Inefficient Industrial Land Identification and Classification System Based on Multidimensional Indicators

Abstract

This invention discloses a system for identifying and classifying inefficient industrial land based on multi-dimensional indicators, belonging to the field of land use classification technology. The method includes: acquiring raw data from a multi-source heterogeneous platform and cleaning and completing it; calculating multi-dimensional indicators of industrial land based on the data obtained from the data acquisition and preprocessing module; establishing a scale-adaptive mapping model to obtain unified-scale mapping values ​​for the multi-dimensional indicators of industrial land; and, based on historical labeled samples, sending the comprehensive score of industrial land into a gray zone judgment strategy, dynamically adjusting the gray zone boundary to obtain classification indicator weights, and summarizing the inefficient and efficient membership degrees, and determining inefficient land, efficient land, and gray land based on a set threshold. This invention solves the problems of multi-scale mismatch, hard threshold fragmentation, temporal lag, and weakened environmental compliance in traditional inefficient industrial land identification methods, achieving high-precision, interpretable, and sustainable inefficient identification and classification of industrial land.

Description

Technical Field

[0001] This invention relates to the field of land use classification technology, specifically to a system for identifying and classifying inefficient industrial land based on multidimensional indicators. Background Technology

[0002] In recent years, with the increasing demands for industrial transformation and upgrading and high-quality development, management departments of various industrial parks have placed higher expectations on the accurate assessment and scientific utilization of industrial land. Inefficient use of industrial land not only wastes a large amount of land resources but also leads to decreased returns on investment, increased environmental pollution risks, and limited urban spatial carrying capacity. Therefore, establishing a technical system capable of quickly and accurately identifying and classifying inefficient industrial land is of significant practical importance and application value for promoting industrial structure optimization, revitalizing existing land resources, and enhancing regional economic competitiveness.

[0003] Most schemes only perform simple stratification at the park level or plot level, which cannot take into account the land use efficiency under different spatial resolutions and is prone to scale mismatch problems.

[0004] Existing technologies generally use output density, energy intensity, and other metrics to binary divide land parcels into high-efficiency or low-efficiency categories, ignoring the continuity of indicators and the ambiguity of boundaries, which leads to frequent grade jumps or misjudgments of marginal land parcels.

[0005] Relying heavily on annual or quarterly statistical data, it cannot reflect short-term fluctuations caused by equipment maintenance, capacity expansion, seasonal projects, etc., and lacks the ability to respond to dynamic scenarios in real time.

[0006] To address this, the present invention provides a system for identifying and classifying inefficient industrial land based on multidimensional indicators. Summary of the Invention

[0007] The purpose of this invention is to provide a system for identifying and classifying inefficient industrial land based on multidimensional indicators, so as to solve the existing problems mentioned in the background art.

[0008] To achieve the above objectives, the present invention provides the following technical solution: a system for identifying and classifying inefficient industrial land based on multi-dimensional indicators, comprising:

[0009] The data acquisition and preprocessing module is used to acquire raw data from multi-source heterogeneous platforms and perform cleaning and completion.

[0010] The indicator acquisition and calculation module is used to calculate multi-dimensional indicators of industrial land based on the data obtained by the data acquisition and preprocessing module.

[0011] The adaptive scale mapping module is used to establish a scale adaptive mapping model to obtain multi-dimensional index mapping values ​​for industrial land at a unified scale.

[0012] The land use fuzzy classification module is used to send the comprehensive score of the industrial land to the gray zone judgment strategy based on historical labeled samples, dynamically adjust the gray zone boundary, obtain the classification index weights, summarize the inefficient membership degree and the efficient membership degree, and determine inefficient land use, efficient land use and gray land use according to the set threshold.

[0013] A further improvement of the present invention is that the data acquisition and preprocessing module includes:

[0014] Raw industrial data is obtained from land registration management systems, environmental monitoring platforms, traffic monitoring systems, remote sensing images, and IoT platforms, including plot boundaries, multi-temporal output value, emissions, heavy vehicle traffic, and soil heavy metal content.

[0015] KNN interpolation was used to fill in missing values ​​in the raw industrial data, and outliers were removed based on three times the standard deviation. All data were resampled to N grid scales and time series alignment was completed.

[0016] Plots of land used without permits or with excessive emissions will be subject to a veto and marked as inefficient land use, and will not be included in subsequent classification analysis.

[0017] A further improvement of this invention is that the multidimensional indicators include production equipment utilization rate, heavy vehicle traffic density, land connectivity, emission compliance deviation, soil carrying capacity health index, land idling rate, and energy consumption intensity.

[0018] The land connectivity is quantified using a graph theory model to determine the quality of link connectivity between land parcels and core logistics hubs; the node set of the study area is... Indicates land parcels and hubs, and boundary rights. Calculate the weighted shortest path distance from each node plot to each hub by multiplying the road grade by the average vehicle speed. , obtain connectivity , This represents the maximum weighted distance of the shortest path from each node to each hub;

[0019] The emission compliance deviation is measured by the ratio of the actual emissions from the site to the upper limit of the industry emission standard, taking into account chemical oxygen demand and exhaust gas emissions. and The results were obtained from the calculation of three indicators.

[0020] A further improvement of the present invention is that the adaptive scale mapping module includes a scale energy spectrum extraction unit, a regional scale mapping unit, and a multi-scale fusion unit;

[0021] The scale-based energy spectrum extraction unit includes index fields on each scale grid. Perform two-dimensional discrete wavelet decomposition to obtain the scale bandpass subband coefficients. ,in Let represent subbands ranging from finer to coarser; then the total energy of each subband is expressed as . ,Will Normalization yields the relative importance of the index in each sub-band; finding the index that makes... Largest subband This corresponds to the scale where the spatial statistical structure is most concentrated; it is used for local adjustments of the subsequent mapping function.

[0022] A further improvement of this invention is that the regional scale mapping unit includes establishing a scale-adaptive mapping model, and the specific construction process of the scale-adaptive mapping model includes:

[0023] In overlapping regions, that is, within the same geographical unit, any two scales coexist. and Data grids construct paired samples ;

[0024] Solving for any two scales using least squares and mapping function The largest subband in the scale energy spectrum extraction unit is extracted as the dominant scale for each region. Then, weights are introduced into the weighted least squares. The chained mapping transforms any two scales, and simultaneously outputs the scale confidence during the transformation. ;

[0025] For all indicators, dynamic and static parameter datasets are established. The dynamic parameter dataset includes equipment utilization and heavy vehicle traffic density. A time weight is assigned to each mapping sample in the dynamic parameter dataset. t represents the timestamp of the sample data. This indicates the current time at which the mapping calculation is performed. This represents the empirical decay rate.

[0026] A further improvement of this invention is that the multi-scale fusion unit includes extracting emission compliance deviation. The emission compliance deviation is mapped to a safety factor. Subsequently, scale confidence scores were extracted. With safety factor Multiply to obtain the corrected confidence level Finally, each indicator is mapped to the highest resolution and normalized, and then weighted and fused according to the modified confidence level to obtain the multi-dimensional indicator mapping value of industrial land. ,in, This represents the mapping function that transforms the k-th grid scale to the densest grid scale.

[0027] A further improvement of the present invention is that the land use fuzzy classification module includes an adaptive dual-parameter membership calculation unit, a de-ashing execution unit, and a membership degree comprehensive decision unit;

[0028] The adaptive dual-parameter membership calculation unit uses historically labeled typical high-efficiency and typical low-efficiency land parcels as samples, and estimates the Gaussian parameters of each quantitative indicator using the EM algorithm to obtain the i-th quantitative indicator. The adaptive two-parameter Gaussian fuzzy membership function is expressed as: , ;in, This is represented as an inefficient membership function. Represents an efficient membership function. These represent the mean and standard deviation of typical inefficient and typical efficient samples, respectively.

[0029] A further improvement of this invention is that the land use fuzzy classification module is also equipped with the gray zone judgment strategy, including calculating the adaptive two-parameter Gaussian fuzzy membership function for all indicators, if and If only the inefficient membership function of index i is retained, it will be sent to the membership comprehensive decision-making unit. and Then only the efficient membership function of index i is retained and sent to the membership comprehensive decision unit. and Then, both the inefficient and efficient membership functions are retained and sent to the membership comprehensive decision-making unit. and At the same time less than or equal to When this happens, it is marked as a gray area and sent to the de-graying execution unit. This indicates the initial gray band boundary set by the system.

[0030] A further improvement of this invention is that the de-ashing execution unit is used to retrieve the average membership degree of surrounding classified land parcels. The gray zone boundary of the j-th indicator is dynamically adjusted according to the following rules: Before running the membership-based comprehensive decision-making unit, if ,or If the result is positive, it indicates that indicator j has been removed from the gray area. At this point, the gray area judgment strategy will be used to determine the gray area. Updated to Then, in the same way, determine how index j should be sent to the membership degree comprehensive decision-making unit. and All greater than If so, then indicator j is still considered fuzzy, and indicator j continues to be defined by its own... and Send it to the membership degree comprehensive decision-making unit.

[0031] A further improvement of this invention is that the membership degree comprehensive decision-making unit includes the system's inefficient membership degree and efficient membership degree for each indicator, with weights calculated using the entropy method. Perform weighted aggregation, represented as , ,like If it is determined to be inefficient industrial land, then it is considered as such. If it is determined to be high-efficiency industrial land, it will be left in the gray area and a manual review will be triggered. This indicates the set scoring threshold.

[0032] Compared with the prior art, the beneficial effects of the present invention are:

[0033] 1. This invention first extracts the wavelet energy spectrum of the index field at each grid scale through multi-scale spectral decomposition and linear mapping, and automatically identifies the dominant spatial frequency band with the most information content, so that each index at different resolutions can be accurately mapped to a unified fine scale, and completely eliminates the deviation caused by scale inconsistency.

[0034] 2. For indicators with obvious time-varying characteristics, an exponentially decaying time weight is adopted to highlight recent data and weaken outdated information, so that the model can reflect the enterprise's capacity adjustment and maintenance status in real time; and environmental safety factors are introduced in the fusion stage to give discounts or rejection to plots with excessive emissions and to give appropriate rewards to clean plots, so as to achieve a dual balance between classification accuracy and environmental compliance.

[0035] 3. By using an adaptive dual-parameter Gaussian membership function, each index value is continuously mapped to the membership degree on either the inefficient or efficient side. Then, with the help of gray zone adaptive calibration, the uncertainty interval is dynamically tightened or relaxed by neighborhood consensus to perform de-graying judgment. Indicators that are obviously biased to one side are quickly classified as deterministic evidence, while fuzzy indicators continue to participate in the pending determination. Attached Figure Description

[0036] Figure 1 This is a framework diagram of the inefficient industrial land identification and classification system based on multi-dimensional indicators of the present invention. Detailed Implementation

[0037] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations thereof. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0038] The term "and / or" is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three cases: A exists alone, A and B exist simultaneously, and B exists alone.

[0039] Example 1

[0040] Figure 1 This embodiment illustrates the framework diagram of the inefficient industrial land identification and classification system based on multi-dimensional indicators disclosed in this embodiment, including:

[0041] The data acquisition and preprocessing module is used to acquire raw data from multi-source heterogeneous platforms and perform cleaning and completion.

[0042] The data acquisition and preprocessing module includes:

[0043] Raw industrial data is obtained from land registration management systems, environmental monitoring platforms, traffic monitoring systems, remote sensing images, and IoT platforms, including land parcel boundaries, multi-temporal output value, energy consumption, water consumption, emissions, heavy vehicle traffic, soil heavy metals, and raw data on safety accidents.

[0044] KNN interpolation was used to fill in missing values ​​in the original data, and outliers were removed based on three times the standard deviation.

[0045] All data are resampled to N grid scales and time series alignment is performed; for example, including park-level s1 (500m×500m), plot-level s2 (100m×100m), and fine-scale s3 (20m×20m).

[0046] Plots of land used without permits or with excessive emissions will be subject to a veto and marked as inefficient land use, and will not be included in subsequent classification analysis.

[0047] The indicator acquisition and calculation module is used to calculate multi-dimensional indicators of industrial land based on the data obtained by the data acquisition and preprocessing module.

[0048] The multidimensional indicators include production equipment utilization rate, heavy vehicle traffic density, land connectivity, emission compliance deviation, soil carrying capacity health index, land idling rate, and energy consumption intensity.

[0049] The land connectivity is quantified using a graph theory model to determine the connectivity quality between land parcels and core logistics hubs, including ports, railway freight yards, and logistics parks; the study area node set is used to... Indicates land parcels and hubs, and boundary rights. Calculate the weighted shortest path distance from each node plot to each hub by multiplying the road grade by the average vehicle speed. , obtain connectivity , This represents the maximum weighted distance of the shortest path from each node to each hub;

[0050] The emission compliance deviation is measured by the ratio of the actual emissions from the site to the upper limit of the industry emission standard, taking into account the chemical oxygen demand (COD) and exhaust gas emissions. and The results were obtained from the calculation of three indicators.

[0051] ;

[0052] If there are excessive emissions, the system can use this parameter to either veto the emission or significantly reduce the efficiency membership in the membership calculation.

[0053] The production equipment utilization rate is obtained by the ratio of actual operating hours to available operating hours per unit time; it directly reflects the efficiency of production resource utilization within the plot and is closer to the actual situation of industrial production than the plot ratio.

[0054] The heavy vehicle traffic density is obtained by dividing the total number of heavy trucks (including trucks and trailers) passing through per unit time by the ratio of road length to the perimeter of the plot; it reflects logistics efficiency and road carrying capacity. A high value may indicate high logistics efficiency, or it may indicate potential traffic congestion and noise pollution.

[0055] The soil carrying capacity health index is obtained by weighted geometric mean based on the ratio of heavy metal (Pb, Cd, Cr) concentrations in on-site soil samples to background values.

[0056] Land idling rate through The energy consumption intensity is obtained through... get.

[0057] The adaptive scale mapping module is used to establish a scale adaptive mapping model to obtain multi-dimensional index mapping values ​​for industrial land at a unified scale.

[0058] The adaptive scale mapping module includes a scale energy spectrum extraction unit, a regional scale mapping unit, and a multi-scale fusion unit.

[0059] The scale-based energy spectrum extraction unit includes index fields on each scale grid. Perform two-dimensional discrete wavelet decomposition (such as Daubechies-4) to obtain the scale bandpass subband coefficients. ,in Let represent subbands ranging from finer to coarser; then the total energy of each subband is expressed as . ,Will After normalization, the relative importance of the index in each sub-band (corresponding to the spatial frequency range) can be obtained;

[0060] Searching for Largest subband This corresponds to the scale where the spatial statistical structure is most concentrated; it is used for local adjustments of the subsequent mapping function.

[0061] This embodiment achieves automatic quantification of information content across different spatial frequency ranges (from fine-grained to overall trends) by performing Daubechies 4 two-dimensional wavelet decomposition on the index field at each grid scale, calculating and normalizing the energy of each subband, and identifying the dominant subband. This eliminates the interference of gridding noise on the mapping, ensuring that subsequent mapping functions focus on the true spatial structure of the index, thereby improving mapping accuracy and robustness.

[0062] The regional scale mapping unit includes establishing a scale-adaptive mapping model, and the specific construction process of the scale-adaptive mapping model includes:

[0063] In overlapping regions, that is, within the same geographical unit, any two scales coexist. and Data grids construct paired samples ;

[0064] Solving for any two scales using least squares and mapping function The largest subband in the scale energy spectrum extraction unit is extracted as the dominant scale for each region. Then, weights are introduced into the weighted least squares. The chained mapping transforms any two scales, and simultaneously outputs the scale confidence during the transformation. It quantifies the intensity of changes in different spatial frequencies (from fine to coarse) and automatically identifies which spatial granularities contain the most information. At the same time, it reduces the weight of subbands with weak energy, which may be noise, to a very low level, and the subsequent mapping focuses on the true structure at the dominant scale.

[0065] As enterprises adjust their production capacity, perform equipment maintenance, or expand production, the latest operational data often better reflects the current true utilization status of a plot of land, while earlier data may be outdated. Therefore, dynamic and static parameter datasets are established for all indicators. The dynamic parameter dataset includes equipment utilization rate and heavy vehicle traffic density. Each mapping sample in the dynamic parameter dataset is assigned a time weight. t represents the timestamp of the sample data. This indicates the current time at which the mapping calculation is performed. This represents the empirical decay rate; amplifying its effect in least-squares fitting allows the mapping function to more accurately capture the truly important spatial patterns between scales, rather than being affected by noise or minute details. Through chain mapping, any two scales can be converted to each other.

[0066] This embodiment constructs scale-paired samples in overlapping regions, uses weighted least squares estimation of the linear mapping function with the dominant band energy as the weight, and outputs the confidence level of the mapping residual transformation to build an interpretable mapping model from arbitrary scale to target fine scale. It also quantifies the confidence level of each scale mapping, eliminates the bias caused by data inconsistency under different spatial resolutions, and ensures that the fusion results are comparable and reliable across the entire grid.

[0067] Secondly, by assigning exponential time weights to dynamic indicators (equipment utilization rate, heavy vehicle traffic flow), the importance of recent operating data in the mapping fit is highlighted, while the impact of outdated data is weakened. This ensures that the mapping model reflects enterprise capacity adjustments and equipment status changes in a timely manner, improving the mapping's adaptability to time-varying scenarios.

[0068] The multi-scale fusion unit includes extracting emissions compliance deviations. Environmental compliance is the bottom line for evaluating industrial land use. If the emissions of a piece of land exceed the standards, it should not be overestimated because other indicators are excellent. Therefore, this invention not only rejects severely non-compliant plots in the pre-processing stage, but also appropriately discounts the contribution of plots with slight non-compliance or close to the threshold in the indexing and fusion stage. Thus, the emission compliance deviation is mapped to a safety factor. By incorporating environmental risks into the confidence level, non-compliant plots can be automatically vetoed or downplayed without sacrificing accuracy.

[0069] when When, it means that the standard has been met. =1, no attenuation; when When this occurs, it indicates that emissions exceed standards. The more severe the exceedance, the smaller the factor; when When this is the case, it indicates low emissions, and appropriate incentives will be given for clean land use;

[0070] This two-way adjustment not only penalizes those exceeding standards but also slightly weights plots with excellent environmental performance. Nonlinear decay makes the effects of exceeding the standard more prominent—slight exceedances have little impact, while severe exceedances have a large impact. The factor remains constant. The changes are smooth within the range, avoiding abrupt changes at the boundary, and are easy to interpret.

[0071] Subsequently, scale confidence was extracted. With safety factor Multiply to obtain the corrected confidence level Finally, each indicator is mapped to the highest resolution and normalized, and then weighted and fused according to the modified confidence level to obtain the multi-dimensional indicator mapping value of industrial land. ,in, This represents the mapping function that transforms the k-th grid scale to the densest grid scale.

[0072] The classification boundary between inefficient and efficient land use is not absolute, and the index values ​​have continuous characteristics. How to set a "gray zone" or flexible membership degree (fuzzy membership function) is a difficult problem. Therefore, a land use fuzzy classification module is set up to send the comprehensive score of industrial land use to the gray zone judgment strategy based on historical labeled samples, dynamically adjust the gray zone boundary, obtain the classification index weights, and summarize the inefficient membership degree and efficient membership degree. Based on the set threshold, inefficient land use, efficient land use, and gray land use are determined.

[0073] This embodiment considers the credibility of each scale while incorporating environmental compliance levels into the overall weighting, achieving a comprehensive score that balances quantitative risk and value. This avoids the risk of environmentally non-compliant land parcels being masked by other excellent indicators, while also preventing the complete negation of the comprehensive value of environmentally excellent land parcels, thus enhancing the fairness and compliance of the comprehensive assessment.

[0074] The land use fuzzy classification module includes an adaptive dual-parameter membership calculation unit, a de-ashing execution unit, and a membership degree comprehensive decision-making unit;

[0075] The adaptive dual-parameter membership calculation unit uses historically labeled typical high-efficiency and typical low-efficiency land parcels as samples, and estimates the Gaussian parameters of each quantitative indicator using the EM algorithm to obtain the i-th quantitative indicator. The adaptive two-parameter Gaussian fuzzy membership function is expressed as: , ;in, This is represented as an inefficient membership function. Represents an efficient membership function. These represent the mean and standard deviation of typical inefficient and typical efficient samples, respectively, which are automatically estimated from historical labeled data using the EM algorithm.

[0076] The land use fuzzy classification module also includes the gray zone judgment strategy, which calculates the adaptive two-parameter Gaussian fuzzy membership function for all indicators. and If only the inefficient membership function of index i is retained, it will be sent to the membership comprehensive decision-making unit. and Then only the efficient membership function of index i is retained and sent to the membership comprehensive decision unit. and If both are relatively certain, then both the inefficient and efficient membership functions are retained and sent to the membership comprehensive decision-making unit. and At the same time less than or equal to When this occurs, the classification of the land parcel by the indicator is considered uncertain—that is, it enters the gray zone, is marked as gray zone, and sent to the de-graying execution unit. This indicates the initial gray band boundary set by the system.

[0077] The de-ashing execution unit is used to retrieve the average membership degree of surrounding classified land parcels. The gray zone boundary of the j-th indicator is dynamically adjusted according to the following rules: ;

[0078] Before running the membership-based comprehensive decision-making unit, if ,or If the result is positive, it indicates that indicator j has been removed from the gray area. At this point, the gray area judgment strategy will be used to determine the gray area. Updated to Then, in the same way, determine how index j should be sent to the membership degree comprehensive decision-making unit. and All greater than If both are uncertain, then indicator j is still considered fuzzy, and indicator j continues to be determined by its own... and Send it to the membership degree comprehensive decision-making unit.

[0079] This embodiment determines whether to remove or retain ambiguity for each indicator under the guidance of initial or dynamic thresholds, and dynamically adjusts the gray band width based on the average membership degree of the surrounding area. Indicators that are extremely biased to one side are removed from the "gray band" and only strong membership degrees are retained, or the gray band is maintained when the neighborhood is ambiguous. This accelerates the decision convergence for "significant" evidence, while avoiding premature classification when global information is insufficient, thus balancing decision efficiency and robustness.

[0080] Through this adaptive mechanism that uses neighbors as a reference, the gray band interval can be automatically narrowed or widened based on local spatial clustering characteristics, thus achieving flexible tightening and widening of the classification boundary; when neighbors are consistent, the gray band is quickly narrowed to avoid isolated noise slowing down the judgment; when neighbors are ambiguous, the gray band is widened to prevent premature assertion.

[0081] The membership degree comprehensive decision-making unit includes the system's inefficient membership degree and efficient membership degree for each indicator, with weights calculated using the entropy method. Perform weighted aggregation, represented as , ,like If it is determined to be inefficient industrial land, then it is considered as such. If it is determined to be high-efficiency industrial land, it will be left in the gray area and a manual review will be triggered. This represents the set scoring threshold. The weights are calculated using the entropy method. The indicators should objectively reflect the amount of information they contain, and those with a larger amount of information will contribute more to the final judgment.

[0082] The inefficient industrial land identification and classification system based on multidimensional indicators also includes an online iteration module, which is used to incorporate manually reviewed and labeled gray zone plots and newly added time series data into the EM training set in the form of labeled samples, and retrain the Gaussian parameters and mapping coefficients in real time or periodically.

[0083] Sub-model sets are divided according to administrative regions or industrial parks, and corresponding parameters are maintained separately to achieve regional adaptation; the GIS visualization interface is rendered with a three-color gradient of inefficient / gray / efficient, and supports retrospective and trend analysis of membership degree changes over time;

[0084] The threshold and weight settings can be set by default according to the present invention, or they can be set by those skilled in the art.

[0085] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0086] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. 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 flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0087] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0088] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0089] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A system for identifying and classifying inefficient industrial land based on multidimensional indicators, characterized in that: include: The data acquisition and preprocessing module is used to acquire raw data from multi-source heterogeneous platforms and perform cleaning and completion. The indicator acquisition and calculation module is used to calculate multi-dimensional indicators of industrial land based on the data obtained by the data acquisition and preprocessing module. The adaptive scale mapping module is used to establish a scale adaptive mapping model to obtain multi-dimensional index mapping values ​​for industrial land at a unified scale. The land use fuzzy classification module is used to send the comprehensive score of the industrial land to the gray zone judgment strategy based on historical labeled samples, dynamically adjust the gray zone boundary, obtain the classification index weights, and summarize the inefficient membership degree and efficient membership degree. Based on the set threshold, it judges inefficient land use, efficient land use and gray land use. The data acquisition and preprocessing module includes: Raw industrial data is obtained from land registration management systems, environmental monitoring platforms, traffic monitoring systems, remote sensing images, and IoT platforms, including plot boundaries, multi-temporal output value, emissions, heavy vehicle traffic, and soil heavy metal content. KNN interpolation was used to fill in missing values ​​in the raw industrial data, and outliers were removed based on three times the standard deviation. All data were resampled to N grid scales and time series alignment was completed. Plots of land used without permits or with excessive emissions will be subject to a veto and marked as inefficient land use, and will not be included in subsequent classification analysis; The multidimensional indicators include production equipment utilization rate, heavy vehicle traffic density, land connectivity, emission compliance deviation, soil carrying capacity health index, land idling rate, and energy consumption intensity. The land connectivity is quantified using a graph theory model to determine the quality of link connectivity between land parcels and core logistics hubs; the node set of the study area is... Indicates land parcels and hubs, and boundary rights. Calculate the weighted shortest path distance from each node plot to each hub by multiplying the road grade by the average vehicle speed. , obtain connectivity , This represents the maximum weighted distance of the shortest path from each node to each hub; The emission compliance deviation is measured by the ratio of the actual emissions from the site to the upper limit of the industry emission standard, taking into account chemical oxygen demand and exhaust gas emissions. and The results were obtained from the calculation of three indicators.

2. The inefficient industrial land identification and classification system based on multi-dimensional indicators according to claim 1, characterized in that: The adaptive scale mapping module includes a scale energy spectrum extraction unit, a regional scale mapping unit, and a multi-scale fusion unit. The scale-based energy spectrum extraction unit includes index fields on each scale grid. Perform two-dimensional discrete wavelet decomposition to obtain the scale bandpass subband coefficients. ,in Let represent subbands ranging from finer to coarser; then the total energy of each subband is expressed as . ,Will Normalization yields the relative importance of the index in each sub-band; finding the index that makes... Largest subband This corresponds to the scale where the spatial statistical structure is most concentrated; it is used for local adjustments of the subsequent mapping function.

3. The inefficient industrial land identification and classification system based on multidimensional indicators according to claim 2, characterized in that: The regional scale mapping unit includes establishing a scale-adaptive mapping model, and the specific construction process of the scale-adaptive mapping model includes: In overlapping regions, that is, within the same geographical unit, any two scales coexist. and Data grids construct paired samples ; Solving for any two scales using least squares and mapping function The largest subband in the scale energy spectrum extraction unit is extracted as the dominant scale for each region. Then, weights are introduced into the weighted least squares. The chained mapping transforms any two scales, and simultaneously outputs the scale confidence during the transformation. ; For all indicators, dynamic and static parameter datasets are established. The dynamic parameter dataset includes equipment utilization and heavy vehicle traffic density. A time weight is assigned to each mapping sample in the dynamic parameter dataset. t represents the timestamp of the sample data. This indicates the current time at which the mapping calculation is performed. This represents the empirical decay rate.

4. The inefficient industrial land identification and classification system based on multidimensional indicators according to claim 3, characterized in that: The multi-scale fusion unit includes extracting emissions compliance deviations. The emission compliance deviation is mapped to a safety factor. Subsequently, scale confidence scores were extracted. With safety factors Multiply to obtain the corrected confidence level Finally, each indicator is mapped to the highest resolution and normalized, and then weighted and fused according to the modified confidence level to obtain the multi-dimensional indicator mapping value of industrial land. ,in, This represents the mapping function that transforms the k-th grid scale to the densest grid scale.

5. The inefficient industrial land identification and classification system based on multidimensional indicators according to claim 4, characterized in that: The land use fuzzy classification module includes an adaptive dual-parameter membership calculation unit, a de-ashing execution unit, and a membership degree comprehensive decision-making unit; The adaptive dual-parameter membership calculation unit uses historically labeled typical high-efficiency and typical low-efficiency land parcels as samples, and estimates the Gaussian parameter of each quantitative indicator using the EM algorithm to obtain the first... i Quantitative indicators The adaptive two-parameter Gaussian fuzzy membership function is expressed as: , ;in, This is represented as an inefficient membership function. Represents an efficient membership function. These represent the mean and standard deviation of typical inefficient and typical efficient samples, respectively.

6. The inefficient industrial land identification and classification system based on multi-dimensional indicators according to claim 5, characterized in that: The land use fuzzy classification module also includes the gray zone judgment strategy, which calculates the adaptive two-parameter Gaussian fuzzy membership function for all indicators. and If only the inefficient membership function of index i is retained and sent to the membership comprehensive decision-making unit, then... and Then only the efficient membership function of index i is retained and sent to the membership comprehensive decision unit. and Then, both the inefficient and efficient membership functions are retained and sent to the membership comprehensive decision-making unit. and At the same time less than or equal to When this happens, it is marked as a gray area and sent to the de-graying execution unit. This indicates the initial gray band boundary set by the system.

7. The inefficient industrial land identification and classification system based on multidimensional indicators according to claim 6, characterized in that: The de-ashing execution unit is used to retrieve the average membership degree of surrounding classified land parcels. The gray zone boundary of the j-th indicator is dynamically adjusted according to the following rules: Before running the membership-based comprehensive decision-making unit, if ,or If the result is positive, it indicates that indicator j has been removed from the gray area. At this point, the gray area judgment strategy will be used to determine the gray area. Updated to Then, in the same way, determine how index j should be sent to the membership degree comprehensive decision-making unit. and All greater than If so, then indicator j is still considered fuzzy, and indicator j continues to be defined by its own... and Send it to the membership degree comprehensive decision-making unit.

8. The inefficient industrial land identification and classification system based on multidimensional indicators according to claim 7, characterized in that: The membership degree comprehensive decision-making unit includes the system's inefficient membership degree and efficient membership degree for each indicator, with weights calculated using the entropy method. Perform weighted aggregation, represented as , ,like If it is determined to be inefficient industrial land, then it is considered as such. If it is determined to be high-efficiency industrial land, it will be left in the gray area and a manual review will be triggered. This indicates the set scoring threshold.

Images