Intelligent identification method and system for low-efficiency industrial land based on multi-dimensional index fusion

Abstract

The application provides a low-efficiency industrial land intelligent identification method and system based on multi-dimensional index fusion, relates to the technical field of intelligent identification, obtains industrial land original data from a multi-source platform; determines the comprehensive score of the industrial land based on the industrial land original data, determines the low-efficiency industrial land according to the comprehensive score; establishes an input-output data set by constructing a virtual input variable, determines the slack efficiency of the low-efficiency industrial land, and determines the dominant type of the low-efficiency industrial land according to the slack efficiency; and matches and outputs a strategy tool combination from a preset strategy tool library based on the dominant type. The application can realize accurate identification and cause classification of the low-efficiency industrial land, and improve the objectivity of identification and the pertinence of strategy matching.

Description

Technical Field

[0001] This application relates to the field of intelligent identification technology, and more specifically, to a method and system for intelligent identification of inefficient industrial land based on the fusion of multi-dimensional indicators. Background Technology

[0002] In recent years, with the increasing demands for industrial transformation and upgrading and high-quality development, the constraints on land resources have become increasingly prominent, and land use patterns are gradually shifting from incremental expansion to tapping into existing resources and improving efficiency. In the field of identifying inefficient industrial land, existing technologies typically evaluate land use intensity or economic performance. At the methodological level, some technologies use indicator systems similar to those for intensive land use evaluation, combining spatial morphology indicators to reflect the development level and utilization intensity of land parcels; others focus on economic performance indicators, forming a methodology primarily based on input-output efficiency. At the classification level, existing technologies are mostly based on historical samples or comprehensive scoring results, using threshold grading to classify and manage land parcel efficiency. At the strategic governance level, existing technologies generally emphasize building a full-cycle governance mechanism covering land use access, process supervision, and exit / redevelopment.

[0003] In the field of identifying inefficient industrial land, existing technologies typically evaluate land use intensity or economic performance. Identification methods often rely on single-dimensional indicators and subjective weighting, with threshold-based hierarchical management at the classification level and a focus on building a full-cycle governance mechanism at the policy governance level. However, existing methods struggle to objectively and accurately identify inefficient land use and trace its causes. Therefore, achieving accurate identification and causal classification of inefficient industrial land, and improving the objectivity of identification and the relevance of policy matching, are challenges facing the industry. Summary of the Invention

[0004] This application provides a method and system for intelligent identification of inefficient industrial land based on the fusion of multi-dimensional indicators, which can achieve accurate identification and cause classification of inefficient industrial land, and improve the objectivity of identification and the pertinence of strategy matching.

[0005] Firstly, this application provides a method for intelligent identification of inefficient industrial land based on the fusion of multi-dimensional indicators, the intelligent identification method comprising the following steps: Obtain raw data on industrial land from multiple sources; Based on the original data of industrial land, the comprehensive score of the industrial land is determined, and inefficient industrial land is identified according to the comprehensive score. For the inefficient industrial land, an input-output dataset is established by constructing virtual input variables, thereby determining the various relaxation efficiencies of the inefficient industrial land, and judging the dominant type of the inefficient industrial land based on the various relaxation efficiencies. Based on the dominant type, a differential matching is performed from the preset strategy tool library, and the corresponding strategy tool combination is output.

[0006] In this embodiment, obtaining raw industrial land data from a multi-source platform specifically includes: Obtain land parcel boundary data, plot ratio data, average number of floors data, building quality data, and development intensity data for industrial land from the land registration management system; Data on energy consumption, water consumption, unit tax revenue and electricity costs, unit tax revenue and water costs, environmental quality, and emissions exceeding standards for industrial land are obtained from the environmental monitoring platform. Obtain data on tax revenue per mu, revenue per mu, and overall labor productivity of industrial land from tax and statistics departments. Data on the proportion of large-scale enterprises, key enterprises, and R&D investment in industrial land were obtained from the systems of the Ministry of Industry and Information Technology and the Ministry of Science and Technology. Obtain infrastructure level data for industrial land from an IoT monitoring platform; Obtain data on unlicensed land use identification for industrial land from the land and resources management system; All the above data are combined into raw data for industrial land use.

[0007] In this embodiment, determining the comprehensive score of the industrial land based on the original industrial land data specifically includes: Spatial benefit indicators, economic benefit indicators, development potential indicators, and environmental benefit indicators are extracted from the original data of industrial land to construct a four-dimensional evaluation indicator system; The entropy weight method is used to determine the information entropy and information utility value of each indicator, and then the objective weight of each indicator is determined. Construct a weighted normalized matrix to determine the positive and negative ideal solutions, and then determine the distances from the industrial land to the positive and negative ideal solutions; The comprehensive score of the industrial land is determined based on the distances from the industrial land to the positive and negative ideal solutions.

[0008] In this embodiment, spatial benefit indicators, economic benefit indicators, development potential indicators, and environmental benefit indicators are extracted from the original industrial land data to construct a four-dimensional evaluation indicator system, specifically including: Data on plot ratio, average number of floors, building quality, development intensity, and infrastructure level are extracted from the original data of the industrial land to form spatial efficiency indicators; The total labor productivity, tax revenue per mu, and revenue per mu are extracted from the original data of the industrial land to form economic benefit indicators. The proportion of enterprises above designated size, the proportion of R&D investment, and the proportion of key enterprises are extracted from the original industrial land data to form development potential indicators. The unit tax revenue electricity fee, unit tax revenue water fee, and environmental quality data are extracted from the original data of the industrial land to form environmental benefit indicators; The spatial benefit indicators, economic benefit indicators, development potential indicators, and environmental benefit indicators are correlated and combined to form a four-dimensional evaluation index vector for industrial land.

[0009] In this embodiment, the entropy weight method is used to determine the information entropy and information utility value of each indicator, and then the objective weight of each indicator is determined, specifically including: The industrial land and all industrial land in its area are combined to form an evaluation object set, and a judgment matrix is ​​constructed. Standardize the data of each indicator in the judgment matrix; The information entropy of each indicator is determined based on the standardized data of each indicator. The information utility value is determined based on the information entropy of each indicator, and then the objective weight of each indicator is determined based on the information utility value of each indicator.

[0010] In this embodiment, for the inefficient industrial land, establishing an input-output dataset by constructing virtual input variables specifically includes: The aforementioned inefficient industrial land, together with other plots in the same area that have been identified as inefficient industrial land, constitute a reference group; The actual output values ​​of the inefficient industrial land in four dimensions—economic benefits, environmental benefits, development potential, and spatial benefits—are determined to form an actual output vector. Determine the virtual input value of the inefficient industrial land; The virtual input values ​​are paired with the actual output vectors to establish an input-output dataset.

[0011] In this embodiment, determining the relaxation efficiencies of the inefficient industrial land specifically includes: Using the virtual input vector of the input-output dataset as the input variable and the actual output vector as the output variable, a linear programming equation is constructed. Solve the linear programming equation to obtain the efficiency score and reference set weight coefficient of the inefficient industrial land; Based on the reference set weight coefficients, the actual output value, and the efficiency score, the relaxation variables of the inefficient industrial land in the four output dimensions of economic benefits, environmental benefits, development potential, and spatial benefits are determined, and then the relaxation efficiencies are determined.

[0012] In this embodiment, the dominant types of inefficient industrial land include: spatially inefficient, output-inefficient, and strategy-mismatched.

[0013] In this embodiment, based on the dominant type, differentially matching and outputting the corresponding strategy tool combination from the preset strategy tool library specifically includes: Based on the dominant type, retrieve matching strategy tool categories from the preset strategy tool library; Select specific strategy tools applicable to the aforementioned inefficient industrial land from each strategy tool category; The selected specific strategy tools are combined to form a strategy tool combination for the inefficient industrial land and then output.

[0014] Secondly, this application provides an intelligent identification system for inefficient industrial land based on multi-dimensional indicator fusion, used to execute an intelligent identification method for inefficient industrial land based on multi-dimensional indicator fusion, the intelligent identification system comprising: The data acquisition module is used to acquire raw data on industrial land use from multiple source platforms; An inefficiency identification module is used to determine the comprehensive score of the industrial land based on the original data of industrial land, and to identify inefficient industrial land according to the comprehensive score. The type determination module is used to establish an input-output dataset by constructing virtual input variables for the inefficient industrial land, thereby determining the various relaxation efficiencies of the inefficient industrial land, and determining the dominant type of the inefficient industrial land based on the various relaxation efficiencies. The strategy output module is used to differentiate and output the corresponding strategy tool combination from the preset strategy tool library based on the dominant type.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: The system acquires raw industrial land data from multiple sources; based on the raw industrial land data, it determines the comprehensive score of the industrial land and identifies inefficient industrial land according to the comprehensive score; for the inefficient industrial land, it establishes an input-output dataset by constructing virtual input variables, thereby determining the various relaxation efficiencies of the inefficient industrial land, and determining the dominant type of the inefficient industrial land according to each relaxation efficiency; based on the dominant type, it differentially matches and outputs the corresponding strategy tool combination from a preset strategy tool library.

[0016] Therefore, this application firstly achieves comprehensive coverage of industrial land across multiple dimensions by acquiring raw industrial land data in parallel from multiple heterogeneous platforms, providing a multi-source data foundation for subsequent identification of inefficient industrial land; secondly, it constructs an evaluation index system by extracting four-dimensional indicators from the raw industrial land data, objectively determines the weight of each indicator using the entropy weight method, calculates the comprehensive score, and finally identifies inefficient industrial land based on the comprehensive score, achieving a quantitative evaluation and objective ranking of the comprehensive performance of industrial land, thus improving the scientific rigor and accuracy of inefficient industrial land identification; then, by... Inefficient industrial land and other inefficient land within the same area form a reference group. The average output of other plots in the reference group is used as a virtual input variable to calculate the relaxation efficiency of each plot in four dimensions and determine the dominant type of the plot, providing a quantitative basis for subsequent differentiated strategy matching. Finally, by mapping the dominant type to the categories of strategy tools in the preset strategy tool library, specific strategy tools applicable to the inefficient industrial land are selected from the matched categories and combined for output, improving the pertinence and operability of strategy matching and providing a decision-making basis for the differentiated governance of inefficient industrial land.

[0017] In summary, the technical solution adopted in this application can achieve accurate identification and cause classification of inefficient industrial land, and improve the objectivity of identification and the pertinence of strategy matching. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this embodiment of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is an exemplary flowchart of the intelligent identification method for inefficient industrial land based on multi-dimensional indicator fusion provided in this application; Figure 2 This is a schematic diagram of a device scenario for an intelligent identification system for inefficient industrial land based on multi-dimensional indicator fusion, provided in this application. Figure 3 This is a module structure diagram of an intelligent identification system for inefficient industrial land based on multi-dimensional indicator fusion, provided in this application. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0021] This application provides a method and system for intelligent identification of inefficient industrial land based on multi-dimensional indicator fusion. Its core is to obtain raw industrial land data from a multi-source platform; based on the raw industrial land data, determine the comprehensive score of the industrial land, and identify inefficient industrial land according to the comprehensive score; for the inefficient industrial land, establish an input-output dataset by constructing virtual input variables, thereby determining the various relaxation efficiencies of the inefficient industrial land, and determining the dominant type of the inefficient industrial land based on each relaxation efficiency; based on the dominant type, differentially match and output the corresponding strategy tool combination from a preset strategy tool library.

[0022] Example 1: To better understand the above technical solution, the following will provide a detailed description of the technical solution in conjunction with the accompanying drawings and specific implementation methods. (Refer to...) Figure 1 As shown in the figure, this is an exemplary flowchart of an intelligent identification method for inefficient industrial land based on multi-dimensional indicator fusion according to this embodiment of the present application. The intelligent identification method includes the following steps: In step S1, raw data of industrial land use is obtained from a multi-source platform.

[0023] In this embodiment, obtaining raw industrial land data from a multi-source platform can be achieved through the following steps: Obtain land parcel boundary data, plot ratio data, average number of floors data, building quality data, and development intensity data for industrial land from the land registration management system; Data on energy consumption, water consumption, unit tax revenue and electricity costs, unit tax revenue and water costs, environmental quality, and emissions exceeding standards for industrial land are obtained from the environmental monitoring platform. Obtain data on tax revenue per mu, revenue per mu, and overall labor productivity of industrial land from tax and statistics departments. Data on the proportion of large-scale enterprises, key enterprises, and R&D investment in industrial land were obtained from the systems of the Ministry of Industry and Information Technology and the Ministry of Science and Technology. Obtain infrastructure level data for industrial land from an IoT monitoring platform; Obtain data on unlicensed land use identification for industrial land from the land and resources management system; All the above data are combined into raw data for industrial land use.

[0024] In practical implementation, the following data can be obtained from the land registration management system: land parcel boundary data, plot ratio data, average number of floors data, building quality data, and development intensity data for industrial land. Specifically, this can be achieved by connecting to the land registration management system of the target area via an interface call, and obtaining spatial attribute data for industrial land in batches based on the land parcel code. Similarly, energy consumption data, water consumption data, unit tax revenue and electricity fee data, unit tax revenue and water fee data, environmental quality data, and emission exceedance detection data for industrial land can be obtained from the environmental monitoring platform. This can be achieved by connecting to the environmental monitoring platform via a data exchange interface, and obtaining environmental attribute data for industrial land based on its geographical location information and enterprise discharge permit number. Furthermore, data on tax revenue per mu, revenue per mu, and overall labor productivity for industrial land can be obtained from the tax and statistics departments' systems. This can be achieved by connecting to the tax management system and statistics department's systems through the government data sharing platform, and obtaining economic benefit data for industrial land based on its unified social credit code. Finally, data on the proportion of large-scale enterprises, key enterprises, and R&D investment proportion for industrial land can be obtained from the industry and information technology departments' systems. This can be achieved by connecting to the industry and information technology management departments' systems via a data interface. The information system of the science and technology management department obtains data on the development potential of industrial land; it obtains infrastructure level data of industrial land from the Internet of Things (IoT) monitoring platform, that is, by connecting to various IoT sensors and monitoring devices deployed in industrial clusters through the IoT data access platform to obtain infrastructure level data of industrial land, including: road accessibility indicators, water and power supply guarantee rate, communication network coverage level, and the completeness of public supporting facilities; it obtains unlicensed land use identification data of industrial land from the land and resources management system, that is, by connecting to the land and resources management information system through the data exchange interface, and obtaining compliance identification data of industrial land based on the land parcel code and land use right certificate number; finally, all the above data are combined into the original industrial land data, that is, using the land parcel code as the unique primary key identifier, the various types of data are horizontally correlated, and for the same land parcel in different data sources, the data formats are inconsistent and the units of measurement are inconsistent, and the data is uniformly converted according to the preset data conversion rules. For numerical data, two decimal places are uniformly retained, and for categorical data, it is uniformly converted into preset coded values, thereby combining all the data into the original industrial land data.

[0025] In step S2, based on the original data of industrial land, the comprehensive score of the industrial land is determined, and inefficient industrial land is identified according to the comprehensive score.

[0026] In this embodiment, based on the original data of industrial land, a comprehensive score for the industrial land is determined, and inefficient industrial land is identified according to the comprehensive score. This can be achieved through the following steps: Spatial benefit indicators, economic benefit indicators, development potential indicators, and environmental benefit indicators are extracted from the original data of industrial land to construct a four-dimensional evaluation indicator system; The entropy weight method is used to determine the information entropy and information utility value of each indicator, and then the objective weight of each indicator is determined. Construct a weighted normalized matrix to determine the positive and negative ideal solutions, and then determine the distances from the industrial land to the positive and negative ideal solutions; The comprehensive score of the industrial land is determined based on the distances from the industrial land to the positive and negative ideal solutions.

[0027] In practice, firstly, spatial benefit indicators, economic benefit indicators, development potential indicators, and environmental benefit indicators are extracted from the original data of industrial land to construct a four-dimensional evaluation indicator system. This process will be elaborated in subsequent steps. Secondly, the entropy weight method is used to determine the information entropy and information utility value of each indicator, and then the objective weight of each indicator is determined. This process will also be elaborated in subsequent steps. Then, a weighted normalization matrix is ​​constructed to determine the positive and negative ideal solutions, and then the distance from the industrial land to the positive and negative ideal solutions is determined. That is, assuming there are N industrial land plots as evaluation objects, and each industrial land plot contains 14 evaluation indicators, the standardized matrix is ​​multiplied by the objective weight vector. That is, for each indicator of each industrial land plot, the standardized value of the indicator is multiplied by the objective weight corresponding to the indicator to obtain the weighted normalization value. The weighted normalization values ​​of all indicators of all evaluation objects together constitute the weighted normalization matrix. For each of the fourteen indicators, find the maximum weighted normalized value of that indicator from all N industrial land parcels. Combine these maximum values ​​into a vector according to the indicator order. This vector is the positive ideal solution, representing the virtual land parcel that performs best across all indicator dimensions. For each of the fourteen indicators, find the minimum weighted normalized value of that indicator from all N industrial land parcels. Combine these minimum values ​​into a vector according to the indicator order. This vector is the negative ideal solution, representing the virtual land parcel that performs worst across all indicator dimensions. Both the positive and negative ideal solutions are virtual vectors and do not correspond to any actual industrial land parcel. Instead, they serve as two benchmark reference points for measuring the relative performance of each industrial land parcel. Next, the weighted normalized value of the industrial land on the fourteen indicators is subtracted one by one from the corresponding indicator value in the positive ideal solution. The square of each difference is then summed, and the square root of the sum is taken as the distance of the industrial land to the positive ideal solution. Similarly, the weighted normalized value of the industrial land on the fourteen indicators is subtracted one by one from the corresponding indicator value in the negative ideal solution. The square of each difference is then summed, and the square root of the sum is taken as the distance of the industrial land to the negative ideal solution. Finally, the comprehensive score of the industrial land is determined based on the distances to both the positive and negative ideal solutions. That is, the distance to the negative ideal solution is divided by the sum of the distances to the positive and negative ideal solutions, and the result is taken as the comprehensive score of the industrial land. The comprehensive score ranges from 0 to 1. The higher the comprehensive score, the closer the industrial land is to the optimal benchmark and the further away it is from the worst benchmark, indicating better overall performance.

[0028] In this embodiment, spatial benefit indicators, economic benefit indicators, development potential indicators, and environmental benefit indicators are extracted from the original industrial land data to construct a four-dimensional evaluation indicator system. This is achieved through the following steps: Data on plot ratio, average number of floors, building quality, development intensity, and infrastructure level are extracted from the original data of the industrial land to form spatial efficiency indicators; The total labor productivity, tax revenue per mu, and revenue per mu are extracted from the original data of the industrial land to form economic benefit indicators. The proportion of enterprises above designated size, the proportion of R&D investment, and the proportion of key enterprises are extracted from the original industrial land data to form development potential indicators. The unit tax revenue electricity fee, unit tax revenue water fee, and environmental quality data are extracted from the original data of the industrial land to form environmental benefit indicators; The spatial benefit indicators, economic benefit indicators, development potential indicators, and environmental benefit indicators are correlated and combined to form a four-dimensional evaluation index vector for industrial land.

[0029] In practical implementation, firstly, data on plot ratio, average number of floors, building quality, development intensity, and infrastructure level can be extracted from the original industrial land data to form spatial efficiency indicators. Specifically, these data are linked according to land parcel codes to create spatial efficiency indicators for the industrial land. All five indicators are positive, meaning higher values ​​indicate better spatial efficiency. Secondly, data on total labor productivity, tax revenue per mu, and revenue per mu can be extracted from the original industrial land data to form economic efficiency indicators. Again, these three indicators are positive, meaning higher values ​​indicate better economic efficiency. Thirdly, data on the proportion of enterprises above designated size, R&D investment, and key enterprises can be extracted from the original industrial land data to form development potential indicators. This involves linking these data according to land parcel codes to create economic efficiency indicators for the industrial land. A development potential index vector is constructed for the industrial land. All three indicators mentioned above are positive, meaning the higher the value, the stronger the development potential. Then, unit tax revenue (electricity and water fees) and environmental quality data are extracted from the original industrial land data to form environmental benefit indicators. Specifically, these data are linked according to the land parcel code to form the environmental benefit index vector for the industrial land. Unit tax revenue (electricity and water fees) and unit tax revenue (water fees) are negative indicators, meaning the higher the value, the worse the environmental benefit; environmental quality is a positive indicator, meaning the higher the value, the better the environmental quality. Finally, the spatial benefit indicators, economic benefit indicators, development potential indicators, and environmental benefit indicators are linked and combined to form a four-dimensional evaluation index vector for the industrial land. This is achieved by using the land parcel code as the unique primary key identifier and horizontally concatenating the spatial benefit indicators, economic benefit indicators, development potential indicators, and environmental benefit indicators for the same land parcel, forming a vector containing 14 components. This vector is then used as the four-dimensional evaluation index vector for the industrial land.

[0030] In this embodiment, the entropy weight method is used to determine the information entropy and information utility value of each indicator, and then the objective weight of each indicator is determined. Specifically, this can be achieved through the following steps: The industrial land and all industrial land in its area are combined to form an evaluation object set, and a judgment matrix is ​​constructed. Standardize the data of each indicator in the judgment matrix; The information entropy of each indicator is determined based on the standardized data of each indicator. The information utility value is determined based on the information entropy of each indicator, and then the objective weight of each indicator is determined based on the information utility value of each indicator.

[0031] In practical implementation, firstly, the industrial land and all industrial land within its area constitute the evaluation object set, and a judgment matrix is ​​constructed. That is, assuming there are N industrial land parcels within the area where the target industrial land is located, all N industrial land parcels are included in the evaluation object set, with each parcel as an independent evaluation object. Each evaluation object contains the aforementioned 14 evaluation indicators. Based on this evaluation object set, a judgment matrix is ​​constructed. The judgment matrix is ​​an N-row, 14-column matrix, where each row corresponds to one industrial land parcel, and each column corresponds to one evaluation indicator. Each element in the judgment matrix represents the value of the corresponding industrial land parcel on the corresponding evaluation indicator. Secondly, the indicator data in the judgment matrix is ​​standardized. That is, for positive indicators (i.e., indicators where a larger value indicates better performance), the value of the i-th evaluation object on the j-th indicator is subtracted from the minimum value of that indicator among all evaluation objects, and then divided by the difference between the maximum and minimum values ​​of that indicator. This process is the standardization process. For negative indicators (i.e., indicators where a smaller value indicates better performance),... The standardization process involves subtracting the original value of the i-th evaluation object for that indicator from the maximum value of the indicator across all evaluation objects, and then dividing by the difference between the maximum and minimum values ​​of the indicator. Next, the information entropy of each indicator is determined based on the standardized data; that is, the information entropy of each indicator is calculated using an information entropy algorithm. Finally, the information utility value is determined based on the information entropy of each indicator, and then the objective weight of each indicator is determined based on the information utility value. This is achieved by subtracting the information entropy of each indicator from 1, and using the result as the information utility value of each indicator. The information utility value reflects the indicator's ability to provide effective information. Finally, the information utility value of the corresponding indicator is divided by the sum of the information utility values ​​of all indicators, and the result is used as the objective weight of that indicator, thus obtaining the objective weights of each indicator.

[0032] In practice, inefficient industrial land is determined based on the comprehensive score. That is, the comprehensive scores of all industrial land in the target area are sorted in ascending order, and the comprehensive scores of all industrial land are divided into five levels using the natural discontinuity method. Industrial land with a comprehensive score falling into the first level is judged as inefficient industrial land.

[0033] In step S3, for the inefficient industrial land, an input-output dataset is established by constructing virtual input variables, thereby determining the relaxation efficiencies of the inefficient industrial land and determining the dominant type of the inefficient industrial land based on the relaxation efficiencies.

[0034] In this embodiment, for the inefficient industrial land, an input-output dataset is established by constructing virtual input variables, which can be achieved through the following steps: The aforementioned inefficient industrial land, together with other plots in the same area that have been identified as inefficient industrial land, constitute a reference group; The actual output values ​​of the inefficient industrial land in four dimensions—economic benefits, environmental benefits, development potential, and spatial benefits—are determined to form an actual output vector. Determine the virtual input value of the inefficient industrial land; The virtual input values ​​are paired with the actual output vectors to establish an input-output dataset.

[0035] In practical implementation, firstly, inefficient industrial land can be grouped with other plots within the same area that have been identified as inefficient industrial land to form a reference group. The construction of the reference group follows these principles: all plots within the reference group are identified inefficient industrial land and are homogeneous, meaning they all exhibit varying degrees of inefficiency; the reference group is limited to the same area to ensure that each plot faces similar strategic environments, market conditions, and factor price levels; and the reference group should contain at least two plots to ensure the statistical validity of subsequent calculations of virtual input variables. Secondly, the actual output values ​​of inefficient industrial land in four dimensions—economic benefits, environmental benefits, development potential, and spatial benefits—are determined to form an actual output vector. Specifically, for the economic benefits dimension, the total labor productivity, tax revenue per mu, and revenue per mu of the inefficient industrial land are weighted and summed according to preset weights to obtain the actual economic output value of the inefficient industrial land. The preset weights can be set based on expert experience. For the environmental benefits dimension, the three raw data points of unit tax revenue (electricity), unit tax revenue (water), and environmental quality of the inefficient industrial land are weighted and summed according to the preset weights to obtain the actual environmental benefits output value of the inefficient industrial land. It should be noted that since unit tax revenue (electricity) and unit tax revenue (water) are negative indicators, they have been positiveized before weighted summation, i.e., 1 is used to subtract the negative indicators. For the development potential dimension, the three raw data points of the inefficient industrial land (the proportion of enterprises above designated size, the proportion of R&D investment, and the proportion of key enterprises) are weighted and summed according to the preset weights to obtain the actual development potential output value of the inefficient industrial land. For the spatial benefits dimension, the five raw data points of the inefficient industrial land (plot ratio, average number of floors, building quality, development intensity, and infrastructure level) are weighted and summed according to the preset weights to obtain the actual spatial benefits output value of the inefficient industrial land. The actual output values ​​of the above four dimensions are then combined in the order of economic benefits, environmental benefits, development potential, and spatial benefits to form the actual output vector of the inefficient industrial land. The preset weights can be set according to expert advice.

[0036] Furthermore, in practical implementation, the virtual input value of inefficient industrial land is determined. That is, for the economic benefit dimension, the arithmetic mean of the actual economic benefit output values ​​of all inefficient industrial land in the reference group, excluding the target plot, is calculated. This arithmetic mean is used as the virtual input value of the target plot in the economic benefit dimension. Based on the above method, the virtual input values ​​in the other three dimensions can be calculated. The virtual input values ​​of the above four dimensions are combined in the order of economic benefit, environmental benefit, development potential, and spatial benefit to form the virtual input vector of the inefficient industrial land. The virtual input value is used to replace the actual input variable. Finally, the virtual input value of inefficient industrial land is paired with the actual output vector to establish an input-output dataset. That is, the input-output dataset contains two parts: the input part is the virtual input vector, which contains four virtual input values, corresponding to the virtual input of economic benefit, virtual input of environmental benefit, virtual input of development potential, and virtual input of spatial benefit, respectively. The output component is an actual output vector containing four actual output values, corresponding to actual economic benefits, actual environmental benefits, actual development potential, and actual spatial benefits, respectively. The input and output components are then linked and stored according to land parcel codes to form an input-output dataset.

[0037] In this embodiment, the relaxation efficiencies of the inefficient industrial land are determined, which can be achieved through the following steps: Using the virtual input vector of the input-output dataset as the input variable and the actual output vector as the output variable, a linear programming equation is constructed. Solve the linear programming equation to obtain the efficiency score and reference set weight coefficient of the inefficient industrial land; Based on the reference set weight coefficients, the actual output value, and the efficiency score, the relaxation variables of the inefficient industrial land in the four output dimensions of economic benefits, environmental benefits, development potential, and spatial benefits are determined, and then the relaxation efficiencies are determined.

[0038] In practical implementation, firstly, the virtual input vector of the input-output dataset can be used as the input variable, and the actual output vector as the output variable to construct a linear programming equation. That is, the constructed linear programming equation contains an objective function and several constraints. The objective function is set to maximize the efficiency score of the inefficient industrial land. The first constraint is an output constraint: the weighted combination of the weight coefficients of each plot within the reference group and the output value of each plot should not be less than the product of the efficiency score of the inefficient industrial land and its actual output value. The second constraint is an input constraint: the weighted combination of the weight coefficients of each plot within the reference group and the input value of each plot should not be greater than the virtual input value of the inefficient industrial land. The third constraint is a non-negativity constraint: all reference set weight coefficients are not less than zero, thus constructing the linear programming equation. Then, the linear programming equation is solved to obtain the efficiency score of the inefficient industrial land and the reference set weight coefficients. The solution process can be implemented through programming, such as calling the corresponding optimization solution library in a Python environment. The solution results include: an efficiency score, which is a value greater than or equal to 1. When the efficiency score is equal to 1, it indicates that the inefficient industrial land is located on the production frontier, meaning that the output has reached its optimum at the current input level. When the efficiency score is greater than 1, it indicates that the output of the inefficient industrial land is insufficient relative to the frontier. Reference set weight coefficients, which indicate which plots in the reference set constitute the production frontier. The larger the weight coefficient, the higher the importance of the corresponding plot as a reference benchmark. Finally, based on the reference set weight coefficients, actual output values, and efficiency scores, the relaxation variables of the inefficient industrial land on the four output dimensions of economic benefits, environmental benefits, development potential, and spatial benefits are determined. The relaxation efficiency is determined by multiplying the optimal weight coefficient of each plot within the reference group by the economic output value of each plot, summing the results, and then subtracting the product of the optimal efficiency score of the inefficient industrial land and its actual economic output value. This yields the economic efficiency relaxation variable, which reflects the additional economic output that can be increased beyond proportional expansion in the economic efficiency dimension. The same method can be used to obtain the corresponding relaxation variables for the other three dimensions. In data envelopment analysis, when the evaluated object is in an inefficient state, the additional output that can be increased beyond proportional expansion of all outputs still exists in each output dimension. For the economic efficiency dimension, the actual economic output value of the inefficient industrial land is divided by the sum of the actual economic output value and the economic efficiency relaxation variable. The result is taken as the economic efficiency relaxation efficiency. The same method can be used to obtain the corresponding relaxation efficiencies for the other three dimensions, thus obtaining the relaxation efficiency for each dimension. Relaxation efficiency is a relative indicator used to measure the degree of closeness between the actual output level and the optimal output level of a certain output dimension. The formula for calculating relaxation efficiency is: actual output value divided by the sum of actual output value and relaxation variables.

[0039] In practical implementation, the dominant type of inefficient industrial land is determined based on various relaxation efficiencies. That is, the dominant type of inefficient industrial land can be determined according to preset mapping rules. Specifically, if the economic benefit relaxation efficiency is the smallest among the four relaxation efficiency values, it indicates that the land parcel has the most severe output deficiency in the economic benefit dimension, and the dominant type of inefficient industrial land is determined to be output inefficient. If the spatial benefit relaxation efficiency is the smallest among the four relaxation efficiency values, it indicates that the land parcel has the most severe output deficiency in the spatial benefit dimension, and the dominant type of inefficient industrial land is determined to be spatial inefficient. If the environmental benefit relaxation efficiency or the development potential relaxation efficiency is the smallest among the four relaxation efficiency values, it indicates that the land parcel has the most severe output deficiency in the environmental benefit dimension or the development potential dimension, and the dominant type of inefficient industrial land is determined to be strategy mismatch. The determined dominant types are recorded and output. The output results include: the plot code of the inefficient industrial land, the specific values ​​of the four relaxation efficiencies, the dimension corresponding to the minimum value, and the final determined dominant type. The dominant types of inefficient industrial land include: spatial inefficiency, output inefficiency, and strategy mismatch.

[0040] In step S4, based on the dominant type, a corresponding combination of strategy tools is differentially matched from a preset strategy tool library and output.

[0041] In this embodiment, based on the dominant type, differentially matching and outputting the corresponding strategy tool combination from the preset strategy tool library can be achieved through the following steps: Based on the dominant type, retrieve matching strategy tool categories from the preset strategy tool library; Select specific strategy tools applicable to the aforementioned inefficient industrial land from each strategy tool category; The selected specific strategy tools are combined to form a strategy tool combination for the inefficient industrial land and then output.

[0042] In practical implementation, firstly, matching strategy tool categories are retrieved from a pre-defined strategy tool library based on the dominant type. This pre-defined strategy tool library is a structured tool library constructed after systematically coding and analyzing strategy documents related to inefficient industrial land use by governments at all levels using grounded theory. The strategy tools in this library are divided into five main categories according to their functional attributes: planning and spatial governance, property rights and exit mechanisms, strategy incentives and constraints, market and multi-party participation, and dynamic monitoring and evaluation. When the dominant type of inefficient industrial land is spatial inefficiency, matching planning and spatial governance tools are retrieved from the preset strategy tool library; when the dominant type of inefficient industrial land is output inefficiency, matching incentive and constraint tools and property rights exit tools are retrieved from the preset strategy tool library; when the dominant type of inefficient industrial land is strategy mismatch, matching dynamic supervision tools and market participation tools are retrieved from the preset strategy tool library. Then, specific strategy tools applicable to inefficient industrial land are selected from each strategy tool category. That is, for spatially inefficient land, the specific strategy tools that can be selected include: floor area ratio improvement tools, which allow the plot to increase the building floor area ratio while complying with planning; mixed-use development tools, which allow the plot to accommodate multiple functions such as industry, R&D, and supporting facilities; and consolidation development tools, which allow adjacent plots to be merged for overall development. For land with low output, specific strategic tools include: per-acre efficiency assessment tools, i.e., establishing an assessment system centered on per-acre tax revenue and per-acre income; differentiated resource allocation tools, i.e., allocating indicators such as land use, energy use, and pollution discharge based on differences in output efficiency; and land acquisition and storage tools, i.e., the government legally reclaiming the right to use inefficient land and re-granting or reorganizing it. For land with mismatched strategies, specific strategic tools include: flexible land supply tools, i.e., supplying land using flexible methods such as flexible-term land grants and lease-to-own arrangements; performance agreement tools, i.e., signing performance supervision agreements with land users, stipulating development and construction standards, input-output levels, and measures for handling breaches; and market-based transfer tools, i.e., encouraging the market-based transfer of inefficient industrial land through methods such as transfer, lease, and equity participation. During the screening process, one or more of the above tools can be selected and combined, and multi-factor judgments can be made based on the specific circumstances of the target plot.

[0043] Furthermore, in practical implementation, the selected strategy tools will be combined to form a strategy tool portfolio specifically for inefficient industrial land, which will then be output. This means prioritizing the inclusion of core strategy tools addressing the primary weaknesses to ensure the most significant issues are resolved. Supplementary strategy tools will then be added as needed to enhance governance effectiveness. The inclusion of redundant or conflicting strategy tools will be avoided to ensure internal consistency within the portfolio. For example, for a piece of land with low output efficiency, the strategy tool portfolio might include: a per-acre efficiency assessment tool as a binding tool to clearly define output efficiency improvement targets; and a differentiated resource allocation tool as an incentive tool to provide resource allocation incentives to enterprises that achieve the assessment targets. The output of this portfolio will include: the land parcel code, the primary type, the matching strategy tool category, and the name of the selected specific strategy tool.

[0044] like Figure 2 As shown in the diagram, this solution provides a device scenario schematic for an intelligent identification system for inefficient industrial land based on multi-dimensional indicator fusion. The target industrial land is connected to the inefficiency identification layer through multi-platform data collection and data fusion. The inefficiency identification layer uses the entropy weight method to calculate the comprehensive score and identify whether the target industrial land is inefficient industrial land. The inefficiency identification layer outputs a strategy tool combination based on the dominant type judgment, indicating that the corresponding strategy tool combination is differentially matched from the preset strategy tool library and output to the end user device.

[0045] Therefore, this application firstly achieves comprehensive coverage of industrial land across multiple dimensions by acquiring raw industrial land data in parallel from multiple heterogeneous platforms, providing a multi-source data foundation for subsequent identification of inefficient industrial land; secondly, it constructs an evaluation index system by extracting four-dimensional indicators from the raw industrial land data, objectively determines the weight of each indicator using the entropy weight method, calculates the comprehensive score, and finally identifies inefficient industrial land based on the comprehensive score, achieving a quantitative evaluation and objective ranking of the comprehensive performance of industrial land, thus improving the scientific rigor and accuracy of inefficient industrial land identification; then, by... Inefficient industrial land and other inefficient land within the same area form a reference group. The average output of other plots in the reference group is used as a virtual input variable to calculate the relaxation efficiency of each plot in four dimensions and determine the dominant type of the plot, providing a quantitative basis for subsequent differentiated strategy matching. Finally, by mapping the dominant type to the categories of strategy tools in the preset strategy tool library, specific strategy tools applicable to the inefficient industrial land are selected from the matched categories and combined for output, improving the pertinence and operability of strategy matching and providing a decision-making basis for the differentiated governance of inefficient industrial land.

[0046] In summary, the technical solution adopted in this application can achieve accurate identification and cause classification of inefficient industrial land, and improve the objectivity of identification and the pertinence of strategy matching.

[0047] Example 2: This application provides an intelligent identification system for inefficient industrial land based on multi-dimensional indicator fusion, referencing... Figure 3 As shown in the figure, this is a modular structure diagram of an intelligent identification system for inefficient industrial land based on multi-dimensional indicator fusion, according to this embodiment of the present application. The intelligent identification system includes: Data acquisition module 100 is used to acquire raw data on industrial land from multiple source platforms; The inefficiency identification module 200 is used to determine the comprehensive score of the industrial land based on the original industrial land data, and to identify inefficient industrial land according to the comprehensive score. The type determination module 300 is used to establish an input-output dataset by constructing virtual input variables for the inefficient industrial land, thereby determining the various relaxation efficiencies of the inefficient industrial land, and determining the dominant type of the inefficient industrial land based on the various relaxation efficiencies. The strategy output module 400 is used to differentiate and output the corresponding strategy tool combination from the preset strategy tool library based on the dominant type.

[0048] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0049] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-Erasable Programmable Read-Only Memory (EEPROM), compactdisc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0050] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

Claims

1. A method for intelligent identification of inefficient industrial land based on multi-dimensional indicator fusion, characterized in that, The intelligent recognition method includes the following steps: Obtain raw data on industrial land from multiple sources; Based on the original data of industrial land, the comprehensive score of the industrial land is determined, and inefficient industrial land is identified according to the comprehensive score. For the inefficient industrial land, an input-output dataset is established by constructing virtual input variables, thereby determining the various relaxation efficiencies of the inefficient industrial land, and judging the dominant type of the inefficient industrial land based on the various relaxation efficiencies. Based on the dominant type, a differential matching is performed from the preset strategy tool library, and the corresponding strategy tool combination is output.

2. The intelligent identification method for inefficient industrial land based on multi-dimensional indicator fusion as described in claim 1, characterized in that, Obtaining raw industrial land data from multi-source platforms specifically includes: Obtain land parcel boundary data, plot ratio data, average number of floors data, building quality data, and development intensity data for industrial land from the land registration management system; Data on energy consumption, water consumption, unit tax revenue and electricity costs, unit tax revenue and water costs, environmental quality, and emissions exceeding standards for industrial land are obtained from the environmental monitoring platform. Obtain data on tax revenue per mu, revenue per mu, and overall labor productivity of industrial land from tax and statistics departments. Data on the proportion of large-scale enterprises, key enterprises, and R&D investment in industrial land were obtained from the systems of the Ministry of Industry and Information Technology and the Ministry of Science and Technology. Obtain infrastructure level data for industrial land from an IoT monitoring platform; Obtain data on unlicensed land use identification for industrial land from the land and resources management system; All the above data are combined into raw data for industrial land use.

3. The intelligent identification method for inefficient industrial land based on multi-dimensional indicator fusion as described in claim 1, characterized in that, Based on the original data of industrial land use, the comprehensive score of the industrial land use is determined specifically as follows: Spatial benefit indicators, economic benefit indicators, development potential indicators, and environmental benefit indicators are extracted from the original data of industrial land to construct a four-dimensional evaluation indicator system; The entropy weight method is used to determine the information entropy and information utility value of each indicator, and then the objective weight of each indicator is determined. Construct a weighted normalized matrix to determine the positive and negative ideal solutions, and then determine the distances from the industrial land to the positive and negative ideal solutions; The comprehensive score of the industrial land is determined based on the distances from the industrial land to the positive and negative ideal solutions.

4. The intelligent identification method for inefficient industrial land based on multi-dimensional indicator fusion as described in claim 3, characterized in that, The spatial benefit indicators, economic benefit indicators, development potential indicators, and environmental benefit indicators are extracted from the original data of industrial land to construct a four-dimensional evaluation indicator system, which specifically includes: Data on plot ratio, average number of floors, building quality, development intensity, and infrastructure level are extracted from the original data of the industrial land to form spatial efficiency indicators; The total labor productivity, tax revenue per mu, and revenue per mu are extracted from the original data of the industrial land to form economic benefit indicators. The proportion of enterprises above designated size, the proportion of R&D investment, and the proportion of key enterprises are extracted from the original industrial land data to form development potential indicators. The unit tax revenue electricity fee, unit tax revenue water fee, and environmental quality data are extracted from the original data of the industrial land to form environmental benefit indicators; The spatial benefit indicators, economic benefit indicators, development potential indicators, and environmental benefit indicators are correlated and combined to form a four-dimensional evaluation index vector for industrial land.

5. The intelligent identification method for inefficient industrial land based on multi-dimensional indicator fusion as described in claim 3, characterized in that, The entropy weight method is used to determine the information entropy and information utility value of each indicator, and then the objective weight of each indicator is determined, specifically including: The industrial land and all industrial land in its area are combined to form an evaluation object set, and a judgment matrix is ​​constructed. Standardize the data of each indicator in the judgment matrix; The information entropy of each indicator is determined based on the standardized data of each indicator. The information utility value is determined based on the information entropy of each indicator, and then the objective weight of each indicator is determined based on the information utility value of each indicator.

6. The intelligent identification method for inefficient industrial land based on multi-dimensional indicator fusion as described in claim 1, characterized in that, For the aforementioned inefficient industrial land, the specific steps of establishing an input-output dataset by constructing virtual input variables include: The aforementioned inefficient industrial land, together with other plots in the same area that have been identified as inefficient industrial land, constitute a reference group; The actual output values ​​of the inefficient industrial land in four dimensions—economic benefits, environmental benefits, development potential, and spatial benefits—are determined to form an actual output vector. Determine the virtual input value of the inefficient industrial land; The virtual input values ​​are paired with the actual output vectors to establish an input-output dataset.

7. The intelligent identification method for inefficient industrial land based on multi-dimensional indicator fusion as described in claim 1, characterized in that, Determining the relaxation efficiencies of the aforementioned inefficient industrial land specifically includes: Using the virtual input vector of the input-output dataset as the input variable and the actual output vector as the output variable, a linear programming equation is constructed. Solve the linear programming equation to obtain the efficiency score and reference set weight coefficient of the inefficient industrial land; Based on the reference set weight coefficients, the actual output value, and the efficiency score, the relaxation variables of the inefficient industrial land in the four output dimensions of economic benefits, environmental benefits, development potential, and spatial benefits are determined, and then the relaxation efficiencies are determined.

8. The intelligent identification method for inefficient industrial land based on multi-dimensional indicator fusion as described in claim 1, characterized in that, The dominant types of inefficient industrial land use include: spatially inefficient, output-inefficient, and strategy-mismatched.

9. The intelligent identification method for inefficient industrial land based on multi-dimensional indicator fusion as described in claim 1, characterized in that, Based on the dominant type, the differential matching and output of corresponding strategy tool combinations from the preset strategy tool library specifically includes: Based on the dominant type, retrieve matching strategy tool categories from the preset strategy tool library; Select specific strategy tools applicable to the aforementioned inefficient industrial land from each strategy tool category; The selected specific strategy tools are combined to form a strategy tool combination for the inefficient industrial land and then output.

10. A smart identification system for inefficient industrial land based on multi-dimensional indicator fusion, used to execute the smart identification method for inefficient industrial land based on multi-dimensional indicator fusion as described in any one of claims 1 to 9, characterized in that, The intelligent recognition system includes: The data acquisition module is used to acquire raw data on industrial land use from multiple source platforms; An inefficiency identification module is used to determine the comprehensive score of the industrial land based on the original data of industrial land, and to identify inefficient industrial land according to the comprehensive score. The type determination module is used to establish an input-output dataset by constructing virtual input variables for the inefficient industrial land, thereby determining the various relaxation efficiencies of the inefficient industrial land, and determining the dominant type of the inefficient industrial land based on the various relaxation efficiencies. The strategy output module is used to differentiate and output the corresponding strategy tool combination from the preset strategy tool library based on the dominant type.

Images