal solution sets for green space type spatial configuration schemes in each functional unit based on the multi-objective function model. The visualization feedback module is used to realize the visualization decision-making and dynamic feedback of the optimal solution set of green space type spatial configuration schemes. Specifically, its functions are as follows: First, using a three-dimensional digital model of green space layout, the Pareto optimal solution set includes the green space type spatial configuration schemes, their functional units, quantitative indicators, economic costs, and construction feasibility, which are then visualized. Decision-makers can weigh and evaluate the comprehensive effects of all schemes. Second, a random forest spatiotemporal clustering model is proposed to predict the changing trends of urban user demand, and the preprocessing module, scheme establishment module, and visualization feedback module are called in a loop until the updated green space spatial layout scheme is obtained.

10. A device for optimizing the layout of urban green space for multi-functional collaborative improvement, characterized in that, include: The system includes a processor, a memory, a communication interface, and a communication bus. The processor, the memory, and the communication interface communicate with each other via the communication bus. The memory stores at least one executable instruction, which causes the processor to perform the operation corresponding to the urban green space layout optimization method for multi-functional collaborative improvement as described in any one of claims 1-8.

Images