Intelligent site selection method and system based on spatio-temporal data platform

Abstract

The invention relates to the technical field of urban planning, and particularly provides an intelligent site selection method and system based on a spatio-temporal data platform, and the method comprises the steps: collecting multi-source spatio-temporal data of key industry categories of a target region, constructing an enterprise demand feature set based on the multi-source spatio-temporal data, and building a quantitative demand standard based on the feature set; constructing an industrial ecology adaptation model based on the current industrial basic data of the target area, calculating the matching degree between the feature set and the industrial ecology of the target area through the industrial ecology adaptation model, and outputting candidate parks according to the matching degree by taking industrial ecology clustering as a target; establishing a multi-element adaptation relationship between the target enterprise and the candidate parks based on the quantitative demand standard, performing matching verification through a multi-dimensional compliance verification unit, and removing parks which do not meet the constraint condition of the quantitative demand standard; and based on the verification result, outputting a site selection scheme, finally realizing high matching of enterprise demands and public resources, improving site selection efficiency and quality, and assisting sustainable development of regional industries.

Description

Technical Field

[0001] The present invention relates to the field of urban planning technology, and in particular to a smart site selection method and system based on a spatiotemporal data platform. Background Art

[0002] With the advancement of smart city development, the application of big data and spatiotemporal analysis technologies in urban planning, industrial layout, and other fields is becoming increasingly widespread. Smart site selection, a key component of optimal resource allocation, has gradually transitioned from traditional manual decision-making to data-driven, intelligent decision-making. Existing technologies typically integrate multiple sources of information, including geospatial and economic and social data, and combine them with GIS spatial analysis, indicator weighting, and visualization to assist in site selection for commercial facilities, public service agencies, and other organizations. This has significantly improved the scientific nature and efficiency of site selection. For example, some methods construct evaluation index systems (such as accessibility, population density, and supporting facilities) and employ models such as the Analytic Hierarchy Process (AHP) and the Entropy Method to calculate comprehensive scores, thereby enabling the ranking and screening of candidate sites.

[0003] However, in the specific scenario of smart investment promotion, existing technologies still have significant limitations: On the one hand, existing site selection methods often focus on static analysis of single factors (such as plot area and transportation conditions) or are designed for general scenarios (such as commercial stores and security facilities). They fail to fully consider the need for a deep, coordinated alignment of "enterprise needs, industrial ecosystem, and spatial resources" in the context of investment attraction. This leads to a disconnect between site selection and the actual operational needs of enterprises. For example, when soliciting investment for industrial projects, existing methods often screen based solely on basic indicators such as land use type and floor area ratio, while ignoring the enterprise's industrial chain-related needs (such as the distribution radius of suppliers and customers), production process characteristics (such as energy consumption and load requirements), and the complementarity of the regional industrial ecosystem (such as the maturity of upstream and downstream supporting facilities). This leads to a disconnect between site selection and the goal of industrial cluster development. On the other hand, existing evaluation models are mostly static, relying on fixed indicator weights or single algorithms (such as genetic algorithms and comprehensive scoring). They fail to iterate and optimize based on the actual results of investment attraction in different industries. This makes it difficult to adapt to dynamic scenarios such as regional industrial evolution and changing enterprise needs, which can lead to insufficient long-term adaptability of site selection plans.

[0004] These issues make it difficult for existing technologies to achieve the goals of "precise matching, industrial collaboration, and dynamic adaptation" in smart investment promotion scenarios, leading to frequent problems such as low project implementation rates and weak industrial agglomeration effects. Therefore, there is an urgent need for a site selection method tailored to smart investment promotion scenarios that integrates the dynamic needs of enterprises, industrial ecosystem characteristics, and spatial resource elements, and provides closed-loop optimization capabilities. Summary of the Invention

[0005] In order to overcome the deficiencies of the prior art, the present application provides a wisdom site selection method and system based on a space-time data platform to solve the problems in the prior art.

[0006] One embodiment of the present application provides a wisdom site selection method based on a space-time data platform, comprising the following steps: Collecting multi-source space-time data of target area key industry categories, constructing an enterprise demand feature set based on the multi-source space-time data, the enterprise demand feature set at least including industry attributes, fixed asset investment, industry chain associated demand and production process demand; establishing a quantitative demand standard based on the enterprise demand feature set, the quantitative demand standard at least including land space scale, building function parameters, municipal and supporting element threshold values; Based on the current industrial base data of the target area, an industrial ecological adaptation model is constructed, and the matching degree of the enterprise demand feature set and the target area industrial ecology is calculated through the industrial ecological adaptation model, the matching degree including industry chain correlation degree, industry cluster fit degree and supporting facility completeness; the industrial ecological adaptation model takes industrial ecological clustering as the goal, and outputs a candidate park according to the matching degree; Based on the quantitative demand standard of the enterprise demand feature set, a multi-element adaptation relationship of land space scale, building function parameters of the target enterprise and the candidate park is established; the municipal and supporting element threshold values and restrictive elements of the candidate park are matched and verified by a multi-dimensional compliance verification unit, and the park that does not meet the constraint conditions of the quantitative demand standard is eliminated; Based on the verification result, an site selection scheme is output, the site selection scheme including recommended park entry ranking, adapted industrial land and industrial housing, and containing customized resource matching suggestions and restrictive element risk response strategies matched with the enterprise demand feature set; According to the feedback data of the target area key industry categories, the quantitative demand standard and the industrial ecological adaptation model parameters are dynamically updated.

[0007] By adopting the above scheme, the deep fusion and efficient utilization of multi-source spatio-temporal data can be achieved, the enterprise demand feature set covering industrial attributes, fixed asset investment, industrial chain related demand and production process demand can be accurately constructed, and the quantitative standards such as land space scale, building function parameters, municipal and supporting element threshold values are converted, thereby providing a solid data foundation and clear measurement benchmark for subsequent processes. The industrial ecology adaptation model is oriented to industrial ecology clustering, and through calculating the industrial chain correlation degree, industrial agglomeration fit degree and supporting facility completeness of the enterprise demand feature set and the target regional industrial ecology, the deep collaborative matching of “enterprise demand-industrial ecology-space resources” is realized, the problem of disconnection between the site selection scheme and the industrial clustering target in the prior art is effectively solved, and the regional industrial collaborative development and clustering construction are effectively promoted. The multi-dimensional compliance verification unit strictly verifies the candidate parks based on the quantitative demand standards, accurately excludes the parks that do not meet the conditions, and greatly improves the feasibility and compliance of site selection. The output site selection scheme covers recommended parks, adapted industrial land and industrial buildings, and is matched with customized resource matching suggestions and restrictive element risk response strategies, which can optimize resource allocation and reduce potential risks. At the same time, through the feedback data of the target regional key industry categories, the quantitative demand standards and model parameters are dynamically updated, so that the site selection method has good adaptability and iteration, effectively overcomes the poor long-term adaptability problem in the prior art, and finally realizes the high matching of enterprise demand and public resources, significantly improves the site selection efficiency and quality, and helps the sustainable development of regional industry.

[0008] In one of the embodiments, before the step of outputting the site selection scheme based on the verification result, the following steps are further included: A virtual simulation scene of the target region is constructed based on digital twin technology, the verified candidate park data, the enterprise demand feature set and the current industrial foundation data of the target region are input into the virtual simulation scene, and the industrial driving effect, economic benefit and supporting measure input-output ratio of the target enterprise in the preset period after entering the candidate park are simulated and simulated. The industrial driving effect includes the number of upstream and downstream enterprise agglomeration and the improvement range of industrial chain synergy efficiency, the economic benefit includes the annual tax contribution value and the number of newly added employment positions, and the supporting measure input-output ratio is the ratio of park infrastructure input to enterprise expected output value. Based on the simulation result, the priority of the candidate park is sorted, the weighted comprehensive score of the industrial driving effect, economic benefit and supporting measure input-output ratio of each candidate park in the preset period of simulation is calculated, and the candidate park with the optimal comprehensive score is selected.

[0009] By adopting the above solution and leveraging digital twin technology to construct a virtual simulation scenario, the simulation analysis incorporates verified candidate park data, enterprise demand feature sets, and the target region's current industrial infrastructure data. This allows for accurate simulation of the industry-driven effects, economic benefits, and input-output ratio of supporting measures within a preset period after the target enterprise's arrival, transforming the long-term impact of industrial development from "empirical prediction" to "data-driven quantitative assessment." By calculating the weighted comprehensive score of each candidate park within a preset period and prioritizing them, the park with the best overall benefits can be selected, further enhancing the site selection plan's in-depth consideration of long-term industrial synergy, economic benefits, and resource input efficiency. This effectively avoids the problem of ignoring long-term development adaptability due to short-term matching, making the final output site selection plan more aligned with the region's industrial sustainable development goals and enhancing the foresight and scientific nature of smart investment site selection.

[0010] In one embodiment, the step of collecting multi-source spatiotemporal data of key industries in the target area and constructing an enterprise demand feature set based on the multi-source spatiotemporal data further includes: Constructing a community feature set for the target area, wherein the community feature set includes at least community population structure data, community public service carrying capacity data, and community environmental sensitive factor data; The step of calculating the matching degree between the enterprise demand feature set and the target region's industrial ecology through the industrial ecology adaptation model specifically includes: Calculating the first matching degree between the enterprise demand feature set and the industrial ecology of the target region, where the first matching degree is determined based on the correlation degree of the industrial chain, the degree of industrial agglomeration fit, and the completeness of supporting facilities; Calculating a second degree of matching between the enterprise demand feature set and the community feature set, where the second degree of matching is determined based on the compatibility of the enterprise's talent needs with the community's demographic structure data, the incremental demand for community public service carrying capacity data from the enterprise's operations, and the compatibility of the enterprise's environmental impact with the community's environmental sensitive factor data; The first matching degree and the second matching degree are weightedly integrated to obtain the final matching degree between the enterprise and the target area, which serves as the basis for outputting candidate parks from the industrial ecological adaptation model.

[0011] By adopting the above solution, a community feature set and a dual-dimensional matching mechanism based on "industrial ecology and community characteristics" are constructed. This ensures that the matching degree calculation not only covers the correlation and adaptation between enterprises and the regional industrial ecology, but also incorporates the compatibility assessment between enterprises and the community's population structure, public service carrying capacity, and environmentally sensitive factors. The weighted fusion of the first and second matching degrees not only reflects the economic value of industrial synergy, but also takes into account the social adaptability and environmental compliance at the community level. This effectively avoids the situation where a focus on industrial matching neglects the potential impact of enterprise operations on the community (such as environmental conflicts and public service overload) or the insufficient support of community resources for enterprises (such as an imbalance in talent supply). The final candidate parks achieve a balance between industrial synergy and harmonious community development, further enhancing the consideration of the comprehensive "economic, social, and environmental" benefits of smart investment and site selection, and improving the sustainability and social recognition of the site selection results.

[0012] In one embodiment, the step of establishing a multi-factor adaptation relationship between the target enterprise and the land use space scale and building function parameters of the candidate park based on the quantitative demand standards of the enterprise demand feature set specifically includes: Constructing a multi-factor quantitative adaptation model, wherein the multi-factor quantitative adaptation model establishes a quantitative mapping relationship between the land use spatial scale parameters and building function parameters of the candidate park and the enterprise demand feature set based on the quantitative demand standards of the enterprise demand feature set; Determining a multi-dimensional adaptation index of the quantitative mapping relationship, wherein the multi-dimensional adaptation index includes at least a land space adaptation dimension and a building function adaptation dimension, and each dimensional adaptation index includes at least one quantitative index parameter and a corresponding weight coefficient; Standardize each quantitative indicator parameter to generate comparable quantitative values, including but not limited to calculation of deviation rate, degree of fit, and compliance rate; Based on the weight coefficient, the quantitative values ​​of each dimension are weighted and aggregated to generate a comprehensive fitness score of multiple factors; Set a suitability threshold, and include candidate parks whose comprehensive suitability scores reach the suitability threshold into the subsequent verification process, and eliminate candidate parks that do not reach the suitability threshold.

[0013] By adopting the above solution, the multi-factor quantitative adaptation model transforms the enterprise demand feature set and candidate park parameters into a computable quantitative mapping relationship, solving the problem of insufficient quantitative accuracy of the multi-factor adaptation relationship. The multi-dimensional adaptation indicator system covers the land use space and building function dimensions, and the weight coefficients reflect differences in industry priorities, enhancing the adaptability of the model. Standardization processing unifies the indicator calculation benchmark to improve the objectivity of the evaluation; weighted aggregation scoring achieves scientific integration of multiple dimensions and provides accurate quantitative evaluation; threshold screening reduces subsequent verification redundancy and optimizes process efficiency. This model transforms empirical judgments into data-driven decision-making, significantly improving the accuracy and efficiency of "enterprise demand-park resource" matching, providing key support for the core solution and enhancing the scientific nature and reliability of smart investment promotion and site selection.

[0014] In one embodiment, the step of constructing an industrial ecology adaptation model based on the current industrial base data of the target area and calculating the matching degree between the enterprise demand feature set and the industrial ecology of the target area through the industrial ecology adaptation model specifically includes the following steps: Collect enterprise demand feature sets and the current industrial infrastructure data of the target area for preprocessing to obtain standardized feature data that can be used for machine learning model training; The standardized feature data is used to train an industrial ecological adaptation model through a machine learning algorithm to obtain a trained industrial ecological adaptation model, and the trained industrial ecological adaptation model is used to: Real-time calculation of the matching degree between the enterprise demand feature set and the target region's industrial ecology; With the goal of industrial ecological clustering, the candidate parks are ranked according to the matching degree.

[0015] By adopting the above solution, the collected enterprise demand feature set and the current industrial base data of the target area are preprocessed to make the data suitable for the input requirements of the machine learning model; the preprocessed standardized feature data is used to train the industrial ecology adaptation model using the machine learning algorithm, so that the obtained model can subsequently output the matching degree between the enterprise and the industrial ecology based on the input real-time data, and output the candidate park ranking results with the goal of industrial ecology clustering, providing a quantitative basis for accurate site selection.

[0016] In one embodiment, the step of collecting enterprise demand feature sets and preprocessing the current industrial base data of the target area to obtain standardized feature data that can be used for machine learning model training specifically includes: Collect the enterprise demand feature set and the current industrial base data of the target area, wherein the current industrial base data includes industrial chain structure data, industrial cluster distribution data and supporting facilities data; Clean the collected data, including filling in missing values, correcting outliers and unifying data formats; Based on the correlation between the enterprise demand feature set and the target region's industrial ecology, feature mapping is performed on the cleaned data to generate an initial feature set that includes industry chain correlation features, industrial agglomeration features, and supporting features. The initial feature set is standardized, and each feature value is normalized to a preset value range to obtain standardized feature data that can be directly used for machine learning model training.

[0017] By adopting the above solution, the collected enterprise demand feature set and the current industrial basic data of the target area are cleaned, feature mapped and standardized, so that the data can be de-noised and correlation relationships can be established, and standardized feature data suitable for machine learning model training can be obtained; feature mapping clarifies the correspondence between enterprise needs and industrial ecology, and standardization unifies the data benchmark, avoiding model input deviations caused by data confusion and ensuring the effectiveness of model training data.

[0018] In one embodiment, the step of training the industrial ecological adaptation model with the standardized feature data through a machine learning algorithm to obtain a trained industrial ecological adaptation model specifically includes: Dividing the standardized feature data into a training set and a validation set; An initial industrial ecological adaptation model is trained based on the training set. The initial industrial ecological adaptation model uses the industrial chain association dimension, the industrial agglomeration fit dimension, and the supporting support dimension as core evaluation dimensions, and sets corresponding quantitative indicators and weight coefficients for each dimension; Calculate the comprehensive matching score of the training set using the initial industrial ecosystem adaptation model. The score is the weighted sum of the standardized quantitative indicators of each dimension. Use the validation set to verify the model performance, requiring the matching accuracy of the core evaluation dimensions to be no less than a preset threshold, and optimize the weight coefficients and indicator parameters of the industrial ecosystem adaptation model based on the validation results; A trained industrial ecological adaptation model is generated based on the optimized weight coefficients and indicator parameters. The industrial ecological adaptation model is configured as follows: Receive real-time enterprise demand feature sets and target region industrial ecosystem data, and output a comprehensive matching score between the two; With the goal of industrial ecological clustering, the comprehensive matching scores are subjected to a secondary weighted calculation to generate a ranking result of candidate parks, among which parks with a high degree of correlation with existing industrial clusters receive additional weight bonuses.

[0019] By adopting the above scheme, the standardized feature data is divided into a training set and a validation set. The training set is used to train the initial industrial ecological adaptation model and calculate the comprehensive matching score. The test set is used to verify the model performance and optimize the parameters so that the matching accuracy of the model meets the preset threshold. The optimized model can receive data and output the comprehensive matching degree in real time, and highlight the correlation of industrial clusters through secondary weighting, so that the ranking results of candidate parks meet the quantitative matching requirements and fit the goal of industrial ecological clustering, thereby improving the accuracy and applicability of site selection matching.

[0020] This application also relates to a smart site selection system based on a spatiotemporal data platform, including: A multi-source data acquisition module is used to collect multi-source spatiotemporal data of key industries in the target area, construct an enterprise demand feature set based on the multi-source spatiotemporal data, and the enterprise demand feature set includes at least industry attributes, fixed asset investment amount, industry chain related demand and production process requirements; and establish quantitative demand standards based on the enterprise demand feature set, and the quantitative demand standards include at least land use spatial scale, building function parameters, and municipal and supporting element thresholds. The candidate park generation module is used to construct an industrial ecosystem adaptation model based on the current industrial infrastructure data of the target area. The industrial ecosystem adaptation model is used to calculate the matching degree between the enterprise demand feature set and the industrial ecosystem of the target area. The matching degree includes the industrial chain correlation, industrial agglomeration fit, and supporting facilities completeness. The industrial ecosystem adaptation model aims to cluster the industrial ecosystem and outputs candidate parks based on the matching degree. The candidate park screening module is used to establish a multi-factor adaptation relationship between the target enterprise and the candidate park's land use spatial scale and building function parameters based on the quantitative demand standards of the enterprise demand feature set; through a multi-dimensional compliance verification unit, the matching verification of the municipal and supporting factor thresholds and restrictive factors of the candidate park is carried out, and parks that do not meet the constraints of the quantitative demand standards are eliminated; A site selection plan generation module is used to output a site selection plan based on the verification results. The site selection plan includes a recommended ranking of the park to be settled, suitable industrial land and industrial buildings, and includes customized resource matching suggestions and restrictive factor risk response strategies that match the enterprise's demand feature set; The dynamic optimization module is used to dynamically update the quantitative demand standards and industrial ecological adaptation model parameters based on the feedback data of key industries in the target area.

[0021] By implementing this approach, we can achieve deep integration and efficient utilization of multi-source spatiotemporal data, accurately constructing a set of enterprise demand characteristics encompassing industrial attributes, fixed asset investment, industrial chain-linked demands, and production process requirements. These characteristics are then converted into quantitative criteria, such as land use spatial scale, building function parameters, and thresholds for municipal and supporting elements, providing a solid data foundation and clear benchmarks for subsequent processes. The industrial ecosystem adaptation model, guided by industrial ecosystem clustering, calculates the degree of industrial chain relevance, industrial agglomeration fit, and the completeness of supporting facilities within the target region's industrial ecosystem. This achieves a deep, coordinated alignment between enterprise demand, industrial ecosystem, and spatial resources, effectively resolving the disconnect between existing site selection methods and industrial clustering goals, and effectively promoting regional industrial collaborative development and clustering. A multi-dimensional compliance verification unit rigorously verifies candidate parks based on quantitative demand criteria, accurately eliminating those that do not meet the requirements, significantly enhancing the feasibility and compliance of site selection. The resulting site selection plan includes recommended entry parks, suitable industrial land, and industrial buildings, along with customized resource matching recommendations and risk mitigation strategies for restrictive factors, to optimize resource allocation and mitigate potential risks. At the same time, by dynamically updating the quantitative demand standards and model parameters through feedback data from key industries in the target area, the site selection method has good adaptability and iterativeness, effectively overcoming the problem of poor long-term adaptability of solutions in existing technologies, and ultimately achieving a high degree of matching between corporate needs and public resources, significantly improving site selection efficiency and quality, and contributing to the sustainable development of regional industries.

[0022] The present application also relates to a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the above-mentioned intelligent site selection method based on a spatiotemporal data platform are implemented.

[0023] The present application also relates to a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the steps of the above-mentioned intelligent site selection method based on a spatiotemporal data platform.

[0024] The above embodiments provide a smart site selection method and system based on a spatiotemporal data platform, which has the following beneficial effects: It enables the deep integration and efficient utilization of multi-source spatiotemporal data, accurately constructing a set of enterprise demand characteristics encompassing industrial attributes, fixed asset investment, industrial chain-related demands, and production process requirements. These characteristics are then converted into quantitative criteria, such as land use spatial scale, building function parameters, and thresholds for municipal and supporting elements, providing a solid data foundation and clear benchmarks for subsequent processes. The industrial ecosystem adaptation model, guided by industrial ecosystem clustering, calculates the degree of industrial chain relevance, industrial agglomeration fit, and the completeness of supporting facilities within the target region's industrial ecosystem. This model achieves a deep, coordinated alignment between enterprise demand, industrial ecosystem, and spatial resources, effectively resolving the disconnect between existing site selection solutions and industrial clustering objectives, and effectively promoting regional industrial collaborative development and clustering. A multi-dimensional compliance verification unit rigorously verifies candidate parks based on quantitative demand criteria, accurately eliminating those that do not meet the requirements, significantly enhancing the feasibility and compliance of site selection. The resulting site selection plan includes recommended entry parks, suitable industrial land, and industrial buildings, along with customized resource matching recommendations and risk mitigation strategies for restrictive factors, to optimize resource allocation and mitigate potential risks. At the same time, by dynamically updating the quantitative demand standards and model parameters through feedback data from key industries in the target area, the site selection method has good adaptability and iterativeness, effectively overcoming the problem of poor long-term adaptability of solutions in existing technologies, and ultimately achieving a high degree of matching between corporate needs and public resources, significantly improving site selection efficiency and quality, and contributing to the sustainable development of regional industries. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A flow chart of a smart site selection method based on a spatiotemporal data platform provided in an embodiment of the present invention; Figure 2 A block diagram of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0026] The technical solutions in the embodiments of the present invention will be described clearly and completely below with reference to the accompanying drawings.

[0027] Reference Figure 1 One embodiment of the present invention provides a smart site selection method based on a spatiotemporal data platform, comprising the following steps: S10. Collect multi-source spatiotemporal data of key industries in the target area, and construct an enterprise demand feature set based on the multi-source spatiotemporal data. The enterprise demand feature set includes at least industry attributes, fixed asset investment amount, industry chain-related demands, and production process demands; establish quantitative demand standards based on the enterprise demand feature set. The quantitative demand standards include at least land use spatial scale, building function parameters, and municipal and supporting element thresholds.

[0028] In this embodiment, the target area is the specific geographical area where the investment attraction site selection assessment is planned. Key industry categories are industry categories within the target area that are identified as requiring special attention, analysis, or research based on factors such as regional development positioning, economic pillar status, strategic planning goals, or policy orientation. For example, if the target area is a new energy industry base, its key industry categories may include photovoltaic manufacturing, energy storage battery research and development, and wind power equipment production; if the target area is a coastal economic zone, its key industry categories may include port logistics, marine engineering, and shipbuilding.

[0029] For multi-source spatiotemporal data, industry attribute-related data can be obtained through industry development white papers and key industry lists published by industry regulatory authorities; fixed asset investment amounts can be referenced from enterprise project filing documents, investment agreements, and project database information from investment promotion departments; industry chain-related demand data can be derived from industry cluster map reports, enterprise cooperation agreements, and transaction records on supply chain management platforms; production process requirements can be based on industry technical standard manuals, enterprise production specifications, and equipment supplier technical parameter specifications. These data can be obtained by technicians in this field through the internet through channels such as corresponding official websites, enterprise websites, and industry platforms.

[0030] When constructing an enterprise demand feature set based on multi-source spatiotemporal data, the collected original data needs to be screened and refined during the construction process. Extract core features such as the industry category and development stage of the industry from the data related to industrial attributes; summarize and calculate the fixed asset investment data to clarify the total amount of funds and the proportion of each use; sort out the industrial chain-related demand data to determine the specific collaborative relationship with upstream and downstream enterprises and supporting enterprises; analyze the production process demand data to obtain key information such as the technical level and equipment specifications required for production, and finally form an enterprise demand feature set that includes industrial attributes, fixed asset investment, industrial chain-related demands and production process demands. These features are the core elements of constructing an enterprise demand feature set, aiming to accurately portray enterprise characteristics from the perspective of the enterprise's own attributes and operational needs, and provide a basic basis for subsequent industrial collaborative matching and resource adaptation. The specific functions are: Industrial attributes: Clarifying the sub-industry field to which the enterprise belongs (such as automotive parts manufacturing, biopharmaceutical research and development) is the basis for judging whether the target enterprise is compatible with the key industries of the target region (for example, if the region focuses on the development of new energy industries, new energy battery research and development companies will be given priority), ensuring that the investment project is consistent with the regional industrial positioning.

[0031] Fixed asset investment amount: reflects the scale and investment intensity of the enterprise (e.g., fixed asset investment ≥ 500 million yuan). It is used to screen enterprises that match the regional carrying capacity (for example, industrial parks can use this indicator to determine whether they have sufficient land and capital supporting capabilities), thereby avoiding implementation difficulties caused by the mismatch between enterprise scale and regional resources.

[0032] Industrial chain linkage requirements: This reflects the company's position in the industrial chain and its dependence on upstream and downstream suppliers (e.g., core suppliers must be located within a 30-kilometer radius and close to logistics hubs). It is the key to calculating the degree of industrial synergy matching (for example, if a region has formed an automobile manufacturing cluster, parts companies with this feature will be prioritized due to the complementarity of the industrial chain), supporting the goal of industrial ecological clustering.

[0033] Production process requirements: These cover special requirements during the company's production process (such as cleanroom grade, high-temperature-resistant workshops, and annual electricity consumption ≥ 10 million kWh). These requirements are directly related to subsequent space and resource adaptation (for example, the explosion-proof process requirements of chemical companies must match the safety facility standards of the park), ensuring that the basic conditions for the company's production operations are met.

[0034] For the establishment of quantitative demand standards based on the enterprise demand feature set, the quantification process needs to combine the actual operational needs of the enterprise with the general industry standards. The land use spatial scale is determined based on the production scale, equipment floor area and future development plan of the enterprise, with reference to the land use situation of similar enterprises in the same industry. The building function parameters are determined based on the special requirements of the production process for the factory, such as floor height, load-bearing capacity, fire protection level, etc., and the specific parameter values ​​are determined by comparing with the building design specifications; the municipal and supporting element thresholds comprehensively consider the enterprise's production water, electricity, and gas needs, as well as the requirements for supporting conditions such as transportation and logistics, and combine the municipal facilities planning standards to determine various thresholds, such as the water supply capacity of tons per hour and the power supply capacity of kilovolt-amperes. For the land use spatial scale, building function parameters, municipal and supporting element thresholds, these characteristics are the core content of the quantitative demand standards, which aim to convert enterprise needs into verifiable and implementable space and resource indicators, and provide a rigid basis for the screening and verification of candidate parks. The specific functions are: Land spatial scale: including plot area, shape, floor area ratio, etc. (e.g., land area ≥ 50 mu, floor area ratio ≤ 1.2), used to match the space occupation needs of enterprises (for example, large equipment manufacturing enterprises require large plots of land, while scientific and technological innovation enterprises may focus more on the regularity of the plot shape to adapt to the layout of the R&D building), ensuring that the enterprise has sufficient physical space to carry out production.

[0035] Building function parameters: functional indicators involving factories, office buildings and other buildings (such as factory floor height ≥8 meters, load-bearing capacity ≥5 tons / square meter, and R&D building floor thickness ≥0.2 meters), used to adapt to the company's production process or office needs (for example, heavy machinery companies require high-load-bearing factories, and biopharmaceutical companies require sterile clean workshops) to avoid production restrictions due to substandard building conditions.

[0036] Municipal and supporting element thresholds: These quantitative standards cover basic infrastructure such as water, electricity, gas, and transportation (e.g., power supply capacity ≥ 2000 kVA, distance to a highway entrance ≤ 5 km, sewage treatment capacity ≥ 500 tons / day). These standards are used to determine whether the park's infrastructure can support business operations (for example, a data center's high requirements for power stability must match the park's dual-circuit power grid threshold), ensuring continued stable operation after the business goes into production.

[0037] It should be noted that in the process of collecting and processing multi-source spatiotemporal data, constructing enterprise demand feature sets and establishing quantitative demand standards, the operations involved, such as data screening, calculation, and analysis, are all conventional means in the field of data processing. Those skilled in the art can use appropriate methods to complete them according to actual conditions, so they will not be elaborated here.

[0038] S20. Based on the current industrial infrastructure data of the target area, an industrial ecological adaptation model is constructed. The matching degree between the enterprise demand feature set and the industrial ecology of the target area is calculated through the industrial ecological adaptation model. The matching degree includes the correlation degree of the industrial chain, the industrial agglomeration fit and the completeness of supporting facilities. The industrial ecological adaptation model aims at industrial ecological clustering and outputs candidate parks according to the matching degree.

[0039] The construction of the industrial ecological adaptation model specifically includes the following steps: S210: Collect and pre-process the enterprise demand feature set and the current industrial base data of the target area to obtain standardized feature data that can be used for machine learning model training. Specifically, the following steps are included: S211. Collect the enterprise demand feature set and the current industrial base data of the target area, where the current industrial base data includes industrial chain structure data, industrial cluster distribution data, and supporting facilities data; S212. Clean the collected data, including filling in missing values, correcting outliers, and unifying data formats; S213. Based on the correlation between the enterprise demand feature set and the target region's industrial ecology, feature mapping is performed on the cleaned data to generate an initial feature set including industry chain correlation features, industrial agglomeration features, and supporting features; S214: Standardize the initial feature set, normalize each feature value to a preset value range, and obtain standardized feature data that can be directly used for machine learning model training.

[0040] S220: Using the standardized feature data to train an industrial ecosystem adaptation model through a machine learning algorithm to obtain a trained industrial ecosystem adaptation model. The trained industrial ecosystem adaptation model is used to: calculate the matching degree between the enterprise demand feature set and the target area industrial ecosystem in real time; and output a ranking result of candidate parks based on the matching degree, with the goal of industrial ecosystem clustering. Specifically, the following steps are included: S221, dividing the standardized feature data into a training set and a validation set; S222. Training an initial industrial ecosystem adaptation model based on the training set. The initial industrial ecosystem adaptation model uses the industrial chain association dimension, the industrial cluster fit dimension, and the supporting dimension as core evaluation dimensions, and sets corresponding quantitative indicators and weight coefficients for each dimension. S223. Calculate the comprehensive matching score of the training set using the initial industrial ecosystem adaptation model. The score is the weighted sum of the standardized quantitative indicators of each dimension. S224. Verify the model performance using the validation set, requiring the matching accuracy of the core evaluation dimension to be no less than a preset threshold, and optimize the weight coefficient and indicator parameters of the industrial ecosystem adaptation model based on the validation results; S225. Generate a trained industrial ecological adaptation model based on the optimized weight coefficients and indicator parameters. The industrial ecological adaptation model is configured as follows: Receive real-time enterprise demand feature sets and target region industrial ecosystem data, and output a comprehensive matching score between the two; With the goal of industrial ecological clustering, the comprehensive matching scores are subjected to a secondary weighted calculation to generate a ranking result of candidate parks, among which parks with a high degree of correlation with existing industrial clusters receive additional weight bonuses.

[0041] In this embodiment, as described in step S20 above, an industrial ecosystem adaptation model is constructed based on the current industrial infrastructure data of the target region. The core of the model is to quantify the degree of match between enterprise needs and the regional industrial ecosystem through standardized data processing and machine learning training, such as the XGBoost algorithm, and ultimately screen out candidate parks that meet the goal of industrial clustering. The specific implementation process is as follows: (1) Data preprocessing and feature engineering (i.e., step S210) In step S211, the collected enterprise demand feature set and the current industrial base data of the target area need to form a complete data chain: the enterprise demand feature set is the structured data constructed in step S10, including industry attributes, fixed asset investment, etc. Current industrial infrastructure data encompasses three core categories: industrial chain structure data (such as the number of upstream and downstream enterprises within the region, the frequency of collaboration, and the reach of core enterprises); industrial cluster distribution data (such as the dominant industry types, enterprise density, and annual output value of each cluster); and supporting facilities data (such as water, electricity, and gas capacity within the park, throughput at surrounding logistics hubs, and the number of scientific research institutions). This data can be obtained through regional industrial big data platforms, park management systems, and statistical yearbooks of relevant departments. Technical personnel can retrieve it through corresponding official websites or government data interfaces.

[0042] In step S212, the collected data is cleaned: Missing values ​​are filled using industry mean interpolation (e.g., if the "gas capacity" in a park's supporting facilities data is missing, it is filled with the average gas capacity of parks of the same type) or by extrapolating data trends from adjacent years; Outlier correction is identified through the 3σ criterion (for example, when a company's operating indicator far exceeds 3 times the standard deviation of its peers, it is marked as an outlier and replaced with the 95th percentile of the industry); The data format is unified into a structured table (for example, the "collaboration frequency" is converted from the text description "5 times per month" to the value "5", and the time format is unified into "YYYY-MM").

[0043] In step S213, feature mapping is performed based on the association relationship: Industrial chain linkage characteristics: Mapping enterprise industrial chain linkage requirements (e.g., "parts suppliers must be within 5 kilometers") with regional industrial chain structure data (e.g., "number of parts companies within 5 kilometers of the industrial park") into characteristics such as "supply-demand matching degree" and "collaboration distance compliance rate"; Industrial agglomeration characteristics: Mapping enterprise industry attributes (such as "new energy battery R&D") and regional industrial cluster distribution data (such as "the proportion of new energy industry enterprises in the park") into characteristics such as "industry type consistency" and "cluster scale adaptability"; Supporting features: The thresholds of enterprise municipal and supporting factors (such as "power supply capacity ≥ 1500KVA") and park supporting facilities data (such as "actual power supply capacity of the park 2000KVA") are mapped to features such as "facility compliance rate" and "resource redundancy", ultimately forming the initial feature set.

[0044] In step S214, the standardization process uses min-max normalization: each feature value in the initial feature set is mapped to the interval [0,1] (such as the original value of 50% of the "collaboration distance compliance rate" is normalized to 0.5, and the original value of 80% of the "industry type matching degree" is normalized to 0.8), eliminating dimensional differences and obtaining standardized feature data that can be directly input into the machine learning model.

[0045] (2) Industrial Ecosystem Adaptation Model Training and Optimization (i.e., Step S220) In step S221, the standardized feature data is divided into a training set (for model parameter learning) and a validation set (for model performance evaluation) in a ratio of 7:3, and the data distribution consistency is maintained during the division (for example, the proportion of each industry type in the training set and the validation set is consistent).

[0046] In step S222, the initial industrial ecosystem adaptation model adopts a multi-dimensional weighted evaluation framework: Industrial chain linkage dimension: Quantitative indicators include "number of upstream and downstream enterprise matches" and "collaboration frequency compliance rate", with a weight coefficient of 0.4 (focusing on industrial collaboration); Industrial agglomeration fit dimension: Quantitative indicators include "density of similar enterprises" and "industry scale fit", with a weight coefficient of 0.3 (focusing on cluster integration); Support dimension: Quantitative indicators include "facility threshold compliance rate" and "resource redundancy", and the weight coefficient is set to 0.3 (focusing on basic guarantees).

[0047] The initial weights of each dimensional indicator were determined through the Analytic Hierarchy Process (AHP) to ensure relevance to the goal of industrial clustering. When constructing the initial industrial ecosystem adaptation model, the XGBoost algorithm was simultaneously introduced. This algorithm uses standardized feature data (such as the "collaboration distance compliance rate" for industrial chain linkage characteristics and the "density of similar enterprises" for industrial agglomeration characteristics) as input features. The importance of each feature is calculated using decision tree splitting rules, and the initial weight coefficients are automatically iteratively optimized (for example, gradually increasing the importance of the "number of upstream and downstream enterprise matches" feature from 0.4 to 0.45) to enhance the model's ability to fit actual industrial ecosystem relationships.

[0048] In step S223, the calculation of the comprehensive matching score relies on the ensemble learning properties of the XGBoost algorithm: the algorithm feeds quantitative indicators for each dimension (such as the "number of matching upstream and downstream enterprises" for the industrial chain correlation dimension and the "density of similar enterprises" for the industrial agglomeration dimension) into 100 decision trees for parallel calculation. Each decision tree outputs a local matching score for that dimension based on the feature splitting rule (for example, a decision tree predicts a score of 0.85 for the "collaboration frequency compliance rate"). Finally, the algorithm's output layer performs a weighted fusion of the local matching scores of all decision trees (the fusion weight is negatively correlated with the tree's training error; trees with smaller errors have higher weights) to generate a comprehensive matching score for the training set. For example, after the algorithm's calculations, the weighted sum of the local matching scores for a certain park's industrial chain correlation dimension is 0.32, 0.21 for the industrial agglomeration dimension, and 0.18 for the supporting dimension, resulting in a comprehensive matching score of 0.71.

[0049] In step S224, model validation requires a matching accuracy of at least 85% for the core assessment dimensions (e.g., the match between the "collaboration feasibility" predicted by the industry chain correlation dimension and the actual collaboration records must be ≥85%). If validation fails, the weight coefficients are optimized using a grid search method (e.g., adjusting the weight of the industry chain correlation dimension to 0.45) and the indicator parameters are modified (e.g., expanding the calculation scope of the "supporting facilities compliance rate").

[0050] In step S224, the model performance verification and optimization process is deeply integrated with the parameter tuning logic of the XGBoost algorithm: Verification phase: After the validation set is input into the model, the algorithm outputs the predicted matching degree of each core evaluation dimension (e.g., the predicted value of the industry chain correlation dimension is 0.78 vs. the actual value is 0.82). The dimension matching accuracy (required to be ≥85%) and the mean square error of the comprehensive score (required to be ≤0.05) are calculated. Optimization phase: If verification fails to meet the criteria, a grid search method is used to adjust key XGBoost parameters—including the maximum tree depth (to control overfitting, range 3-8), the learning rate (to control the iteration step size, range 0.01-0.1), and the minimum weight of leaf nodes (to control the complexity of the tree, range 0.1-0.5). Simultaneously, the parameters of indicators in each dimension are optimized (for example, the calculation factor for the "supporting facilities compliance rate" is expanded to include data on logistics nodes within a 3-kilometer radius). For example, when the accuracy of the industrial agglomeration dimension was only 82%, adjusting the maximum tree depth from 5 to 6 and reducing the learning rate from 0.05 to 0.03 increased the dimension's accuracy to 87%.

[0051] In step S225, the trained industrial ecosystem adaptation model relies on the efficient reasoning capability of the XGBoost algorithm to achieve two core functions: Real-time computing: After receiving real-time enterprise demand feature sets (such as "new annual production capacity of 2 million units") and regional industrial ecosystem data (such as "the park's existing supporting production capacity is 5 million units"), the model quickly completes feature matching (inference time ≤ 0.5 seconds) through the pre-trained decision tree model library and outputs a comprehensive matching score (such as 0.82). Sorting output: With the goal of industrial ecosystem clustering, the algorithm performs secondary weighting on the basis of the comprehensive score. Through the feature gain mechanism of XGBoost, an additional gain coefficient is set for the "correlation with existing industrial clusters" feature (such as the proportion of shared core suppliers and the frequency of historical collaboration) (when the correlation is greater than 60%, the feature gain is increased by 20%). Finally, the candidate parks are ranked from high to low according to the weighted scores (for example, Park A scores 0.89, Park B scores 0.81, and Park C scores 0.73), ensuring that the ranking results not only meet the needs of enterprises but also strengthen the orientation of industrial clustering.

[0052] Through the above process, the industrial ecosystem adaptation model achieves accurate quantitative matching between enterprise needs and regional industrial ecology, providing a data-driven decision-making basis for subsequent park screening, ensuring that candidate parks not only meet the individual needs of enterprises but also conform to the development direction of regional industrial clusters.

[0053] S30. Based on the quantitative demand standards of the enterprise demand feature set, establish a multi-factor adaptation relationship between the land space scale and building function parameters of the target enterprise and the candidate park; use a multi-dimensional compliance verification unit to perform matching verification on the municipal and supporting factor thresholds and restrictive factors of the candidate park, and eliminate parks that do not meet the constraints of the quantitative demand standards.

[0054] In this embodiment, the core of this step is to achieve accurate matching and compliance screening between candidate parks and enterprise needs: on the one hand, based on the quantitative demand standards of the enterprise demand feature set, a multi-factor adaptation relationship between the target enterprise and the candidate parks in terms of land space scale and building function parameters is established to clarify the degree of matching between the two; on the other hand, through the multi-dimensional compliance verification unit, the municipal and supporting factor thresholds and restrictive factors of the candidate parks are checked for matching, and parks that do not meet the constraints of the quantitative demand standards are eliminated. Finally, the candidate parks that meet the basic requirements are retained to provide screened alternative objects for the subsequent site selection plan output.

[0055] S40. Based on the verification results, a site selection plan is output. The site selection plan includes a recommended ranking of industrial parks, suitable industrial land and industrial buildings, and includes customized resource matching suggestions and restrictive factor risk response strategies that match the enterprise's demand feature set.

[0056] In this embodiment, this step generates a final site selection plan based on the aforementioned verification results. The core of this step is to integrate the matching information and risk response strategies to provide enterprises with a comprehensive decision-making basis for settlement. The specific contents are as follows: The site selection plan is based on the results of multiple rounds of verification. First, a ranking of recommended parks is formed. The ranking is determined from high to low based on the comprehensive adaptability (such as multiple factor adaptation scores and industrial ecological matching). For example, Park A (comprehensive score 0.89), Park B (0.81), and Park C (0.73) are ranked in turn; at the same time, the suitable industrial land and industrial buildings in each recommended park are clarified, such as Park A's Plot No. 3 (area of ​​50 mu, floor area ratio 1.2) and Standard Factory Building No. 2 (story height 9 meters, load-bearing capacity 15 tons / square meter), which accurately match the company's land space scale and building function requirements.

[0057] The scheme also includes two types of core auxiliary information: first, customized resource matching recommendations, which provide targeted resource connection solutions for enterprise demand characteristics, such as recommending a list of core suppliers within 3 kilometers and logistics collaboration solutions for enterprises with industry chain-related needs; for enterprises with high energy consumption needs, match double-loop power grids and energy-saving subsidy strategies. Second, restrictive factor risk response strategies, which propose solutions for restrictive factors found during verification, such as ecological red line constraints and upper limit of production capacity restrictions. For example, when a certain park has a sewage discharge index limit, it is recommended that the enterprise use water recycling technology to reduce emissions, or assist in applying for regional sewage rights replacement quota.

[0058] Through the above, the site selection scheme not only clearly defines specific entry options, but also provides resource connection and risk resolution paths, realizing a closed loop from "screening results" to "landing plans".

[0059] S50, according to the feedback data of the target area key industry categories, dynamically update the quantitative demand standard and the industry ecological adaptation model parameters.

[0060] In this embodiment, this step realizes the dynamic iteration of the quantitative demand standard and the industry ecological adaptation model by continuously collecting feedback data of the target area key industry categories, ensuring that the site selection method always fits the actual regional industrial development, and the specific implementation is as follows: (1) Core sources and types of feedback data Feedback data focuses on the actual operation and development of the target area's key industries, mainly including three categories: Enterprise operation feedback: actual production data of the entering enterprise (such as land utilization rate, actual adaptation rate of building function), operating indicators (such as production capacity compliance rate, industry chain collaboration efficiency), and demand deviation records (such as whether the originally set municipal element threshold meets the actual energy consumption demand).

[0061] Industry ecology change data: changes in the cluster size of the region's key industries (such as the number of new upstream and downstream enterprises), supporting facility bearing pressure (such as the difference between the actual throughput of the transportation hub and the original threshold), and the impact of strategy adjustment (such as the restriction of newly added ecological protection strategies on industrial layout).

[0062] Model prediction deviation data: the deviation between the prediction matching degree of the early industry ecological adaptation model and the actual entry effect (such as the model predicting that the industry chain correlation degree of a certain park is 0.8, while the actual correlation degree is only 0.6).

[0063] (2) Dynamic update of quantitative demand standard Based on the feedback data, the quantitative demand standard set in step S10 is adjusted accordingly to ensure that the standard is synchronized with the actual demand of the enterprise and the carrying capacity of the region: If the majority of enterprises report that the "original land use spatial scale standard is too small" (e.g., the actual land utilization rate reaches 110%), the land area threshold will be raised (e.g., from 50 mu to 60 mu), or the floor area ratio parameter will be optimized (e.g., from 1.2 to 1.5); If the actual adaptation rate of the building function is low (for example, the actual requirement for the clean workshop level of a certain type of enterprise is 2 levels higher than the original standard), the building function parameters will be updated (for example, Class10000 will be adjusted to Class1000); If municipal and supporting elements are frequently overloaded (for example, the power supply capacity of a certain park often reaches above 95% of the threshold), the corresponding threshold will be raised (for example, from 2000KVA to 2500KVA), or new supplementary indicators will be added (such as adding "emergency power supply switching time ≤ 10 seconds").

[0064] (3) Dynamic optimization of industrial ecosystem adaptation model parameters In response to model prediction deviations found in feedback data, the core parameters of the industrial ecosystem adaptation model are adjusted to improve the model's prediction accuracy: If the feedback shows that "the impact of industrial chain correlation on actual settlement effect is greater than the original setting of the model" (for example, for every 10% increase in correlation, the survival rate of enterprises increases by 15%), then increase the weight coefficient of "industrial chain correlation dimension" in the model (for example, adjust it from 0.4 to 0.5); If the prediction error of a certain type of feature is significant (e.g., the deviation between the predicted value and the actual value of "completeness of supporting facilities" exceeds 15%), the quantitative indicator of this feature should be optimized (e.g., from "number of facilities" to "actual utilization rate of facilities + response time"); If there are major adjustments to regional industrial strategies (such as the addition of mandatory requirements for “green manufacturing”), a new “environmental compliance” feature dimension will be added to the model and assigned a corresponding weight (such as 0.2) to ensure that the model adapts to the latest strategic direction.

[0065] Through the above-mentioned dynamic update mechanism, the quantitative demand standards always fit the actual operational needs of the enterprise, and the industrial ecological adaptation model continuously adapts to the changes in the regional industrial ecology, ultimately achieving self-optimization and iterative upgrading of the site selection method, providing more accurate and adaptable decision-making support for subsequent enterprise site selection.

[0066] In one embodiment, before step S40, the following steps are further included: S301. Construct a virtual simulation scenario of the target area based on digital twin technology, input the verified candidate park data, enterprise demand feature set, and current industrial base data of the target area into the virtual simulation scenario, and simulate the industrial driving effect, economic benefits, and supporting measures input-output ratio of the target enterprise after it settles in the candidate park within a preset period; wherein, the industrial driving effect includes the number of upstream and downstream enterprises agglomerated and the extent of the improvement in the collaborative efficiency of the industrial chain; the economic benefits include the annual tax contribution value and the number of new jobs; and the supporting measures input-output ratio is the ratio of the park infrastructure investment to the expected output value of the enterprise; In this embodiment, this step relies on digital twin technology to build a virtual simulation scene of the target area. The core is to predict the long-term effects of the enterprise after its entry through virtual-reality mapping. The specific process is as follows: Virtual simulation scene construction: A digital twin foundation is constructed based on the target area's geographic information (e.g., topography, road network, park layout) and industry data (e.g., existing enterprise distribution and industrial chain structure). Using 3D modeling technology, a virtual image of the park's buildings, infrastructure, and surrounding environment is created, ensuring that the simulation scene aligns with the physical world in terms of spatial scale and resource distribution. Geographic information data can be obtained through a GIS system, while industry data can be obtained online from the corresponding official websites.

[0067] Input data integration: Import the verified candidate park data (such as land use spatial parameters and building function indicators), enterprise demand feature sets (such as production capacity scale and industrial chain-related demand) and the current industrial base data of the target area (such as the number of existing upstream and downstream enterprises and the carrying capacity of supporting facilities) into the virtual simulation scene to establish a data association model (such as the mapping relationship between enterprise production capacity and surrounding logistics capacity).

[0068] Multi-dimensional simulation: simulates the dynamic changes of the target enterprise over a preset period (such as 3 years or 5 years) in a virtual scenario, focusing on outputting three types of indicators, including: Industry-driven effect: Calculate the number of upstream and downstream enterprises clustered using an industry chain transmission algorithm (e.g., expecting to attract five parts suppliers and two logistics service providers within three years). Calculate the extent of improvement in industry chain collaborative efficiency using a collaborative efficiency model (e.g., shortening the raw material transportation cycle from seven days to three days, a 57% efficiency increase). Economic benefits: Calculate the annual tax contribution (e.g., an estimated annual tax payment of 20 million yuan) based on the company's fixed asset investment, production capacity planning, and regional taxation strategy. Also, estimate the number of new jobs (e.g., 300 direct jobs and 500 indirect jobs) based on the company's employment scale and the impact of the industrial chain. Input-output ratio of supporting measures: Calculate the additional infrastructure investment required by the park to meet the needs of enterprises (such as an investment of 5 million yuan in expanding a substation and 3 million yuan in building a new sewage treatment plant), combined with the expected annual output value of the enterprises (such as 500 million yuan), and calculate the ratio ((500+300) / 50000=1.6%).

[0069] During the simulation process, dynamic adjustment parameters (such as market demand fluctuations and regional strategy changes) are set synchronously to ensure the robustness of the simulation results (for example, when market demand increases by 20%, the number of new jobs is recalculated to 900).

[0070] S302. Prioritize the candidate parks based on the simulation results, calculate the weighted comprehensive score of the industrial driving effect, economic benefits and supporting measures input-output ratio of each candidate park within a preset simulation period, and screen out the candidate park with the best comprehensive score.

[0071] In this embodiment, this step achieves accurate ranking of candidate parks through weighted comprehensive scoring. The core is to convert multi-dimensional simulation indicators into comparable comprehensive scores. The specific process is as follows: Indicator weight setting: The weight of each indicator is determined based on regional development priorities (such as industrial upgrading orientation and economic benefit orientation), for example: The weight of the industrial driving effect is 0.4 (if the region focuses on improving the industrial chain); Economic benefit weight is 0.35 (if the region focuses on taxation and employment); The weight of the input-output ratio of supporting measures is 0.25 (if the region focuses on resource utilization efficiency).

[0072] Comprehensive score calculation: Standardize the simulation index values ​​of each candidate park (for example, the "number of upstream and downstream enterprises clustered" in the industrial driving effect is normalized to the range of 0 to 1 according to the maximum possible value of the region), and then sum them according to the weight. The formula is: Comprehensive score = (standardized value of industrial driving effect × 0.4) + (standardized value of economic benefits × 0.35) + (standardized value of input-output ratio × 0.25).

[0073] For example, the three standardized values ​​of Park A are 0.9, 0.85, and 0.8 respectively, so the comprehensive score = 0.9×0.4+0.85×0.35+0.8×0.25=0.36+0.2975+0.2=0.8575.

[0074] Screening for the best park: Sort by comprehensive score from high to low and select the candidate park with the highest score (for example, Park A scored 0.8575, which is higher than Park B's 0.82 and Park C's 0.79, making it the best candidate).

[0075] Through the above steps, an in-depth evaluation from "static adaptation verification" to "dynamic effect simulation" is achieved, which not only takes into account the current adaptability of the park and the enterprise, but also predicts the long-term industrial value and economic and social benefits, providing a more comprehensive decision-making basis for outputting a scientific site selection plan in step S40.

[0076] In one embodiment, in step S10, while constructing the enterprise demand feature set, the following steps are also included: Constructing a community feature set for the target area, wherein the community feature set includes at least community population structure data, community public service carrying capacity data, and community environmental sensitive factor data; Among them, in step S20, the step of calculating the matching degree between the enterprise demand feature set and the target region's industrial ecology through the industrial ecology adaptation model specifically includes: S201. Calculate a first matching degree between the enterprise demand feature set and the target region's industrial ecology, where the first matching degree is determined based on the industrial chain correlation, industrial cluster fit, and supporting facilities completeness. S202. Calculate a second matching degree between the enterprise demand feature set and the community feature set, where the second matching degree is determined based on the compatibility between the enterprise's talent needs and the community's demographic structure data, the incremental demand for community public service carrying capacity data from enterprise operations, and the compatibility between the enterprise's environmental impact and community environmental sensitive factor data. S203: Perform weighted fusion of the first matching degree and the second matching degree to obtain a final matching degree between the enterprise and the target area, which serves as a basis for outputting candidate parks by the industrial ecosystem adaptation model.

[0077] In this embodiment, by constructing a community feature set and incorporating it into the matching degree calculation, an upgrade from "industry adaptation" to "industry-community collaborative adaptation" is achieved. The specific implementation is as follows: (1) Construction logic of community feature set The community characteristic set serves as an implicit constraint for enterprise location selection, focusing on capturing the interactive impact of regional social resources and enterprise operations: Community demographic data: This covers age distribution (e.g., working-age population accounts for 70%), education level (e.g., 35% of the population has a college degree or higher), and professional skill composition (e.g., the number of skilled workers in the manufacturing industry). This is used to assess the feasibility of local talent recruitment for companies. Community public service capacity data: This includes data on the number of beds in medical facilities (e.g., 200 beds in a community hospital), educational resource capacity (e.g., 3,000 primary school places), and daily throughput at transportation hubs (e.g., 50,000 passengers per day at a subway station). This is used to predict the incremental demand for public services from new employees in enterprises. Community environmental sensitive element data: including environmental function zoning (such as whether located in the ecological protection red line area), noise sensitive point distribution (such as the nearest distance of 500 meters from residential area), atmospheric pollutant emission threshold (such as the upper limit of sulfur dioxide emission 100 tons / year), used to evaluate the compatibility of enterprise environmental impact and community.

[0078] (2) Calculation dimension and method of second matching degree The second matching degree quantitatively evaluates the collaborative adaptability of enterprise and community through multiple dimensions: Personnel demand adaptability: Calculation logic: cosine similarity calculation is performed on enterprise job demand (such as technical research and development position proportion 40%) and community population occupation skill composition (such as technical talent proportion 30%), the formula is: For example: if the enterprise needs 100 technical personnel, and the community has 80 similar talents, then the adaptability = 80 / 100 = 0.8.

[0079] Public service incremental demand: Calculation logic: according to the size and structure of enterprise employees, predict the new public service demand (such as new medical demand = employee number × annual average number of visits), and compare with community carrying capacity, the formula is: Incremental demand pressure = new demand / (existing carrying capacity × safety redundancy coefficient).

[0080] For example: the enterprise adds 500 employees, and it is expected to add 10 person-times of daily medical demand, the community hospital has a daily reception capacity of 200 person-times (safety redundancy coefficient 1.2), then the pressure = 10 / (200×1.2)≈0.042 (the smaller the pressure value, the more adaptive).

[0081] Environmental compatibility: Calculation logic: compare enterprise pollutant emission list (such as annual emission COD 50 tons) with community environmental sensitive element threshold (such as regional COD capacity remaining 100 tons / year), the formula is: Compatibility = 1 - (enterprise emission / environmental capacity remaining).

[0082] For example: the noise emission compliance rate of the enterprise is 95%, and the distance from the residential area meets the requirements, then the compatibility = 0.95 × 1 = 0.95.

[0083] (3) Weighted fusion mechanism of double matching degree The final matching degree is realized by weighted fusion of the first matching degree (industrial ecology) and the second matching degree (community collaboration), the formula is: Final matching degree = first matching degree × + second matching degree × .

[0084] in, 、 is the weight coefficient, which is dynamically adjusted according to the regional development strategy: If the region focuses on industrial agglomeration (such as national-level economic development zones), then =0.7, =0.3; If the region emphasizes the integration of industry and city (such as urban new district), then =0.5, =0.5; If the community environment is sensitive (such as ecological conservation areas), =0.4, =0.6.

[0085] Through this mechanism, the candidate parks output by the industrial ecological adaptation model not only meet the needs of industrial chain coordination (such as supporting rate ≥ 80%), but also take into account the sustainable development of the community (such as public service pressure ≤ 0.3). For example, a biopharmaceutical company finally chooses a location in =0.5, =0.5, the final matching degree of candidate park A = 0.8 (industry) × 0.5 + 0.9 (community) × 0.5 = 0.85, which is better than park B (matching degree 0.82) that is only adapted to the industry but has great community pressure.

[0086] Through this mechanism, we can avoid the decline in employees' quality of life (such as traffic congestion, degree shortage) or environmental conflicts (such as noise complaints) caused by insufficient community carrying capacity, give priority to parks that complement the community's resources (such as corporate training centers open to the community, and community medical care providing green channels for corporate employees), and regularly optimize community feature set indicators through community feedback data (such as public service satisfaction surveys and environmental monitoring data) to ensure the timeliness of matching calculations.

[0087] In one embodiment, step S30 specifically includes: S310. Construct a multi-factor quantitative adaptation model, wherein the multi-factor quantitative adaptation model establishes a quantitative mapping relationship between the land space scale parameters and building function parameters of the candidate park and the enterprise demand feature set based on the quantitative demand standard of the enterprise demand feature set.

[0088] In this embodiment, this step is based on the quantitative demand standards of the enterprise demand feature set determined in step S10 to build a multi-factor quantitative adaptation model. The core is to establish an accurate quantitative mapping relationship between the candidate park parameters and the enterprise needs, such as: The land use spatial scale parameters of the candidate park include the actual area of ​​the plot, volume ratio, length-to-width ratio of the plot, terrain slope, etc. (these data can be obtained by those skilled in the art through corresponding channels, such as data from park land planning archives and GIS systems); Building functional parameters include plant floor height, floor load-bearing capacity, cleanroom grade, fire protection facility configuration, etc. (these data can be obtained by those skilled in the art through corresponding channels, such as the completion acceptance report and structural design drawings of the park buildings); Quantitative mapping relationships are achieved through feature association algorithms. For example, the enterprise's demand for "land area ≥ 50 mu" and the park's "actual land area 55 mu" are mapped to "area adaptation coefficient 1.1", and the enterprise's demand for "factory building height ≥ 8 meters" and the park's "actual floor height 9 meters" are mapped to "floor height adaptation coefficient 1.125", thereby converting abstract demands into computable numerical associations.

[0089] The construction of the multi-factor quantitative adaptation model is based on the quantitative demand standards of the enterprise demand feature set, and its core is to establish a quantitative relationship between candidate parks and enterprise needs. Specifically, the quantitative standards related to land use space and building functions in the enterprise demand feature set (such as land area requirements, floor height requirements, functional zoning ratio, etc.) are first clarified; secondly, the land use space scale parameters (such as actual land area, volume ratio, etc.) and building function parameters (such as building type, usable functional area, etc.) of the candidate parks are sorted out; finally, through quantitative methods (such as parameter correspondence and numerical association rules), the above parameters of the candidate parks are matched with the quantitative demand standards of the enterprises, forming a clear quantitative mapping relationship - that is, a specific quantitative expression of how the park parameters correspond to, meet, or deviate from the enterprise needs. This constructs a model framework that can quantitatively measure the adaptation relationship between the two, providing a foundation for the subsequent calculation of adaptation through indicators.

[0090] S320: Determine a multi-dimensional adaptation index for the quantitative mapping relationship, where the multi-dimensional adaptation index includes at least a land space adaptation dimension and a building function adaptation dimension, and each dimensional adaptation index includes at least one quantitative index parameter and a corresponding weight coefficient; In this embodiment, this step clarifies the specific dimensions and indicator system of the adaptation assessment to ensure that the assessment direction is consistent with the core needs of the enterprise: Land-use spatial adaptation dimension: This includes at least one quantitative indicator parameter, such as "area deviation rate" ((actual park area - enterprise required area) / enterprise required area; the smaller the absolute value, the better the adaptability), "floor area ratio fit" (actual park floor area ratio / enterprise required floor area ratio; the closer to 1, the better the adaptability). The weight coefficient can be set to 0.4 (based on the importance of land use requirements to the enterprise); Building Function Adaptation Dimension: This dimension includes at least one quantitative indicator parameter, such as "Story Height Compliance Rate" (1 when the actual floor height of the park is ≥ the required floor height of the enterprise, otherwise 0), "Load-bearing Matching Degree" (actual load-bearing capacity of the park divided by the required load-bearing capacity of the enterprise, calculated as 1 when ≥1, and calculated as the actual ratio when <1). The weight coefficient can be set to 0.6 (if the enterprise has higher requirements for building functions); The weight coefficient is determined through the hierarchical analysis method, combined with the production characteristics of the enterprise (for example, heavy manufacturing enterprises may increase the weight of "load-bearing matching") and the general industry standards to ensure that the importance of the indicator is consistent with actual needs.

[0091] S330: Standardize each quantitative indicator parameter to generate comparable quantitative values, wherein the standardization includes but is not limited to calculation of deviation rate, degree of fit, and compliance rate; In this embodiment, this step eliminates the dimensional differences of different indicators through standardization to generate directly comparable quantitative values. The specific method includes: Deviation rate calculation: Applicable to continuous parameters (such as area and floor area ratio). The formula is "(actual park value - enterprise demand value) / enterprise demand value". The result is normalized to the range of "-1 to 1" (for example, if the actual area is 55 mu and the demand is 50 mu, the deviation rate = (55-50) / 50 = 0.1, which is 0.1 after normalization). Compatibility calculation: Applicable to proportional parameters (such as load-bearing matching), directly take the ratio of "actual value of the park / demand value of the enterprise" (if ≥1, it is calculated as 1), and the result falls in the range of "0-1" (for example, if the actual load-bearing value is 10 tons / square meter and the demand is 8 tons / square meter, the compatibility = 1); Compliance rate calculation: Applicable to discrete parameters (such as fire protection facility configuration), calculated as "number of indicators that meet the requirements / total number of indicators" (for example, if 2 out of 3 fire protection indicators meet the standards, the compliance rate = 2 / 3 ≈ 0.67).

[0092] After standardization, the quantitative values ​​of all indicators are unified to the range of "0~1", providing a comparable basis for subsequent weighted aggregation.

[0093] S340, performing weighted aggregation on the quantified values ​​of each dimension based on the weight coefficient to generate a multi-factor comprehensive fitness score; In this embodiment, this step performs weighted aggregation on the quantized values ​​of each dimension after normalization in S330 based on the weight coefficient determined in S320. The calculation formula is: Comprehensive adaptability score = (quantified value of land spatial adaptability dimension × land weight) + (quantified value of building function adaptability dimension × building weight).

[0094] For example: the standardized value of the land use space adaptation dimension of a candidate park is 0.9 (area deviation rate 0.1, floor area ratio matching degree 0.8, take the average), with a weight of 0.4; the standardized value of the building function adaptation dimension is 0.85 (story height standard rate 1, load-bearing matching degree 0.7, take the average), with a weight of 0.6, then the comprehensive adaptability score = 0.9×0.4+0.85×0.6=0.36+0.51=0.87.

[0095] S350. Set a fitness threshold, include candidate parks whose comprehensive fitness scores reach the fitness threshold into the subsequent verification process, and eliminate candidate parks that do not reach the fitness threshold.

[0096] In this embodiment, this step completes the first round of screening by setting thresholds to ensure that the parks entering the subsequent process meet the basic adaptation requirements: The suitability threshold is set based on the suitability score distribution of historical successful site selection cases (for example, the lowest score of historical successful cases, 0.7, is used as the threshold), or determined through expert review (taking into account industry characteristics, such as the threshold for precision manufacturing enterprises can be raised to 0.8); If the comprehensive suitability score of the candidate park is ≥ the threshold (such as 0.87 ≥ 0.7), it will be included in the subsequent verification process; if the score is < the threshold (such as 0.65 < 0.7), it will be directly eliminated.

[0097] Through the above steps, a full-process quantitative assessment from "parameter mapping" to "quantitative scoring" to "threshold screening" can be achieved, accurately retaining candidate parks that are compatible with the company's land use and building needs, laying the foundation for subsequent compliance verification.

[0098] In one embodiment, the current industrial base data of the target area further includes a regional resource feature set, wherein the regional resource feature set is composed of at least basic guarantee elements, industrial synergy elements, and development potential elements; The basic guarantee elements include data on the completeness of infrastructure and supporting public services in the target area; The industrial synergy factors include the scale and structure of the industrial clusters already formed in the target area, the density of connections between upstream and downstream enterprises, and information on shareable technology R&D platforms; The development potential factors include future industry orientation data, land reserves and development planning data, and data on the expected impact of transportation hub construction progress on the industry's radiation range.

[0099] In this example, the inclusion of the target region's current industrial infrastructure data into the regional resource feature set further expands the dimensions of industrial ecosystem assessment, enabling a more comprehensive reflection of the potential for adaptation between regional resource endowments and industrial development. Its specific composition and functions are as follows: (1) Basic security elements The basic guarantee elements focus on the region's ability to supply basic resources to support industrial operations, which is a prerequisite for the establishment of enterprises: Infrastructure completeness data: This covers quantitative indicators of hardware facilities such as transportation, energy, and water conservancy. These indicators include road network density (5 kilometers of road length per square kilometer), power supply reliability (annual power outage duration ≤ 8 hours), and sewage treatment plant capacity (daily processing capacity of 20,000 tons). These indicators are used to assess whether a region can meet the basic production and operation needs of enterprises. Public service supporting data: including the supply of talent apartments (e.g., 5,000 units), the number of vocational skills training institutions (e.g., 3 national-level high-skilled talent training bases), the coverage rate of legal service institutions (e.g., 1 per 10 square kilometers), etc., reflecting the region's ability to support the living and development needs of corporate employees.

[0100] This type of data mainly comes from the annual reports of municipal management departments and GIS spatial databases. It is the key input for calculating the "completeness of supporting facilities" (the core indicator of the first matching degree in step S201). For example, if the infrastructure completeness data of a certain park meets the standard rate of 85%, it will directly improve its basic adaptability to corporate needs.

[0101] (2) Industrial synergy factors Industrial synergy factors focus on the synergistic potential of the region's existing industrial ecosystem and serve as a core reference for enterprises to integrate into the local industrial network: The scale and structure of established industrial clusters: For example, an electronic information industry cluster with an annual output value of 50 billion yuan, covering chip design, terminal manufacturing, software services, and other sub-sectors, can be used to determine the complementarity of an enterprise with an existing cluster (for example, a terminal manufacturing enterprise can rely on chip suppliers within the cluster to reduce costs); Density of connections between upstream and downstream companies: quantified by the number of core suppliers within a 30-kilometer radius and the proportion of annual collaborative transaction volume (e.g., an auto parts company's annual collaborative volume with a local automaker accounts for 60%), which directly impacts the calculation of the degree of industry chain connection. Shareable information on technology R&D platforms: including the number of national key laboratories (e.g., 2), the annual conversion rate of industry-university-research cooperation projects (e.g., 10 projects / year), etc., reflecting the region's ability to support the technological innovation needs of enterprises.

[0102] This type of data comes from the cluster development report and enterprise collaboration ledger of the Industrial Park Management Bureau. It is the core basis for calculating the "industrial chain correlation" and "industrial agglomeration fit" in the first matching degree. For example, if the industrial collaboration factor score of a park is 0.85, its matching degree with the demand-oriented enterprises in the industrial chain will be significantly improved.

[0103] (3) Development potential factors The development potential factor focuses on the evolutionary trends of the region's future resource supply and industrial layout, providing a forward-looking assessment for the long-term development of enterprises: Future industry guidance data: Based on the industrial plans released by relevant departments, the region's key industries (such as new energy and artificial intelligence) and strategic support areas (such as R&D subsidies for high-tech enterprises) are clearly defined to assess the alignment of enterprises with the region's long-term development strategy. Land reserves and development planning data: including the area of ​​industrial land available for sale (e.g., 1,000 mu), the planned floor area ratio adjustment space (e.g., from 1.2 to 2.0), etc., reflecting whether the region can meet the needs of enterprises for production expansion; The expected impact of the progress of transportation hub construction on the scope of industrial radiation: For example, the planned high-speed rail station is expected to open to traffic in 2026. After completion, the regional radiation radius will be expanded from 50 kilometers to 100 kilometers, which will directly affect the prediction of corporate logistics costs and market coverage capabilities.

[0104] This type of data mainly comes from the land use planning of the Natural Resources Bureau and the major project database of the National Development and Reform Commission. Its role is reflected in the dynamic correction of the first matching degree. For example, due to the construction of a transportation hub, the industrial radiation range of a certain park will expand by 50% in the next three years, and its matching degree with logistics-dependent enterprises can be increased by 10%.

[0105] The inclusion of regional resource feature sets enables the industrial ecosystem adaptation model to not only assess “current adaptability” but also take into account “long-term sustainability”: Basic guarantee factors ensure that enterprises can "settle down", industrial synergy factors ensure that enterprises can "integrate", and development potential factors ensure that enterprises can "grow"; In the first matching degree calculation, these data are weighted and integrated into indicators such as "completeness of supporting facilities" and "correlation of the industrial chain" (such as a weight of 0.3 for basic guarantee factors, a weight of 0.5 for industrial synergy factors, and a weight of 0.2 for development potential factors), so that the final matching degree is more in line with the actual regional resources and the needs of the enterprise throughout its life cycle.

[0106] In one embodiment, after step S50, the method further includes: Step S60: Build a dynamic adaptation and adjustment mechanism. This mechanism aims to build dynamic optimization capabilities for the entire life cycle of site selection plans. Through real-time monitoring, triggered recalculation, dynamic adjustment, and difference verification, it ensures that site selection decisions are always synchronized with changes in enterprise needs and regional industrial ecology. The specific implementation is as follows: S610. Set a dynamic monitoring cycle to collect real-time change data of the enterprise demand feature set and dynamic update data of the target area industrial ecology.

[0107] In this embodiment, this step captures dynamic changes in enterprises and regions through regular monitoring, providing a data basis for the adjustment mechanism: Dynamic monitoring cycle setting: Determine the cycle based on industry characteristics and regional industry change rates (e.g., quarterly monitoring for heavy industrial enterprises, monthly monitoring for high-tech enterprises, and immediate monitoring in case of sudden strategy adjustments); Data type collected: Real-time changes in the enterprise's demand feature set: including capacity adjustments (e.g., increasing annual output from 1 million units to 1.5 million units), process upgrades (e.g., upgrading the cleanroom from Class 1000 to Class 100), and changes in talent requirements (e.g., adding 50 R&D personnel). Dynamically updated data on the industrial ecology of the target area: including newly added industrial chain enterprises (such as two new core suppliers within 30 kilometers), supporting facilities upgrades (such as a new 110KV substation in the park), and strategic adjustments (such as tightening of environmental emission standards).

[0108] S620: When it is monitored that the change range of the key indicators of the enterprise demand feature set exceeds the preset threshold or the structural indicators of the regional industrial ecology are adjusted, the recalculation of the industrial ecology adaptation model is triggered; In this embodiment, this step clearly defines the thresholds and scenarios that trigger dynamic adjustment to avoid invalid recalculations: Change thresholds for key enterprise demand indicators: These are set based on industry stability (e.g., fixed asset investment change ≥ 20%, core process parameter adjustment ≥ 1 level, talent demand increase ≥ 30%). For example, if an auto parts company plans to increase its annual production capacity from 500,000 to 700,000 units (an increase of 40% > 20%), this triggers an adjustment. Adjustments to regional industrial ecological structural indicators: including changes in industrial planning direction (such as the region's shift from "traditional manufacturing" to "smart manufacturing"), major infrastructure changes (such as the addition of a high-speed rail hub that changes the logistics radiation range), and upgraded environmental constraints (such as the addition of PM2.5 emission limits). For example, a region's release of a "ban on new chemical production capacity" policy directly triggers the adaptation and recalculation of related companies.

[0109] The trigger logic is automatically judged by the rule engine, and the subsequent adjustment process is initiated when any condition is met.

[0110] S630: Based on the collected dynamically updated data, re-execute the matching degree calculation step and re-perform the matching verification, eliminate the parks that do not meet the quantitative demand standard constraints, generate a dynamically adjusted candidate park ranking, and output a matching degree change analysis report; In this embodiment, this step updates the evaluation results based on the latest data to ensure the suitability of the candidate parks: Recalculate the matching degree: Call the industrial ecosystem adaptation model, input the real-time needs of enterprises (such as the newly added "annual electricity consumption ≥ 15 million kWh") and regional dynamic data (such as the park power supply capacity has been upgraded to 2000KVA), and recalculate the first matching degree (such as industrial chain correlation) and the second matching degree (such as community adaptability). Re-matching verification: Reuse the multi-dimensional compliance verification unit of step S30 to re-verify the municipal factors (such as newly added sewage treatment capacity) and restrictive factors (such as the latest ecological red line) of the candidate parks, and eliminate parks that do not meet the standards due to changes (for example, a park that has been included in the ecological red line due to policy adjustments will be directly eliminated); Generate output results: form a dynamically adjusted ranking of candidate parks (for example, a park originally ranked 3rd is promoted to 1st due to supporting upgrades), and output a matching degree change analysis report (including the reason for the change, such as "Park A's matching degree increased from 0.89 to 0.92 due to the addition of two new upstream suppliers").

[0111] S640: Compare the dynamically adjusted candidate park ranking with the recommended park ranking in the initial output site selection plan, calculate the difference between the two, and output an optimized site selection plan if the difference exceeds a preset threshold. The optimized site selection plan includes: If the enterprise has not yet been established, the recommended park ranking and supporting resource matching suggestions will be updated directly; If the enterprise has already been established, output resource adaptation and optimization strategies or phased migration and transition plans based on the existing park.

[0112] In this embodiment, this step determines whether to update the site selection plan based on the difference, taking into account both decision stability and flexibility: Calculation of difference: quantified by "ranking change rate" (e.g., if three of the top five parks change, the change rate is 60%), or by calculating the mean square error of the comprehensive score (e.g., the mean difference between the dynamic ranking and the initial ranking is greater than 0.1); Preset threshold setting: Determine based on the company's development stage (e.g., 30% for start-ups and 50% for mature companies) to avoid frequent adjustments that affect operations; Output solutions by scenario: If the enterprise has not yet settled: directly update the ranking of recommended parks (e.g., adjust Park A to be the first choice) and simultaneously update the supporting resource matching suggestions (e.g., add "Sign a collaboration agreement with two new suppliers of Park A"); If the enterprise has already been established: output resource adaptation and optimization strategy (such as coordinating the access to temporary power generation equipment when the existing park has insufficient power supply) or phased migration transition plan (such as completing the relocation to Park B within 6 months and simultaneously retaining the existing factory building as a transitional warehouse).

[0113] Through the above mechanism, the dynamic adaptive adjustment mechanism realizes the upgrade of the site selection decision from a "one-time static scheme" to a "full-cycle dynamic optimization", ensuring the long-term adaptability of enterprises and regional resources.

[0114] It should be understood that the size of the serial number of each step in the above embodiment does not mean the order of execution, and the execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0115] In one embodiment, a smart site selection system based on a space-time data platform is provided, which corresponds to the smart site selection method based on a space-time data platform in the above embodiment. The smart site selection system based on a space-time data platform comprises: A multi-source data acquisition module is configured to acquire multi-source space-time data of key industries in a target area, construct an enterprise demand feature set based on the multi-source space-time data, and establish a quantitative demand standard based on the enterprise demand feature set. The enterprise demand feature set at least includes industrial attributes, fixed asset investment, industrial chain associated demand, and production process demand. The quantitative demand standard at least includes land space scale, building function parameter, municipal and supporting element threshold. A candidate park generation module is configured to construct an industrial ecological adaptation model based on the current industrial foundation data of the target area, calculate the matching degree of the enterprise demand feature set and the industrial ecology of the target area through the industrial ecological adaptation model, and output a candidate park according to the matching degree. The matching degree includes industrial chain correlation degree, industrial agglomeration fit degree, and supporting facility completeness. The industrial ecological adaptation model aims to industrial ecological clustering, and outputs the candidate park according to the matching degree. A candidate park screening module is configured to establish a multi-element adaptive relationship of land space scale and building function parameter between a target enterprise and the candidate park based on the quantitative demand standard of the enterprise demand feature set, and perform matching verification on the municipal and supporting element threshold and restrictive elements of the candidate park through a multi-dimensional compliance verification unit, and eliminate parks that do not meet the constraint conditions of the quantitative demand standard. A site selection scheme generation module is configured to output a site selection scheme based on the verification result, wherein the site selection scheme includes recommended park entry ranking, adaptive industrial land and industrial housing, and contains customized resource matching suggestions and restrictive element risk response strategies matched with the enterprise demand feature set. A dynamic optimization module is configured to dynamically update the quantitative demand standard and the industrial ecological adaptation model parameters according to the feedback data of key industries in the target area.

[0116] Regarding the specific definition of a smart site selection system based on a spatiotemporal data platform, please refer to the definition of a smart site selection method based on a spatiotemporal data platform above, which will not be repeated here. The various modules in the above-mentioned smart site selection system based on a spatiotemporal data platform can be implemented in whole or in part through software, hardware, and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0117] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Figure 2 As shown. The computer device includes a processor, a memory, a network interface and a database connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used for data storage, data processing, data analysis, etc. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, it realizes a smart site selection method based on a spatiotemporal data platform.

[0118] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, an intelligent site selection method based on a spatiotemporal data platform is implemented.

[0119] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, an intelligent site selection method based on a spatiotemporal data platform is implemented.

[0120] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0121] Those skilled in the art will clearly understand that for the sake of convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0122] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. All equivalent structural transformations made by using the contents of the present invention description and drawings under the inventive concept of the present invention, or direct / indirect application in other related technical fields are included in the patent protection scope of the present invention.

Claims

1. A smart site selection method based on a spatiotemporal data platform, characterized in that: The steps include: Collect multi-source spatiotemporal data on key industries in the target area, and construct a set of enterprise demand characteristics based on the multi-source spatiotemporal data. The set of enterprise demand characteristics includes at least industry attributes, fixed asset investment amount, industry chain-related demands, and production process requirements. Quantitative demand standards are established based on the set of enterprise demand characteristics. The quantitative demand standards include at least land use spatial scale, building function parameters, and municipal and supporting element thresholds. Based on the current industrial infrastructure data of the target region, an industrial ecosystem adaptation model is constructed. The matching degree between the enterprise demand feature set and the industrial ecosystem of the target region is calculated through the industrial ecosystem adaptation model. The matching degree includes the correlation degree of the industrial chain, the degree of industrial agglomeration fit, and the completeness of supporting facilities. The industrial ecosystem adaptation model aims to cluster the industrial ecosystem and outputs candidate parks based on the matching degree. Based on the quantitative demand standards of the enterprise demand feature set, a multi-factor adaptation relationship between the target enterprise and the land use spatial scale and building function parameters of the candidate parks is established; through a multi-dimensional compliance verification unit, the matching verification of the municipal and supporting factor thresholds and restrictive factors of the candidate parks is carried out, and parks that do not meet the constraints of the quantitative demand standards are eliminated; Based on the verification results, a site selection plan is output. The plan includes a recommended ranking of industrial parks, suitable industrial land and industrial buildings, and includes customized resource matching suggestions and restrictive factor risk response strategies that match the enterprise's demand characteristics. Based on the feedback data from key industries in the target area, the quantitative demand standards and industrial ecosystem adaptation model parameters are dynamically updated.

2. The smart site selection method based on the spatiotemporal data platform according to claim 1, characterized in that: Before the step of outputting the site selection plan based on the verification result, the following steps are also included: A virtual simulation scenario of the target area is constructed based on digital twin technology. The verified candidate park data, enterprise demand feature set, and current industrial base data of the target area are input into the virtual simulation scenario. The simulation simulates the industrial driving effect, economic benefits, and input-output ratio of supporting measures within a preset period after the target enterprise settles in the candidate park. The industrial driving effect includes the number of upstream and downstream enterprises agglomerated and the degree of improvement in the collaborative efficiency of the industrial chain. The economic benefits include the annual tax contribution value and the number of new jobs. The input-output ratio of supporting measures is the ratio of the park infrastructure investment to the expected output value of the enterprise. Based on the simulation results, the candidate parks are prioritized, and the weighted comprehensive scores of the industrial driving effect, economic benefits and input-output ratio of supporting measures of each candidate park within the preset simulation period are calculated to screen out the candidate park with the best comprehensive score.

3. The smart site selection method based on the spatiotemporal data platform according to claim 1, characterized in that: The step of collecting multi-source spatiotemporal data of key industries in the target area and constructing an enterprise demand feature set based on the multi-source spatiotemporal data also includes: Constructing a community feature set for the target area, wherein the community feature set includes at least community population structure data, community public service carrying capacity data, and community environmental sensitive factor data; The step of calculating the matching degree between the enterprise demand feature set and the target region's industrial ecology through the industrial ecology adaptation model specifically includes: Calculating the first matching degree between the enterprise demand feature set and the industrial ecology of the target region, where the first matching degree is determined based on the correlation degree of the industrial chain, the degree of industrial agglomeration fit, and the completeness of supporting facilities; Calculating a second degree of matching between the enterprise demand feature set and the community feature set, where the second degree of matching is determined based on the compatibility of the enterprise's talent needs with the community's demographic structure data, the incremental demand for community public service carrying capacity data from the enterprise's operations, and the compatibility of the enterprise's environmental impact with the community's environmental sensitive factor data; The first matching degree and the second matching degree are weightedly integrated to obtain the final matching degree between the enterprise and the target area, which serves as the basis for outputting candidate parks from the industrial ecological adaptation model.

4. The smart site selection method based on a spatiotemporal data platform according to claim 1, characterized in that: The step of establishing a multi-factor adaptation relationship between the target enterprise and the land use space scale and building function parameters of the candidate park based on the quantitative demand standards of the enterprise demand feature set specifically includes: Constructing a multi-factor quantitative adaptation model, wherein the multi-factor quantitative adaptation model establishes a quantitative mapping relationship between the land use spatial scale parameters and building function parameters of the candidate park and the enterprise demand feature set based on the quantitative demand standards of the enterprise demand feature set; Determining a multi-dimensional adaptation index of the quantitative mapping relationship, wherein the multi-dimensional adaptation index includes at least a land space adaptation dimension and a building function adaptation dimension, and each dimensional adaptation index includes at least one quantitative index parameter and a corresponding weight coefficient; Standardize each quantitative indicator parameter to generate comparable quantitative values, including but not limited to calculation of deviation rate, degree of fit, and compliance rate; The quantified values ​​of each dimension are weighted and aggregated based on the weight coefficient to generate a comprehensive fitness score of multiple factors; Set a suitability threshold, and include candidate parks whose comprehensive suitability scores reach the suitability threshold into the subsequent verification process, and eliminate candidate parks that do not reach the suitability threshold.

5. The smart site selection method based on a spatiotemporal data platform according to claim 1, characterized in that: In the step of constructing an industrial ecology adaptation model based on the current industrial basic data of the target area and calculating the matching degree between the enterprise demand feature set and the industrial ecology of the target area through the industrial ecology adaptation model, the construction of the industrial ecology adaptation model specifically includes the following steps: Collect enterprise demand feature sets and the current industrial infrastructure data of the target area for preprocessing to obtain standardized feature data that can be used for machine learning model training; The standardized feature data is used to train an industrial ecological adaptation model through a machine learning algorithm to obtain a trained industrial ecological adaptation model, and the trained industrial ecological adaptation model is used to: Real-time calculation of the matching degree between the enterprise demand feature set and the target region's industrial ecology; With the goal of industrial ecological clustering, the candidate parks are ranked according to the matching degree.

6. The smart site selection method based on the spatiotemporal data platform according to claim 5, characterized in that: The step of collecting the enterprise demand feature set and the current industrial base data of the target area and pre-processing them to obtain standardized feature data that can be used for machine learning model training specifically includes: Collect the enterprise demand feature set and the current industrial base data of the target area, wherein the current industrial base data includes industrial chain structure data, industrial cluster distribution data and supporting facilities data; Clean the collected data, including filling in missing values, correcting outliers and unifying data formats; Based on the correlation between the enterprise demand feature set and the target region's industrial ecology, feature mapping is performed on the cleaned data to generate an initial feature set that includes industry chain correlation features, industrial agglomeration features, and supporting features. The initial feature set is standardized, and each feature value is normalized to a preset value range to obtain standardized feature data that can be directly used for machine learning model training.

7. The smart site selection method based on the spatiotemporal data platform according to claim 6, characterized in that: The step of training the industrial ecological adaptation model with the standardized feature data through a machine learning algorithm to obtain a trained industrial ecological adaptation model specifically includes: Dividing the standardized feature data into a training set and a validation set; An initial industrial ecological adaptation model is trained based on the training set. The initial industrial ecological adaptation model uses the industrial chain association dimension, the industrial agglomeration fit dimension, and the supporting support dimension as core evaluation dimensions, and sets corresponding quantitative indicators and weight coefficients for each dimension; Calculate the comprehensive matching score of the training set using the initial industrial ecosystem adaptation model. The score is the weighted sum of the standardized quantitative indicators of each dimension. Use the validation set to verify the model performance, requiring the matching accuracy of the core evaluation dimensions to be no less than a preset threshold, and optimize the weight coefficients and indicator parameters of the industrial ecosystem adaptation model based on the validation results; A trained industrial ecological adaptation model is generated based on the optimized weight coefficients and indicator parameters. The industrial ecological adaptation model is configured as follows: Receive real-time enterprise demand feature sets and target region industrial ecosystem data, and output a comprehensive matching score between the two; With the goal of industrial ecological clustering, the comprehensive matching scores are subjected to a secondary weighted calculation to generate a ranking result of candidate parks, among which parks with a high degree of correlation with existing industrial clusters receive additional weight bonuses.

8. A smart site selection system based on a spatiotemporal data platform, used to implement the steps of a smart site selection method based on a spatiotemporal data platform as described in any one of claims 1 to 7, characterized in that: include: A multi-source data acquisition module is used to collect multi-source spatiotemporal data of key industries in the target area, construct an enterprise demand feature set based on the multi-source spatiotemporal data, and the enterprise demand feature set includes at least industry attributes, fixed asset investment amount, industry chain related demand and production process requirements; and establish quantitative demand standards based on the enterprise demand feature set, and the quantitative demand standards include at least land use spatial scale, building function parameters, and municipal and supporting element thresholds. The candidate park generation module is used to construct an industrial ecosystem adaptation model based on the current industrial infrastructure data of the target area. The industrial ecosystem adaptation model is used to calculate the matching degree between the enterprise demand feature set and the industrial ecosystem of the target area. The matching degree includes the industrial chain correlation, industrial agglomeration fit, and supporting facilities completeness. The industrial ecosystem adaptation model aims to cluster the industrial ecosystem and outputs candidate parks based on the matching degree. The candidate park screening module is used to establish a multi-factor adaptation relationship between the target enterprise and the candidate park's land use spatial scale and building function parameters based on the quantitative demand standards of the enterprise demand feature set; through a multi-dimensional compliance verification unit, the matching verification of the municipal and supporting factor thresholds and restrictive factors of the candidate park is carried out, and parks that do not meet the constraints of the quantitative demand standards are eliminated; A site selection plan generation module is used to output a site selection plan based on the verification results. The site selection plan includes a recommended ranking of the park to be settled, suitable industrial land and industrial buildings, and includes customized resource matching suggestions and restrictive factor risk response strategies that match the enterprise's demand feature set; The dynamic optimization module is used to dynamically update the quantitative demand standards and industrial ecological adaptation model parameters based on the feedback data of key industries in the target area.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the intelligent site selection method based on the spatiotemporal data platform as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the intelligent site selection method based on a spatiotemporal data platform as described in any one of claims 1 to 7 are implemented.

Images