Industrial land intensive utilization evaluation system and method

Abstract

The invention discloses an industrial land intensive utilization evaluation system and method, and relates to the technical field of land resource management, space, economic and ecological data are collected in real time through a sensor, the intensive degree is calculated by adopting a dynamic weight model, the weight is adaptively adjusted in combination with an LSTM neural network, and a land optimization scheme is automatically generated based on a genetic algorithm. According to the industrial land intensive utilization evaluation system and method, the problems that a traditional evaluation method is single in index, lagged in data and disjointed in planning are solved, and the industrial land management efficiency is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of land resource management, and in particular to an industrial land intensive utilization evaluation system and method. Background Art

[0002] Industrial land refers to land used for independent factories, workshops, handicraft workshops, construction and installation production sites, and slag (ash) disposal sites. The layout of industrial land directly impacts a city's functional structure and urban form. In urban master plans, it is crucial to prioritize the allocation of industrial land and comprehensively consider the relationship between industrial land and other land uses, such as residential and transportation, ensuring that each is appropriately utilized.

[0003] Current status of industrial land management: Current industrial land assessments mainly rely on static indicators (such as volume ratio and investment intensity), which have three major flaws: Data lag: Relying on manual reporting and annual updates, it cannot reflect the land use status in real time; Single indicator: ignoring ecological efficiency (such as output value per unit of energy consumption) and spatial dynamic adaptability (such as the degree of industrial chain coordination); Planning disconnect: It is difficult to directly generate executable land use optimization plans based on the assessment results.

[0004] Existing technologies are unable to achieve closed-loop management of “data collection-dynamic evaluation-plan generation”, resulting in inefficient allocation of land resources. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the present invention provides an industrial land intensive utilization evaluation system and method, which solves the problem of inefficient land resource allocation caused by the inability to achieve closed-loop management of "data collection-dynamic evaluation-planning generation".

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: an industrial land intensive utilization evaluation system, including a data acquisition unit, a data processing unit, a dynamic evaluation unit, an optimization scheme generation unit and a visualization unit; The data acquisition unit includes: A three-dimensional space scanning device for obtaining three-dimensional contour information of buildings on industrial land; Energy consumption monitoring device, which collects real-time data on electricity and water consumption per unit area; Surface cover monitoring device, which obtains vegetation distribution information through multispectral imaging; The data processing unit performs spatiotemporal matching of spatial data, economic data, and ecological data, and calculates three basic indicators: space utilization rate, output value per unit area, and ecological efficiency ratio; The dynamic evaluation unit comprises: The weight prediction module dynamically generates weight coefficients based on historical land use data and policy text analysis; The intensity calculation module outputs the land use intensity evaluation value through weighted summation; The optimization solution generating unit includes: Case database, storing successful cases of industrial land transformation; Intelligent optimization engine generates land use adjustment plans with the goal of improving intensification; The visualization unit is used to display land use assessment results and planning schemes in three dimensions.

[0007] Preferably, the weight prediction module adopts a long short-term memory neural network, and its input data includes the basic indicator change sequence and the frequency of occurrence of policy keywords in the past 60 months.

[0008] Preferably, the eco-efficiency ratio is determined by the ratio of vegetation cover index to carbon dioxide emissions, wherein the vegetation cover index is obtained by multispectral image analysis.

[0009] Preferably, the intelligent optimization engine performs a genetic algorithm operation, and its solution coding structure includes coding segments of three dimensions: plot function type, building density and industrial correlation.

[0010] Preferably, during the execution of the genetic algorithm, the fitness function is defined as the weighted sum of the intensiveness evaluation value and the industrial chain coordination degree.

[0011] Preferably, it also includes a data storage unit for generating an unalterable electronic certificate from the collected original data and its time information.

[0012] The present invention also discloses an evaluation method for an industrial land intensive utilization evaluation system, comprising the following steps: Step 1: Collect building spatial data, enterprise economic data, and ecological environment data in the target area through the sensor network; Step 2: Calculate the spatial utilization efficiency index, unit area output index and ecological balance index; Step 3: Use machine learning models to analyze historical trends and policy directions and dynamically determine the weight coefficients of each indicator; Step 4: Compare the weighted intensive evaluation value with the preset threshold to identify inefficient land use; Step 5: Generate at least three land use adjustment plans based on the case library knowledge and optimization algorithm; Step 6: Visualize the comparison between the current situation and the planned solution through a 3D simulation platform.

[0013] Preferably, the training process of the machine learning model in step 3 includes: Use attention mechanism to identify key features; Construct a loss function by comparing the difference between the predicted weights and the actual weights; Introduce weight distribution balance constraints.

[0014] Preferably, the step five includes: (1) Randomly generate multiple sets of land use configuration scheme codes; (2) Calculate the comprehensive benefit value of each plan; (3) Iteratively optimize the solution through selection, crossover, and mutation operations; (4) Output the optimal solution set.

[0015] Preferably, the step six includes: The spatial distribution of intensiveness is presented in the form of a gradient color heat map; Use semi-transparent overlay technology to simultaneously display existing buildings and planned buildings.

[0016] How the system works 1. Data Acquisition Unit Workflow Three-dimensional space scanning device The system uses a pulsed laser ranging principle: a laser emits 905nm wavelength pulses, which are reflected off building surfaces and received by an APD avalanche diode. Distance is calculated using time-of-flight (ToF) and combined with the two-dimensional rotation of the scanning mirror (horizontally 0-360° / vertically -15° to +15°), a 3D point cloud with a point spacing of ≤0.1m is constructed.

[0017] Data usage: Extract building base area and floor height information → calculate space utilization rate I1.

[0018] Energy consumption monitoring device Based on the Hall Effect principle, a reciprocating current sensor (range 0-1000A) is installed on the three-phase circuit of the power distribution cabinet. The voltage signal is converted into a digital quantity by the AD7793 chip. This is transmitted to the gateway via the ZigBee protocol and converted into energy consumption per unit area according to the formula: standard coal consumption = (active energy × 0.1229) / plot area. Data usage: Calculate economic efficiency I2 in combination with enterprise output value.

[0019] Surface cover monitoring device Multispectral imaging principle: The drone's onboard imager simultaneously captures images in the blue (450nm), green (560nm), red (650nm), and near-infrared (840nm) bands. Vegetation cover is quantified using the Normalized Difference Vegetation Index (NDVI) algorithm: (NIR-Red) / (NIR+Red).

[0020] Data usage: Input for eco-efficiency ratio I3 calculation 2. Data processing unit workflow Space-time matching engine Establish a unified space-time coordinate system: Register the LiDAR point cloud to the WGS-84 coordinate system (RMS error < 0.3m using ground control points); (2) Energy consumption data is aligned to 0:00 every day by timestamp; (3) Multispectral images are superimposed on the point cloud through georegistration (registration error < 1 pixel).

[0021] Technical effect: Achieve "time and space matching".

[0022] Basic indicator calculation logic

[0023]

[0024] 3. Dynamic Assessment Unit Workflow Weight prediction module (LSTM) Time series data processing: Slice the [I1, I2, I3] sequence of the past 60 months by month to form a 60×3 matrix.

[0025] Policy text analysis: The TF-IDF algorithm is used to extract the frequency of keywords such as “floor area ratio”, “carbon emissions”, and “per mu output value” in policy documents to generate a 10-dimensional policy vector.

[0026] Dynamic weight output: The number of neurons in the LSTM hidden layer is 128, and the final layer Softmax activation outputs a weight vector of w1+w2+w3=1.

[0027] Intensity calculation and inefficiency determination Perform weighted summation: JI = w1×I1+w2×I2+w3×I3 When JI < threshold K (default K = 0.6), a low-efficiency land use alarm is triggered; Ecological constraint mechanism: If the GIS database identifies that the plot is located within the ecological red line, w3=Max(w3,0.4) is set forcibly.

[0028] 4. Optimize the solution generation unit workflow Case database query Use vector similarity retrieval: convert the current plot features (type = mechanical manufacturing, I1 = 0.41, I3 = 0.16) into a feature vector and match the top 3 similar cases in the knowledge base).

[0029] Genetic algorithm optimization engine Chromosome code: [Plot function (3 bits)] [Building density (8 bits)] [Industry correlation (4 bits)]; Example: 001011001000101 → Function = Warehousing, Density = 60%, Synergy = Level 5.

[0030] Fitness function: Fitness=JI+0.2×industrial chain synergy (industrial chain synergy is calculated by the logistics distance / raw material matching degree between enterprises).

[0031] Optimization iteration: Initialize 50 individuals → Selection (tournament size = 5) → Crossover (single-point Pc = 0.8) → Mutation (bit flip Pm = 0.05) → Iterate until convergence in 100 generations.

[0032] 5. Visualization of unit workflow Heatmap generation Grid the plot into 1m×1m cells and map them to the HSV color space according to the JI value: JI=0.0→H=0° (red); JI=1.0→H=120° (green); Output gradient raster layer.

[0033] Plan comparison view Using ray casting technology: The current status model is loaded as a gray entity; (2) The planning scheme is rendered as a blue semi-transparent object (transparency α = 0.6); (3) Visualization of the intersection part is achieved through depth buffering.

[0034] 6. Blockchain Evidence Storage Unit Data on-chain process: Sensor data - generate SHA-256 hash - call smart contract - write to Fabric block.

[0035] The evidence structure is as follows: {"blockNumber":10245, "timestamp":"2025-06-17T14:30:00Z", "dataHash":"0x8a3d...f2c1", "signature": "Enterprise A private key signature"}

[0036] Method execution principle Step 3: Principle of dynamic weight determination LSTM model training phase Input data: 5-year historical data of 100 industrial zones (a total of 100 × 60 = 6,000 time series records) Attention mechanism works: Calculate feature importance score: Score_i=V·tanh(W·h_i+b); Weighted context vector: C=Σ(Score_i·h_i).

[0037] Loss function optimization: Main loss term: Euclidean distance between the predicted weight vector W_pred and the actual optimal weight W_actual; Regularization term: KL divergence constrains the weight distribution to not deviate too much from (1 / 3, 1 / 3, 1 / 3) Step 5: Solution Generation Principle

[0038]

[0039] Step 6: Visualize the principle Translucent overlay technology Using the Alpha blending formula: C_out=α·C_plan+(1-α)·C_current; Where α=0.6 is the transparency of the planning scheme.

[0040] Heatmap rendering optimization Use GPU acceleration: (1) Pass the JI value into the CUDA kernel function; (2) Calculate the HSV value of each pixel in parallel; (3) Output to the three-dimensional scene through texture mapping.

[0041] The present invention provides an industrial land intensive utilization assessment system and method. Compared with the existing technology, it has the following advantages: This industrial land intensive utilization assessment system and method uses sensors to collect spatial, economic and ecological data in real time, adopts a dynamic weight model to calculate the intensity, combines an LSTM neural network to adaptively adjust weights, and automatically generates land optimization plans based on a genetic algorithm. It solves the problems of traditional assessment methods such as single indicators, data lags, and planning disconnection, and significantly improves the efficiency of industrial land management. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 Schematic diagram of the system of the present invention.

[0043] In the figure: 1. Data acquisition unit; 2. Data processing unit; 3. Dynamic evaluation unit; 4. Optimization solution generation unit; 5. Visualization unit. DETAILED DESCRIPTION

[0044] The technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0045] See also Figure 1 ,The embodiment of the present invention provides a technical solution: an industrial land intensive utilization evaluation system, comprising a data acquisition unit 1, a data processing unit 2, a dynamic evaluation unit 3, an optimization solution generation unit 4 and a visualization unit 5; Data acquisition unit 1, including: A three-dimensional space scanning device for obtaining three-dimensional contour information of buildings on industrial land; Energy consumption monitoring device, which collects real-time data on electricity and water consumption per unit area; Surface cover monitoring device, which obtains vegetation distribution information through multispectral imaging; Data processing unit 2 performs spatiotemporal matching of spatial data, economic data, and ecological data, and calculates three basic indicators: space utilization rate, output value per unit area, and ecological efficiency ratio; Dynamic Assessment Unit 3, including: The weight prediction module dynamically generates weight coefficients based on historical land use data and policy text analysis; The intensity calculation module outputs the land use intensity evaluation value through weighted summation; The optimization solution generating unit 4 includes: Case database, storing successful cases of industrial land transformation; Intelligent optimization engine generates land use adjustment plans with the goal of improving intensification; The visualization unit 5 is used to display the land use assessment results and planning schemes in three dimensions.

[0046] Furthermore, the weight prediction module adopts a long short-term memory neural network, whose input data includes the basic indicator change sequence and the frequency of occurrence of policy keywords in the past 60 months.

[0047] Furthermore, the eco-efficiency ratio is determined by the ratio of vegetation cover index to carbon dioxide emissions, where the vegetation cover index is obtained by multispectral image analysis.

[0048] Furthermore, the intelligent optimization engine performs genetic algorithm operations, and its solution coding structure includes coding segments of three dimensions: plot function type, building density, and industrial relevance.

[0049] Furthermore, during the execution of the genetic algorithm, the fitness function is defined as the weighted sum of the intensiveness evaluation value and the industrial chain coordination degree.

[0050] Furthermore, it also includes a data storage unit for generating an unalterable electronic certificate from the collected original data and its time information.

[0051] An embodiment of the present invention provides an evaluation method for an industrial land intensive utilization evaluation system, characterized by comprising the following steps: Step 1: Collect building spatial data, enterprise economic data, and ecological environment data in the target area through the sensor network; Step 2: Calculate the spatial utilization efficiency index, unit area output index and ecological balance index; Step 3: Use machine learning models to analyze historical trends and policy directions and dynamically determine the weight coefficients of each indicator; Step 4: Compare the weighted intensive evaluation value with the preset threshold to identify inefficient land use; Step 5: Generate at least three land use adjustment plans based on the case library knowledge and optimization algorithm; Step 6: Visualize the comparison between the current situation and the planned solution through a 3D simulation platform.

[0052] Furthermore, the training process of the machine learning model in step 3 includes: Use attention mechanism to identify key features; Construct a loss function by comparing the difference between the predicted weights and the actual weights; Introduce weight distribution balance constraints.

[0053] Furthermore, step five includes: (1) Randomly generate multiple sets of land use configuration scheme codes; (2) Calculate the comprehensive benefit value of each plan; (3) Iteratively optimize the solution through selection, crossover, and mutation operations; (4) Output the optimal solution set.

[0054] Furthermore, step six includes: The spatial distribution of intensiveness is presented in the form of a gradient color heat map; Use semi-transparent overlay technology to simultaneously display existing buildings and planned buildings.

[0055] Specific implementation method (Example: a plot of land in an industrial park) 1. Data acquisition unit 1 execution process 3D spatial scanning Using a Velodyne VLP-16 lidar (deployed on a 25m high monitoring tower), scanning parameters: Angular resolution: horizontal 0.1° / vertical 2°; Point cloud density: 300 points / ㎡.

[0056] Obtain point cloud data of three factory buildings and calculate: Total construction area: 8,200 m2 Land area: 15 mu = 10,000 m2 → Space utilization ratio I1 = 8,200 / 10,000 = 0.82 Energy consumption monitoring Install Honeywell SPT-WA02 sensor in the enterprise power distribution room: Monitoring period: January 1, 2025 - December 31, 2025; Total power consumption: 1,850,000 kWh → equivalent to 227.65 tons of standard coal.

[0057] Combined with the output value data of the tax system: Annual total output value: 126 million yuan → economic efficiency I2 = 126 million / 15 = 8.4 million yuan / mu.

[0058] Land cover monitoring DJI M300 drone performs aerial photography (parameters: flight altitude 100m, resolution 3cm): Multispectral analysis: NDVI mean = 0.35 (west green area); Fixed sensor: CO2 emissions = 28.5 tons / month.

[0059] Calculate eco-efficiency: Green area: 2,100 m2; Ecological efficiency ratio I3=(0.35×2,100) / (28.5×12)=0.215.

[0060] 2. Data processing unit 2 execution process Spatiotemporal matching operation Establish a unified space-time benchmark:

[0061] 3. Dynamic evaluation unit 3 execution process LSTM weight prediction input: Historical sequence (2020-2024): I1:[0.68,0.71,0.75,0.78,0.82]; I2:[520,640,730,790,840]; I3:[0.18,0.19,0.20,0.21,0.215]; Policy vector: [1,0,1] (including the keywords “industrial chain upgrading” and “low-carbon park”).

[0062] Weight output w1=0.45 (space utilization weight); w2=0.30 (economic efficiency weight); w3=0.25 (ecological efficiency weight); Ecological constraint triggered: Due to the ecological zone of the plot → forced w3=0.40.

[0063] Intensity calculation: JI=0.45×0.82+0.30×840+0.40×0.215=0.369+252+0.086=252.455; Judgment: Below the regional benchmark value of 280 → marked as Class II low-efficiency land.

[0064] 4. Optimization solution generation unit 4 execution process Case library matching:

[0065] Genetic algorithm optimization Chromosome encoding example: Function code: 010 (electronic manufacturing) Density code: 01100100 (building density 50%) Collaboration code: 1101 (collaboration level 13).

[0066] Fitness calculation: Fitness=JI+0.2×(logistics distance score+raw material matching score)=252.455+0.2×(0.7+0.8)=252.755.

[0067] Iterative optimization:

[0068] The plan is as follows Solution (recommended): 5,000 square meters of photovoltaic panels were installed on the factory roof (estimated to generate 600,000 kWh of electricity annually); Demolish the old warehouse in the northwest corner and build a new 8-story vertical factory building (floor area ratio → 1.35); Build an intelligent warehousing center with neighboring enterprise Y (shortening the logistics distance by 300m); The JI is forecast to increase to 293.566 (+16.3%).

[0069] 5. Visualization Unit 5 Execution Process Heatmap rendering Divide the plot into 10m×10m grids:

[0070] Plan comparison view Implemented using Unity engine: Current model: gray solid (0% transparency); Planning scheme: blue translucent body (transparency 60%); Overlay method: DepthPeeling algorithm handles occlusion.

[0071] 6. Blockchain Evidence Execution Process Contents of the evidence: { "blockHash":"0x4d7a...c3b9", "timestamp":"2025-06-17T14:30:00Z", "sensorData":{ "lidar":"sha256:8f3c...d2a1", "energy":"sha256:7e9b...4f0c" }, "signer":"Park Management Committee"}.

[0072] This invention achieves three core benefits through a dynamic weight evaluation model (LSTM real-time adjustment of indicator weights), multi-source data fusion (lidar + energy consumption monitoring + multispectral), and intelligent solution generation (genetic algorithm output optimization solution): Precision management: The accuracy rate of identifying inefficient land use has increased to 94%, and the decision-making cycle has been shortened by 98.9%; Economic efficiency: The average value per unit of land increased by 15.8%, and the investment payback period was shortened to 1.3-2.1 years; Ecological sustainability: Enforce ecological weight constraints (w3 ≥ 0.4), driving carbon emissions down by 22.5% and increasing vegetation carbon sequestration capacity by 37%.

[0073] Completely solve the problems of static industrial land assessment, lagging planning and low efficiency.

[0074] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0075] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. An industrial land intensive utilization assessment system, characterized by: It includes a data acquisition unit (1), a data processing unit (2), a dynamic evaluation unit (3), an optimization solution generation unit (4) and a visualization unit (5); The data acquisition unit (1) comprises: A three-dimensional space scanning device for obtaining three-dimensional contour information of buildings on industrial land; Energy consumption monitoring device, which collects real-time data on electricity and water consumption per unit area; Surface cover monitoring device, which obtains vegetation distribution information through multispectral imaging; The data processing unit (2) performs spatiotemporal matching of spatial data, economic data and ecological data, and calculates three basic indicators: space utilization rate, output value per unit area and ecological efficiency ratio; The dynamic evaluation unit (3) comprises: The weight prediction module dynamically generates weight coefficients based on historical land use data and policy text analysis; The intensity calculation module outputs the land use intensity evaluation value through weighted summation; The optimization solution generating unit (4) comprises: Case database, storing successful cases of industrial land transformation; Intelligent optimization engine generates land use adjustment plans with the goal of improving intensification; The visualization unit (5) is used to display the land use assessment results and planning scheme in three dimensions.

2. The industrial land intensive utilization evaluation system according to claim 1, characterized in that: The weight prediction module adopts a long short-term memory neural network, and its input data includes the basic indicator change sequence and the frequency of occurrence of policy keywords in the past 60 months.

3. The industrial land intensive utilization evaluation system according to claim 1, characterized in that: The eco-efficiency ratio is determined by the ratio of the vegetation cover index to the carbon dioxide emissions, wherein the vegetation cover index is obtained by multispectral image analysis.

4. The industrial land intensive utilization evaluation system according to claim 1, characterized in that: The intelligent optimization engine executes a genetic algorithm operation, and its solution coding structure includes coding segments of three dimensions: plot function type, building density, and industrial relevance.

5. The industrial land intensive utilization evaluation system according to claim 1, characterized in that: During the execution of the genetic algorithm, the fitness function is defined as the weighted sum of the intensiveness evaluation value and the industrial chain coordination degree.

6. The industrial land intensive utilization evaluation system according to claim 1, characterized in that: It also includes a data storage unit for generating an unalterable electronic certificate from the collected original data and its time information.

7. An evaluation method for an industrial land intensive utilization evaluation system according to any one of claims 1 to 6, characterized in that: The following steps are involved: Step 1: Collect building spatial data, enterprise economic data, and ecological environment data in the target area through the sensor network; Step 2: Calculate the spatial utilization efficiency index, unit area output index and ecological balance index; Step 3: Use machine learning models to analyze historical trends and policy directions and dynamically determine the weight coefficients of each indicator; Step 4: Compare the weighted intensive evaluation value with the preset threshold to identify inefficient land use; Step 5: Generate at least three land use adjustment plans based on the case library knowledge and optimization algorithm; Step 6: Visualize the comparison between the current situation and the planned solution through a 3D simulation platform.

8. The method for evaluating an industrial land intensive utilization evaluation system according to claim 7, characterized in that: The training process of the machine learning model in step 3 includes: Use attention mechanism to identify key features; Construct a loss function by comparing the difference between the predicted weights and the actual weights; Introduce weight distribution balance constraints.

9. The method for evaluating an industrial land intensive utilization evaluation system according to claim 7, characterized in that: The step five includes: (1) Randomly generate multiple sets of land use configuration scheme codes; (2) Calculate the comprehensive benefit value of each plan; (3) Iteratively optimize the solution through selection, crossover, and mutation operations; (4) Output the optimal solution set.

10. The evaluation method of the industrial land intensive utilization evaluation system according to claim 7, characterized in that: The step six comprises: The spatial distribution of intensiveness is presented in the form of a gradient color heat map; Use semi-transparent overlay technology to simultaneously display existing buildings and planned buildings.

Images