Urban green land community ecosystem adjustment service optimization factor identification method and device based on fusion algorithm

Abstract

The embodiment of the invention provides an urban green land community ecosystem regulation service optimization factor identification method and device based on a fusion algorithm and electronic equipment, and belongs to the technical field of ecological data processing, and the method comprises the steps: collecting a plant community, an elevation model and meteorological monitoring data to form a basic data set; uniformly quantifying the data and evaluating the ecological service in combination with the measurement and calculation model to form a standardized data set; inputting the features into a plurality of machine learning models, and screening out a key feature set through operations such as guided aggregation and the like; constructing an IOFESRM-LMEM model, outputting an optimal model through a series of operations, and analyzing key factors; and finally, outputting an optimization result, a key factor and an evaluation result by using a multi-dimensional index evaluation model. According to the invention, data support is provided for improving the ecological function of the urban green land, formulating ecological planning and optimizing urban management.

Description

Technical Field

[0001] The present invention relates to the technical field of ecological data processing, and in particular to a method, device and electronic equipment for identifying optimization factors of regulating services of an urban green space community ecosystem based on a fusion algorithm. Background Art

[0002] Ecosystem services are key to addressing climate change and urbanization. Urban green spaces, as the main service provider, play a vital role in maintaining urban ecological balance. In the context of rapid urban expansion, there is a serious spatial inequality in green spaces. Plant communities are the main carriers of ecosystem services provided by urban green spaces. Identifying and quantifying the impact of plant community structure on ecosystem services can clarify the direction of future green space optimization and provide replicable evaluation strategies for urban management and planning. Traditional identification and evaluation of ecosystem service influencing factors mainly evaluate the spatiotemporal changes in green space supply and demand, equity, and ecosystem service characteristics at the landscape scale through remote sensing data combined with land use data inversion or the use of process models such as the invest model. This type of method not only makes it difficult to identify and quantify the impact of green space vegetation on ecosystem services, but also lacks guiding suggestions for optimizing the composition of existing urban green space vegetation.

[0003] With the rapid development of data science and machine learning technology, fusion algorithms based on multiple machine learning methods have gradually become one of the core technologies in dealing with complex data analysis problems. Fusion algorithms (FA) can significantly improve the accuracy, stability and robustness of models by combining the prediction results of multiple different algorithms. Especially in ecosystem service assessment, they can effectively make up for the shortcomings of a single algorithm in dealing with complex data and provide a more comprehensive analysis perspective. Bootstrap Aggregating (BA) is an ensemble learning method that trains multiple independent models by performing multiple random sampling of the original data, and synthesizes the prediction results of each model by voting or averaging, ultimately obtaining a more stable and accurate prediction. This method reduces the variance of the model and enhances the model's tolerance to data fluctuations, making it particularly suitable for dealing with high-dimensional complex data. Linear mixed effects models (LMEMs) are multi-level regression models that introduce random effects to deal with differences between and within groups, so that the model can better adapt to hierarchical data. In the fields of environmental science, forestry, and ecology, machine learning algorithms have been used in many fields such as landscape ecological evaluation, pollutant simulation, forest growth simulation, and large-scale ecosystem service assessment. However, the application of machine learning algorithms and LMEMs in the identification of factors affecting the ecological services of urban green spaces has not yet appeared.

[0004] The above problems become technical problems that need to be solved. Summary of the invention

[0005] In view of this, the embodiments of the present invention provide a method, device and electronic device for identifying optimization factors of urban green space community ecosystem regulation services based on a fusion algorithm, which at least partially solve the problems existing in the prior art.

[0006] In a first aspect, an embodiment of the present invention provides a method for identifying optimization factors of regulation services of an urban green space community ecosystem based on a fusion algorithm, comprising:

[0007] Collect plant community survey data, including multiple indicators of trees, shrubs, and herbaceous plants and vertical lengths from water bodies, obtain DEM digital elevation model data, extract topographic and geomorphic features of the assessed urban green space plant community points, collect meteorological monitoring data, extract basic indicators of ecological regulation services, and aggregate the collected data to form a basic data set;

[0008] The data in the basic data set are uniformly quantified to a range of 0-1 to form a standardized data set. At the same time, the measurement model is combined to evaluate multiple ecological services and quantify the ecological service supply efficiency of different green spaces;

[0009] The standardized data set is input into the random forest, gradient boosting regression tree and support vector machine machine learning models. The training data set is randomly sampled using the guided aggregation method to generate multiple sub-datasets. The feature importance evaluation index is used to quantify the feature contribution, and a feature importance ranking list is constructed. The cross-validation technology is used to verify the impact of the feature set on the performance of different models, and redundant and small-impact features are eliminated. Key features are retained to form an optimized feature set.

[0010] The IOFESRMLMEM model based on the linear mixed effects regression method was constructed, which included a fixed effect and random effect combination framework, a linear basic model library, and covariance matrix construction. The standardized data set after feature screening was used as input to enter the IOFESRMLMEM module. By comparing the fitting effects of multiple alternative models, the information criterion was used to select the optimal model, and the optimal model that can describe the relationship between the ecosystem service optimization factors and the target variables was output. The effects of the optimal model parameters were deeply analyzed to identify the key factors affecting ecosystem services.

[0011] Multi-dimensional performance evaluation indicators are used to evaluate the accuracy and generalization ability of the model, and the contribution values ​​of key factors of quantitative optimization of fixed effect and random effect parameters are analyzed to output the optimized prediction results, key optimization factors and model performance evaluation results.

[0012] According to a specific implementation of an embodiment of the present invention, the tree data in the plant community survey data includes average tree height, height under branches, average breast diameter, crown width, canopy density, carbon storage, tree species, and tree quantity, wherein the tree carbon storage calculation formula is:

[0013]

[0014] C is the carbon storage of trees, n is the number of tree species, V i is the volume of the i-th tree, ρ i is the wood density of the i-th tree, B i is the biomass expansion factor of the i-th tree species.

[0015] According to a specific implementation of an embodiment of the present invention, the shrub data in the plant community survey data includes the maximum shrub height, the average shrub height, the maximum crown width, the average shrub crown width, the maximum shrub base diameter, the average shrub base diameter, the number of shrubs, and the shrub species, wherein the shrub crown area calculation formula is:

[0016]

[0017] S is the crown area of ​​the shrub, D max is the maximum crown width, D min The minimum crown width.

[0018] According to a specific implementation of an embodiment of the present invention, the herbaceous plant data in the plant community survey data includes average herb height, herb coverage, herb species, and herb cluster number, wherein the herbaceous plant community coverage is calculated using a visual estimation method, and is estimated based on the proportion of herbaceous plants covering the ground in the sample plot. Assuming that the proportion of herbaceous plants covering the ground is p, the herbaceous plant community coverage C herb =p×100%.

[0019] According to a specific implementation of an embodiment of the present invention, the topographic features extracted by the DEM digital elevation model include slope aspect, slope, and altitude, wherein the slope calculation formula is:

[0020]

[0021] α is the slope, dz is the change in elevation, and dx is the change in horizontal distance.

[0022] According to a specific implementation of an embodiment of the present invention, the meteorological monitoring data includes all-day temperature, air humidity, solid particle concentration, noise level, and negative oxygen ion concentration, wherein the relationship between air humidity and temperature satisfies a simplified form of the Clausius Clapeyron equation:

[0023]

[0024] e s is the saturated water vapor pressure, T is the temperature, and a and b are constants.

[0025] According to a specific implementation of an embodiment of the present invention, the BoxCox transformation in the data preprocessing module is used to reduce data skewness and variance heterogeneity, making the data distribution closer to the normal distribution, so as to improve the model training effect. The transformation formula is:

[0026]

[0027] x is the original data, y(λ) is the transformed data, and λ is the transformation parameter.

[0028] According to a specific implementation of an embodiment of the present invention, the feature importance evaluation index in the fusion machine learning feature screening module includes Gini importance based on random forest, split gain based on gradient boosting tree, and SHAP value. The contribution of the feature to the ecosystem service index is determined by combining these indicators, wherein:

[0029] The Gini importance calculation formula based on random forest is:

[0030]

[0031] K is the number of categories, p k is the proportion of samples belonging to the kth class to the total samples;

[0032] The calculation formula of split gain based on gradient boosting tree is:

[0033]

[0034] IG(D p ) represents the information gain of the parent node, N p is the number of parent node samples, m is the number of child nodes, N j is the number of samples of the jth child node, IG(D j ) is the information gain of the jth child node;

[0035] The calculation formula of information gain IG is:

[0036]

[0037] n k is the number of samples belonging to the kth class, and N is the total number of samples;

[0038] For the SHAP value of feature i on sample x The formula is:

[0039]

[0040] F is the feature set, M is the number of features, S is the feature subset, and v(S) is the predicted value function of subset S;

[0041] In feature selection, information entropy related formulas are introduced to assist in evaluating feature importance. Information entropy H(X) is expressed as:

[0042]

[0043] X is a random variable, x i is the value of X, p(x i ) is x i Probability of occurrence.

[0044] In a second aspect, an embodiment of the present invention further provides a device for identifying optimization factors of regulation services of an urban green space community ecosystem based on a fusion algorithm, comprising:

[0045] The collection module collects plant community survey data, including multiple indicators of trees, shrubs, and herbaceous plants and vertical lengths from water bodies, obtains DEM digital elevation model data, extracts topographic and geomorphic features of the assessed urban green space plant community points, collects meteorological monitoring data, extracts basic indicators of ecological regulation services, and aggregates the collected data to form a basic data set;

[0046] The quantification module quantifies the data in the basic data set to a range of 0-1 to form a standardized data set. At the same time, it combines the measurement model to evaluate various ecological services and quantify the ecological service supply efficiency of different green spaces;

[0047] The input module inputs the standardized data set into the random forest, gradient boosting regression tree and support vector machine machine learning models, uses the guided aggregation method to randomly sample the training data set to generate multiple sub-datasets, uses the feature importance evaluation index to quantify the feature contribution, builds a feature importance ranking list, and combines the cross-validation technology to verify the impact of the feature set on the performance of different models, eliminates redundant and small-impact features, and retains key features to form an optimized feature set;

[0048] The construction module constructs the IOFESRMLMEM model based on the linear mixed effects regression method, including a fixed effect and random effect combination framework, a linear basic model library, and a covariance matrix construction. The standardized data set after feature screening is used as input to enter the IOFESRMLMEM module. By comparing the fitting effects of multiple alternative models, the information criterion is used to select the optimal model, and the optimal model that can describe the relationship between the ecosystem service optimization factors and the target variables is output. The effects of the optimal model parameters are deeply analyzed to identify the key factors that affect ecosystem services.

[0049] The output module uses multi-dimensional performance evaluation indicators to evaluate the accuracy and generalization ability of the model, analyzes the contribution values ​​of key factors for quantitative optimization of fixed effect and random effect parameters, and outputs the optimized prediction results, key optimization factors and model performance evaluation results.

[0050] In a third aspect, an embodiment of the present invention further provides an electronic device, the electronic device comprising:

[0051] at least one processor; and,

[0052] a memory communicatively connected to the at least one processor; wherein,

[0053] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method described in any of the first aspects or any of the implementations of the first aspect.

[0054] In a fourth aspect, an embodiment of the present invention further provides a non-transitory computer-readable storage medium, which stores computer instructions, and the computer instructions are used to enable the computer to execute the method described in the first aspect or any implementation of the first aspect.

[0055] In the fifth aspect, an embodiment of the present invention further provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer executes the method described in the first aspect or any implementation of the first aspect.

[0056] The scheme in the embodiment of the present invention includes: collecting plant community survey data, including multiple indicators of trees, shrubs, and herbaceous plants and the vertical length from the water body, obtaining DEM digital elevation model data, extracting the topographic and geomorphic characteristics of the urban green space plant community points to be evaluated, collecting meteorological monitoring data, extracting basic indicators of ecological regulation services, and aggregating the collected data to form a basic data set; quantifying the data in the basic data set to a range of 0-1 to form a standardized data set, and at the same time combining the measurement model to evaluate multiple ecological services and quantify the ecological service supply efficiency of different green spaces; inputting the standardized data set into the random forest, gradient boosting regression tree and support vector machine machine learning models, using the guided aggregation method to randomly sample the training data set to generate multiple sub-data sets, using the feature importance evaluation index to quantify the feature contribution, constructing a feature importance ranking list, and combining the cross-validation technology to verify the feature importance. Collect the impact on the performance of different models, eliminate redundant and small-impact features, retain key features to form an optimized feature set; construct an IOFESRMLMEM model based on the linear mixed effects regression method, including a fixed effect and random effect combination framework, a linear basic model library, and a covariance matrix construction. Use the standardized data set after feature screening as input to enter the IOFESRMLMEM module, compare the fitting effects of multiple alternative models, use the information criterion to select the optimal model, output the optimal model that can describe the relationship between ecosystem service optimization factors and target variables, conduct in-depth analysis of the effects of the optimal model parameters, and identify the key factors that affect ecosystem services; use multi-dimensional performance evaluation indicators to evaluate the accuracy and generalization ability of the model, analyze the contribution values ​​of fixed effect and random effect parameters to quantitatively optimize key factors, and output the optimized prediction results, key optimization factors, and model performance evaluation results. This solution has the following beneficial effects:

[0057] The solution in the embodiment of the present invention has the following beneficial effects:

[0058] By collecting plant community survey data (covering multiple indicators of trees, shrubs, herbaceous plants and distances from water bodies), DEM digital elevation model data (obtaining topographic and geomorphological features) and meteorological monitoring data (obtaining basic indicators of ecological regulation services), a multi-source and comprehensive basic data set has been formed. These data reflect the ecosystem status of urban green space communities from different angles and provide rich information for subsequent precise analysis. For example, the various indicators in the plant community data can be directly related to the vegetation structure and ecological function of the green space, the elevation model data helps to understand the impact of terrain on the ecosystem, and the meteorological data is closely related to the regulation service effect.

[0059] The normalization process of quantifying the data to a range of 0 to 1 makes data of different types and magnitudes comparable, effectively avoiding model bias caused by differences in data scales. At the same time, combining the measurement model to evaluate multiple ecological services can preliminarily quantify the ecological service supply efficiency of green spaces at the data level, providing a more intuitive and comparable data basis for subsequent model analysis, and improving the data quality and availability of the entire analysis process.

[0060] The guided aggregation method is used in combination with a variety of machine learning models (random forest, gradient boosting regression tree and support vector machine) for feature screening, giving full play to the advantages of different models in feature importance assessment. By randomly sampling to generate multiple sub-datasets and training multiple times, the relationship between features and target variables can be captured more comprehensively. A variety of feature importance evaluation indicators (such as Gini importance based on random forest, split gain based on gradient boosting tree, SHAP value, etc.) are used to quantify the contribution and construct a feature importance ranking list, making the selected key features more representative and reliable. Combined with cross-validation technology, the robustness and generalization ability of feature selection are further ensured, and redundant and small-impact features are effectively eliminated. The retained key features can accurately reflect the benefits of ecosystem services, provide high-quality input for model training, and improve the prediction accuracy and explanatory power of the model.

[0061] The proposed IOFESRM LMEM model based on linear mixed effects regression (LMEMs) innovatively introduces a combination framework of fixed effects and random effects. Fixed effects can accurately describe the overall trend of ecosystem services, quantify the global contribution of each feature to the target variable, and enable the model to grasp the macro-laws of ecosystem services; random effects effectively capture the differences between individual ecosystem service units, reflect the unique impact of local features on the target variable, and thus more carefully characterize the heterogeneity of the ecosystem. The linear basic model library accurately simulates the complex relationship between optimization factors and target variables through linear regression technology, and the covariance matrix construction takes into account the correlation between feature variables and the variability of data structure, further improving the fitting accuracy and robustness of the model. This model construction method can accurately identify the key factors affecting ecosystem services and provide a solid scientific basis for optimizing the supply of ecosystem services.

[0062] The model is evaluated using multi-dimensional performance evaluation indicators (determination coefficient, mean square error, information criterion, etc.) to comprehensively measure the accuracy and generalization ability of the model from different angles. The determination coefficient reflects the degree of explanation of the target variable by the model, the mean square error evaluates the deviation between the predicted value and the true value, and the information criterion balances the model complexity and fitting accuracy. Through the comprehensive evaluation of these indicators, the performance of the model can be accurately judged, ensuring that the model has good applicability on complex ecological data, and providing a clear direction for the optimization and improvement of the model.

[0063] By deeply analyzing the fixed effect and random effect parameters and quantifying the contribution of key optimization factors, we can not only clarify the degree of influence of each factor on ecosystem services, but also reveal the internal driving mechanism of ecosystem service supply. The final output of optimization prediction results, key optimization factors and model performance evaluation results provide comprehensive and accurate data support and reliable decision-making basis for the optimization management of ecosystem services. Managers and decision makers can formulate targeted green space planning, vegetation optimization strategies and resource allocation plans based on these results, so as to achieve efficient optimization and sustainable development of urban green space community ecosystem services. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0065] Figure 1 A schematic flow chart of a method for identifying optimization factors of regulating services of an urban green space community ecosystem based on a fusion algorithm provided in an embodiment of the present invention;

[0066] Figure 2 A schematic flow chart of another method for identifying optimization factors of regulating services of an urban green space community ecosystem based on a fusion algorithm provided by an embodiment of the present invention;

[0067] Figure 3 A schematic diagram of the structure of an urban green space community ecosystem regulation service optimization factor identification device based on a fusion algorithm provided in an embodiment of the present invention;

[0068] Figure 4 A schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0069] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0070] The following describes the embodiments of the present disclosure through specific examples, and those skilled in the art can easily understand other advantages and effects of the present disclosure from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all of the embodiments. The present disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present disclosure. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present disclosure.

[0071] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein may be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on the present disclosure, it should be understood by those skilled in the art that an aspect described herein may be implemented independently of any other aspect, and two or more of these aspects may be combined in various ways. For example, any number of aspects described herein may be used to implement the device and / or practice the method. In addition, other structures and / or functionalities other than one or more of the aspects described herein may be used to implement this device and / or practice this method.

[0072] It should also be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present disclosure. The drawings only show components related to the present disclosure rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0073] Additionally, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, it will be understood by those skilled in the art that the aspects described may be practiced without these specific details.

[0074] The disclosed embodiment provides a method for identifying optimization factors of regulating services of an urban green space community ecosystem based on a fusion algorithm. The method for identifying optimization factors of regulating services of an urban green space community ecosystem based on a fusion algorithm provided in this embodiment can be executed by a computing device, which can be implemented as software, or as a combination of software and hardware, and which can be integrated in a server, a terminal device, etc.

[0075] See also Figure 1 and Figure 2The embodiment of the present disclosure provides a method for identifying optimization factors of regulating services of an urban green space community ecosystem based on a fusion algorithm, including: A method for identifying optimization factors of regulating services of an urban green space community ecosystem based on a fusion algorithm, including:

[0076] S101, collect plant community survey data, including multiple indicators of trees, shrubs, and herbaceous plants and vertical lengths from water bodies, obtain DEM digital elevation model data, extract topographic and geomorphic features of the assessed urban green space plant community points, collect meteorological monitoring data, extract basic indicators of ecological regulation services, and aggregate the collected data to form a basic data set;

[0077] S102, quantify the data in the basic data set to a range of 0-1 to form a standardized data set, and combine the measurement model to evaluate multiple ecological services and quantify the ecological service supply efficiency of different green spaces;

[0078] S103, input the standardized data set into the random forest, gradient boosting regression tree and support vector machine machine learning models, use the guided aggregation method to randomly sample the training data set to generate multiple sub-datasets, use the feature importance evaluation index to quantify the feature contribution, build a feature importance ranking list, combine the cross-validation technology to verify the impact of the feature set on the performance of different models, eliminate redundant and small-impact features, retain key features to form an optimized feature set;

[0079] S104, constructing the IOFESRMLMEM model based on the linear mixed effects regression method, including the fixed effect and random effect combination framework, the linear basic model library, and the covariance matrix construction. The standardized data set after feature screening is used as input to enter the IOFESRMLMEM module. By comparing the fitting effects of multiple alternative models, the information criterion is used to select the optimal model, and the optimal model that can describe the relationship between the ecosystem service optimization factors and the target variables is output. The effects of the optimal model parameters are deeply analyzed to identify the key factors that affect ecosystem services;

[0080] S105, uses multi-dimensional performance evaluation indicators to evaluate the accuracy and generalization ability of the model, analyzes the contribution values ​​of key factors for quantitative optimization of fixed effect and random effect parameters, and outputs the optimized prediction results, key optimization factors and model performance evaluation results.

[0081] In the process of optimizing the ecosystem services of urban green spaces, traditional evaluation methods usually rely on remote sensing data and process models. These technical applications combine the coverage rate and vegetation index of green spaces to analyze the supply and demand of ecological services of urban green spaces at different landscape scales. However, existing methods have certain limitations when dealing with complex green space structures and diverse ecological service benefits: (1) Remote sensing data cannot reflect the vegetation types and specific ecological service functions within green spaces in detail, especially when the plant communities within green spaces are diverse and complex, it is difficult to accurately identify the contributions of various types of vegetation to different ecological services; (2) Traditional process models focus on the macro-ecosystem service characteristic assessment and lack in-depth analysis of vegetation structure optimization and ecological service potential enhancement; (3) Most technical applications focus on increasing green space area or building infrastructure, ignoring how to provide ecological services more efficiently by adjusting and optimizing vegetation structure.

[0082] In addition, current technology applications have not yet provided a systematic solution to optimize the ecological service benefits of vegetation communities of different green space types. Although existing technologies have attempted to improve the fairness of ecological services by increasing green space area or infrastructure, they have failed to deeply explore the relationship between plant species and community structure and ecological benefits, and have failed to quantify the specific potential for optimizing different vegetation communities. Existing evaluation methods mainly focus on the optimization of a single ecological service, ignoring the synergistic effect of plant community structure on multiple ecological services. Therefore, it is difficult for existing technologies to effectively identify and quantify the specific impact of plant community structure on ecosystem services, and there is a lack of effective guidance for the optimization of green space vegetation.

[0083] In view of the limitations of the prior art, the present invention provides a technology for identifying ecosystem service optimization factors based on the fusion of linear mixed effects models and machine learning methods. This technology aims to improve the accuracy, scientificity and generalization ability of ecosystem service optimization factor identification by introducing a linear mixed effects regression model (LMEM) and a bootstrap aggregation (Bootstrap Aggregating) machine learning algorithm, while reducing reliance on traditional empirical judgments. When processing and analyzing complex ecosystem data, this technical solution adopts a modular process design, combined with efficient data preprocessing methods, multi-model feature screening and multi-level modeling technology, to optimize the entire process from data cleaning to model building. By comprehensively characterizing the overall trends and individual differences of ecosystem services and quantifying the impact of key optimization factors on target variables, this technology significantly improves the scientificity and practicality of ecosystem service optimization management and decision support.

[0084] like Figure 1 The overall architecture diagram of the present invention is shown, including constructing a basic data set, which includes:

[0085] ① Plant community survey data

[0086] The urban green space management unit selects plant community structure data with certain representativeness and spatial distribution in combination with the park positioning and planning and design objectives, including the average tree height, height under branches, average breast diameter, crown width, canopy density, carbon storage, tree species, and tree quantity of trees; the maximum shrub height, average shrub height, maximum crown width, average crown width, maximum shrub base diameter, average shrub base diameter, number of shrubs, and shrub species of shrubs; the average herb height, herb cover, herb species, and number of herb bushes of herbaceous plants; and the vertical distance from the water body. A total of 24 indicators.

[0087] ②DEM digital elevation model

[0088] It is obtained from the official website of the United States Geological Survey (USGS) and is used to extract the topographic and geomorphic characteristics of the assessed urban green space plant community sites, including slope aspect, slope gradient, and altitude.

[0089] ③Meteorological monitoring data

[0090] The data are obtained from the National Meteorological Information Center-China Meteorological Data Network (https: / / data.cma.cn / ) and monitoring stations of green space management units, and are used to extract basic indicators of ecological regulation services for identifying optimization factors of ecosystem regulation services in urban green space communities, including all-day temperature, air humidity, solid particle concentration, noise level, and negative oxygen ion concentration.

[0091] The ①②③ data sets are aggregated and represented by Basic data sets.

[0092] Data preprocessing module

[0093] like Figure 2 As shown in the figure, during the data preprocessing, the basic data set is processed to reduce the impact of data errors on the training effect, accuracy and generalization ability of the model. The main steps include data cleaning, conversion, standardization, and normalization processing, so as to better adapt to subsequent analysis and model training. First, in the data cleaning stage, the data integrity and accuracy are ensured by removing outliers, filling missing values ​​and unifying the format; then the Box-Cox transformation is applied to normalize the data distribution to reduce skewness and variance heterogeneity; finally, the normalization method is used to unify the data range to form a standardized data set. At the same time, the ecological service measurement is based on the basic data set, combined with the measurement model to evaluate services such as cooling and humidification effects, dust retention, noise reduction and negative oxygen ion supplementation, and by comparing the differences between the park and non-park areas, the ecological service supply efficiency of different green spaces is quantified. The data processed through the above steps are represented by Standardised datasets.

[0094] In the present invention, the Min-max scaling method is used to quantize the data indicators to a range of 0-1. The calculation formula is:

[0095]

[0096] Where X' is the normalized result, X is the original value, min(X) and max(X) are the minimum and maximum values ​​in the data set, respectively.

[0097] Fusion machine learning feature screening module

[0098] The main function of this module is to perform the next step of feature screening on Standardised datasets. The core of feature screening is based on the Bootstrap Aggregating (BA) technology in the Fusion Algorithms (FA). By building a variety of machine learning models, the impact of various ecosystem service indicators is simulated and analyzed, and finally the primary important contribution features are determined to provide key independent variable inputs for IOFESRM-LMEM model training. The main steps include inputting Standardised datasets into a variety of machine learning models (random forest, gradient boosting regression tree, support vector machine), using the bootstrap aggregation method to randomly sample the training dataset, generate multiple sub-datasets, and capture the relative importance of each feature to the target variable through repeated training. Secondly, the contribution of each feature to the ecosystem service indicator is quantified using feature importance evaluation indicators (Gini importance based on random forest, split gain based on gradient boosting tree, SHAP value), and the output of different models is combined to construct a feature importance ranking list. Thirdly, combined with cross-validation technology, the impact of the selected feature set on the performance of different models is verified to ensure the robustness and generalization ability of feature selection. Finally, redundant features and features with little impact on the target variable are eliminated, and the contribution features that best reflect the benefits of ecosystem services are retained to form an optimized feature set, which is represented by feature set.

[0099] IOFESRM model fitting module

[0100] The present invention adopts linear mixed effects models (LMEMs) to construct a method and device for identifying optimization factors of urban green space community ecosystem regulation services. The core is the IOFESRM-LMEM model constructed based on the linear mixed effects regression method (LMEM), which realizes the comprehensive quantification of complex influencing factors of ecosystem services and the accurate identification of key optimization factors. The main steps of the IOFESRM model are as follows:

[0101] 1. Data input after feature screening

[0102] First, the standardized dataset after feature screening is used as input to the IOFESRM-LMEM module. This dataset contains key feature variables that can significantly affect the core factors of ecosystem services and provide basic data support for model construction.

[0103] 2. Modular Design of Linear Mixed Effects Regression Model

[0104] In the IOFESRM-LMEM module, the model is built based on the following three important parts:

[0105] ① Fixed effect and random effect combination framework: Fixed effects and random effects are combined, with fixed effects used to describe global trends and random effects used to reflect individual differences. This combination framework can more comprehensively capture the heterogeneity and regularity of ecosystem service indicators.

[0106] ②Linear basic model library: Through linear regression technology, multiple basic models are established to simulate the linear relationship between ecosystem service optimization factors and target variables. These basic models can accurately depict the complex ecological service function mechanism.

[0107] ③ Covariance matrix construction: Based on the input data and model structure, a covariance matrix is ​​constructed to describe the correlation between feature variables and the intrinsic variability of the data structure, ensuring the stability and accuracy of the model during parameter estimation.

[0108] 3. Output the optimal model

[0109] Based on linear regression and mixed effect modeling, the IOFESRM-LMEM module compares the fitting effects of multiple candidate models and selects the optimal model using information criteria (AIC, BIC). The output optimal model can best describe the relationship between ecosystem service optimization factors and target variables.

[0110] Example of the optimal optimization factor identification model:

[0111] Cooling and humidification:

[0112] TEMP ij =-1.0454-0.2398×TC ij +0.3868×SOW ij +u j +e ij u j ~N(0,1.065)

[0113]

[0114] HUM ij=0.4947+0.4009×TC ij -0.2625×FVC ij +u j +e ij you j ~N(0,0.331)

[0115]

[0116] Noise reduction:

[0117] NOISE ij =0.2122-0.2550×NOT ij -0.2039×CBH ij +u j +e ij you j ~N(0,0.352)

[0118]

[0119] TSP ij =0.7243-2.2745×DC ij +2.7016×DMC ij +u j +e ij you j ~N(0,0.724)

[0120]

[0121] PM10 ij =1.0814+0.0410×SOW ij +2.9755×DMC ij +u j +e ij you j ~N(0,1.183)

[0122]

[0123] PM2.5 ij =1.2711+0.3305×DC ij -0.4087×MH ij +u j +e ij you j ~N(0,1.744)

[0124]

[0125] PM1 ij=1.2927+0.0921×SOW ij -0.3338×MH ij +u j +e ij u j ~N(0,1.802)

[0126]

[0127] Negative oxygen ion supply

[0128] NAI ij =0.2315×NOT ij +0.1679×SLP ij +u j +e ij u j ~N(0,0.000)

[0129]

[0130] Trying: u j is a random effect; is the random effect error; is the random effect variance.

[0131] 4. Identification of influencing parameters

[0132] After the optimal model is output, the effects of the parameters in the model are analyzed in depth, and the key factors that have a significant impact on ecosystem services are identified through parameter values, covariance matrices, and parameter significance. These factors will serve as the basis for optimizing ecosystem service functions and decision-making management.

[0133] The algorithm flow of the training module is shown in the following table:

[0134]

[0135]

[0136]

[0137] The present invention has the following beneficial effects:

[0138] (1) Basic data processing module: In the basic data processing module, the present invention solves the quality problem of multi-source heterogeneous data through a series of cleaning, normalization and normalization technologies, and improves the consistency and applicability of the data. The Box-Cox transformation is used to normalize the non-normal distribution data to eliminate skewness and variance heterogeneity. Through normalization technology, the data range is unified to ensure that the features have the same weight, providing a high-quality standardized data set for subsequent analysis, significantly reducing the impact of data errors on subsequent model training and analysis, and is the basic guarantee for building an efficient ecosystem regulation service optimization factor identification method.

[0139] (2) Key feature extraction technology based on fusion algorithm: The feature screening module uses Bootstrap Aggregating technology, combined with multiple machine learning methods such as random forest (RF), support vector machine (SVM) and gradient boosted regression tree (GBRT), to improve the accuracy and robustness of feature selection through model fusion. This module quantifies the contribution of features to ecosystem services through the importance score of model output, and uses cross-validation technology to verify the robustness of feature screening. Finally, it extracts the primary key feature set that has a significant contribution to the target variable, providing accurate input for model training and performance optimization.

[0140] (3) IOFESRM-LMEM model module: This paper proposes an IOFESRM model based on linear mixed effects regression (LMEMs), which comprehensively characterizes the overall trend and individual differences of ecosystem services by introducing fixed effects and random effects at the same time. Fixed effects are used to quantify the global contribution of optimization factors to the target variables, and random effects are used to capture individual differences between different ecological units. In addition, the introduction of the covariance matrix further describes the correlation between features and the variability of data, thereby improving the fitting accuracy and robustness of the model. This module can accurately identify the key factors affecting ecosystem services and provide a scientific basis for optimizing the supply of ecosystem services.

[0141] (4) Result evaluation and optimization module: In the model evaluation stage, the present invention uses multi-dimensional performance evaluation indicators (such as determination coefficient, mean square error, and information criterion) to verify the accuracy and generalization ability of the model, ensuring the applicability of the model to complex ecological data. At the same time, by analyzing the fixed effect and random effect parameters, the contribution value of the key optimization factors is further quantified to reveal the driving mechanism of ecosystem service supply. This module ultimately outputs the optimized prediction results, key optimization factors, and model performance evaluation results, providing accurate data support and reliable decision-making basis for the optimal management of ecosystem services.

[0142] The above four key technical points constitute the core content of the present invention. From data processing to feature screening, model construction and result evaluation, a systematic and hierarchical key technology framework is formed, which enables the present invention to provide all-round support for the identification and quantification of ecosystem service optimization factors, and improve the scientificity and practicality of ecosystem service management and optimization.

[0143] According to a specific implementation of an embodiment of the present invention, the tree data in the plant community survey data includes average tree height, height under branches, average breast diameter, crown width, canopy density, carbon storage, tree species, and tree quantity, wherein the tree carbon storage calculation formula is:

[0144]

[0145] C is the carbon storage of trees, n is the number of tree species, V i is the volume of the i-th tree, ρ i is the wood density of the i-th tree, B i is the biomass expansion factor of the i-th tree species.

[0146] According to a specific implementation of an embodiment of the present invention, the shrub data in the plant community survey data includes the maximum shrub height, the average shrub height, the maximum crown width, the average shrub crown width, the maximum shrub base diameter, the average shrub base diameter, the number of shrubs, and the shrub species, wherein the shrub crown area calculation formula is:

[0147]

[0148] S is the crown area of ​​the shrub, D max is the maximum crown width, D min The minimum crown width.

[0149] According to a specific implementation of an embodiment of the present invention, the herbaceous plant data in the plant community survey data includes average herb height, herb coverage, herb species, and herb cluster number, wherein the herbaceous plant community coverage is calculated using a visual estimation method, and is estimated based on the proportion of herbaceous plants covering the ground in the sample plot. Assuming that the proportion of herbaceous plants covering the ground is p, the herbaceous plant community coverage C herb =p×100%.

[0150] According to a specific implementation of an embodiment of the present invention, the topographic features extracted by the DEM digital elevation model include slope aspect, slope, and altitude, wherein the slope calculation formula is:

[0151]

[0152] α is the slope, dz is the change in elevation, and dx is the change in horizontal distance.

[0153] According to a specific implementation of an embodiment of the present invention, the meteorological monitoring data includes all-day temperature, air humidity, solid particle concentration, noise level, and negative oxygen ion concentration, wherein the relationship between air humidity and temperature satisfies a simplified form of the Clausius Clapeyron equation:

[0154]

[0155] e s is the saturated water vapor pressure, T is the temperature, and a and b are constants.

[0156] According to a specific implementation of an embodiment of the present invention, the BoxCox transformation in the data preprocessing module is used to reduce data skewness and variance heterogeneity, making the data distribution closer to the normal distribution, so as to improve the model training effect. The transformation formula is:

[0157]

[0158] x is the original data, y(λ) is the transformed data, and λ is the transformation parameter.

[0159] According to a specific implementation of an embodiment of the present invention, the feature importance evaluation index in the fusion machine learning feature screening module includes Gini importance based on random forest, split gain based on gradient boosting tree, and SHAP value. The contribution of the feature to the ecosystem service index is determined by combining these indicators, wherein:

[0160] The Gini importance calculation formula based on random forest is:

[0161]

[0162] K is the number of categories, p k is the proportion of samples belonging to the kth class to the total samples;

[0163] The calculation formula of split gain based on gradient boosting tree is:

[0164]

[0165] IG(D p ) represents the information gain of the parent node, N p is the number of parent node samples, m is the number of child nodes, N j is the number of samples of the jth child node, IG(D j ) is the information gain of the jth child node;

[0166] The calculation formula of information gain IG is:

[0167]

[0168] nk is the number of samples belonging to the kth class, and N is the total number of samples;

[0169] For the SHAP value of feature i on sample x The formula is:

[0170]

[0171] F is the feature set, M is the number of features, S is the feature subset, and v(S) is the predicted value function of subset S;

[0172] In feature selection, information entropy related formulas are introduced to assist in evaluating feature importance. Information entropy H(X) is expressed as:

[0173]

[0174] X is a random variable, x i is the value of X, p(x i ) is x i Probability of occurrence.

[0175] Corresponding to the above method embodiment, see Figure 3 The embodiment of the present invention further discloses a device 30 for identifying optimization factors of regulating services of an urban green space community ecosystem based on a fusion algorithm, comprising:

[0176] The collection module 301 collects plant community survey data, including multiple indicators of trees, shrubs, and herbaceous plants and vertical lengths from water bodies, obtains DEM digital elevation model data, extracts topographic and geomorphic features of the assessed urban green space plant community points, collects meteorological monitoring data, extracts basic indicators of ecological regulation services, and aggregates the collected data to form a basic data set;

[0177] The quantification module 302 quantizes the data in the basic data set to a range of 0-1 to form a standardized data set, and combines the measurement model to evaluate multiple ecological services and quantify the ecological service supply efficiency of different green spaces;

[0178] Input module 303, inputs the standardized data set into the random forest, gradient boosting regression tree and support vector machine machine learning models, uses the guided aggregation method to randomly sample the training data set to generate multiple sub-datasets, uses the feature importance evaluation index to quantify the feature contribution, constructs a feature importance ranking list, combines the cross-validation technology to verify the impact of the feature set on the performance of different models, eliminates redundant and small-impact features, and retains key features to form an optimized feature set;

[0179] Building module 304, constructing an IOFESRMLMEM model based on a linear mixed effects regression method, including a fixed effect and random effect combination framework, a linear basic model library, and a covariance matrix construction, using the standardized data set after feature screening as input to enter the IOFESRMLMEM module, comparing the fitting effects of multiple alternative models, using the information criterion to select the optimal model, outputting the optimal model that can describe the relationship between the ecosystem service optimization factors and the target variables, deeply analyzing the effects of the optimal model parameters, and identifying the key factors that affect the ecosystem services;

[0180] The output module 305 uses multi-dimensional performance evaluation indicators to evaluate the accuracy and generalization ability of the model, analyzes the contribution values ​​of the key factors of the fixed effect and random effect parameters to quantify the optimization, and outputs the optimized prediction results, key optimization factors and model performance evaluation results.

[0181] See also Figure 4 The embodiment of the present invention further provides an electronic device 60, the electronic device comprising:

[0182] at least one processor; and,

[0183] a memory communicatively connected to the at least one processor; wherein,

[0184] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method for identifying optimization factors for regulating services of urban green space community ecosystems based on a fusion algorithm in the aforementioned method embodiment.

[0185] An embodiment of the present invention further provides a non-transitory computer-readable storage medium, which stores computer instructions, and the computer instructions are used to enable the computer to execute the aforementioned method embodiment.

[0186] An embodiment of the present invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer executes the method for identifying optimization factors of urban green space community ecosystem regulation services based on a fusion algorithm in the aforementioned method embodiment.

[0187] Reference below Figure 4, which shows a schematic diagram of the structure of an electronic device 60 suitable for implementing the embodiment of the present disclosure. The electronic device in the embodiment of the present disclosure may include, but is not limited to, mobile terminals such as mobile phones, notebook computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle terminals (such as vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 4 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0188] like Figure 4 As shown, the electronic device 60 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage device 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the electronic device 60 are also stored. The processing device 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0189] Typically, the following devices may be connected to the I / O interface 605: input devices 606 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, a gage, a gyroscope, etc.; output devices 607 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 608 including, for example, a magnetic tape, a hard disk, etc.; and communication devices 609. The communication devices 609 may allow the electronic device 60 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 4 The electronic device 60 is shown with various devices, but it should be understood that it is not required to implement or have all the devices shown. More or fewer devices may be implemented or have instead.

[0190] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device 609, or installed from a storage device 608, or installed from a ROM 602. When the computer program is executed by the processing device 601, the above-mentioned functions defined in the method of the embodiment of the present disclosure are executed.

[0191] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A method for identifying optimization factors of urban green space community ecosystem regulation services based on fusion algorithm, characterized in that: include: Collect plant community survey data, including multiple indicators of trees, shrubs, and herbaceous plants and vertical lengths from water bodies, obtain DEM digital elevation model data, extract topographic and geomorphic features of the assessed urban green space plant community points, collect meteorological monitoring data, extract basic indicators of ecological regulation services, and aggregate the collected data to form a basic data set; The data in the basic data set are uniformly quantified to a range of 0-1 to form a standardized data set. At the same time, the measurement model is combined to evaluate multiple ecological services and quantify the ecological service supply efficiency of different green spaces; The standardized data set is input into the random forest, gradient boosting regression tree and support vector machine machine learning models. The training data set is randomly sampled using the guided aggregation method to generate multiple sub-datasets. The feature importance evaluation index is used to quantify the feature contribution, and a feature importance ranking list is constructed. The cross-validation technology is used to verify the impact of the feature set on the performance of different models, and redundant and small-impact features are eliminated. Key features are retained to form an optimized feature set. The IOFESRMLMEM model based on the linear mixed effects regression method was constructed, which included a fixed effect and random effect combination framework, a linear basic model library, and covariance matrix construction. The standardized data set after feature screening was used as input to enter the IOFESRMLMEM module. By comparing the fitting effects of multiple alternative models, the information criterion was used to select the optimal model, and the optimal model that can describe the relationship between the ecosystem service optimization factors and the target variables was output. The effects of the optimal model parameters were deeply analyzed to identify the key factors affecting ecosystem services. Multi-dimensional performance evaluation indicators are used to evaluate the accuracy and generalization ability of the model, and the contribution values ​​of key factors of quantitative optimization of fixed effect and random effect parameters are analyzed to output the optimized prediction results, key optimization factors and model performance evaluation results.

2. The method according to claim 1, characterized in that The tree data in the plant community survey data include average tree height, height under branches, average breast diameter, crown width, canopy density, carbon storage, tree species, and tree number. The calculation formula for tree carbon storage is: C is the carbon storage of trees, n is the number of tree species, V i is the volume of the i-th tree, ρ i is the wood density of the i-th tree, B i is the biomass expansion factor of the i-th tree species.

3. The method according to claim 2, characterized in that The shrub data in the plant community survey data include the maximum shrub height, average shrub height, maximum crown width, average shrub crown width, maximum shrub base diameter, average shrub base diameter, number of shrubs, and shrub species. The calculation formula for shrub crown area is: S is the crown area of ​​the shrub, D max is the maximum crown width, D min The minimum crown width.

4. The method according to claim 3, characterized in that The herbaceous plant data in the plant community survey data include average herb height, herb coverage, herb species, and herb cluster number. The herbaceous plant community coverage is calculated by visual estimation based on the proportion of herbaceous plants covering the ground in the sample plot. Assuming that the proportion of herbaceous plants covering the ground is p, the herbaceous plant community coverage C is herb =p×100%.

5. The method according to claim 4, characterized in that The topographic features extracted by the DEM digital elevation model include slope aspect, slope, and altitude, where the slope calculation formula is: α is the slope, dz is the change in elevation, and dx is the change in horizontal distance.

6. The method according to claim 5, characterized in that The meteorological monitoring data include all-day temperature, air humidity, solid particle concentration, noise level, and negative oxygen ion concentration, wherein the relationship between air humidity and temperature satisfies the simplified form of the Clausius-Clapeyron equation: e s is the saturated water vapor pressure, T is the temperature, and a and b are constants.

7. The method according to claim 6, characterized in that The BoxCox transformation in the data preprocessing module is used to reduce data skewness and variance heterogeneity, making the data distribution closer to the normal distribution to improve the model training effect. The transformation formula is: x is the original data, y(λ) is the transformed data, and λ is the transformation parameter.

8. The method according to claim 7, characterized in that The feature importance evaluation indicators in the fusion machine learning feature screening module include Gini importance based on random forest, split gain based on gradient boosting tree, and SHAP value. The contribution of features to ecosystem service indicators is determined by combining these indicators. The Gini importance calculation formula based on random forest is: K is the number of categories, p k is the proportion of samples belonging to the kth class to the total samples; The calculation formula of split gain based on gradient boosting tree is: IG(D p ) represents the information gain of the parent node, N p is the number of parent node samples, m is the number of child nodes, N j is the number of samples of the jth child node, IG(D j ) is the information gain of the jth child node; The calculation formula of information gain IG is: n k is the number of samples belonging to the kth class, and N is the total number of samples; For the SHAP value of feature i on sample x The formula is: F is the feature set, M is the number of features, S is the feature subset, and v(S) is the predicted value function of subset S; In feature selection, information entropy related formulas are introduced to assist in evaluating feature importance. Information entropy H(X) is expressed as: X is a random variable, x i is the value of X, p(x i ) is x i Probability of occurrence.

9. A device for identifying optimization factors of urban green space community ecosystem regulation services based on fusion algorithm, characterized in that: include: The collection module collects plant community survey data, including multiple indicators of trees, shrubs, and herbaceous plants and vertical lengths from water bodies, obtains DEM digital elevation model data, extracts topographic and geomorphic features of the assessed urban green space plant community points, collects meteorological monitoring data, extracts basic indicators of ecological regulation services, and aggregates the collected data to form a basic data set; The quantification module quantifies the data in the basic data set to a range of 0-1 to form a standardized data set. At the same time, it combines the measurement model to evaluate various ecological services and quantify the ecological service supply efficiency of different green spaces; The input module inputs the standardized data set into the random forest, gradient boosting regression tree and support vector machine machine learning models, uses the guided aggregation method to randomly sample the training data set to generate multiple sub-datasets, uses the feature importance evaluation index to quantify the feature contribution, builds a feature importance ranking list, and combines the cross-validation technology to verify the impact of the feature set on the performance of different models, eliminates redundant and small-impact features, and retains key features to form an optimized feature set; The construction module constructs the IOFESRMLMEM model based on the linear mixed effects regression method, including a fixed effect and random effect combination framework, a linear basic model library, and a covariance matrix construction. The standardized data set after feature screening is used as input to enter the IOFESRMLMEM module. By comparing the fitting effects of multiple alternative models, the information criterion is used to select the optimal model, and the optimal model that can describe the relationship between the ecosystem service optimization factors and the target variables is output. The effects of the optimal model parameters are deeply analyzed to identify the key factors that affect ecosystem services. The output module uses multi-dimensional performance evaluation indicators to evaluate the accuracy and generalization ability of the model, analyzes the contribution values ​​of key factors for quantitative optimization of fixed effect and random effect parameters, and outputs the optimized prediction results, key optimization factors and model performance evaluation results.

10. An electronic device, characterized in that: The electronic device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method for identifying optimization factors for regulating services of urban green space community ecosystems based on a fusion algorithm as described in any one of claims 1 to 8.

Images