Ecological protection and restoration planning method and system

Abstract

The invention discloses an ecological protection and restoration planning method and system, and the method comprises the steps: obtaining ecological multi-source data of a target region, and carrying out the spatial registration and format standardization processing through a geographic information system; constructing a three-dimensional ecological background model, and fusing the ecological multi-source data to generate a gridding geographic unit; performing dynamic coupling analysis on the ecological multi-source data by adopting a convolutional neural network, identifying an ecological degradation hot spot region, and outputting a spatial probability distribution diagram; and based on the spatial probability distribution of the ecological degradation hot spot region and the ecological background model, generating a multi-level ecological restoration scheme through a self-organizing feature mapping neural network, and outputting a visual planning map. According to the scheme disclosed by the invention, the accuracy and efficiency of ecological assessment are improved, the economic cost of ecological protection and restoration in the region is reduced, the benefit is improved, a better expected effect can be obtained through tests, and the method has a good popularization prospect.

Description

Technical Field

[0001] This application relates to the technical field of intelligent environmental data processing, and particularly to an ecological protection and restoration planning method and system. Background Art

[0002] With the intensification of global ecological environment degradation, the existing ecological protection and restoration planning has the following limitations. First, traditional methods rely on single remote sensing or field survey data (such as NDVI or soil sampling), and fail to integrate multi-source heterogeneous data such as high-resolution remote sensing, DEM, meteorological dynamics, and human activities, resulting in one-sided ecological assessment results. For example, in soil erosion analysis, the urban expansion reflected by night light data and the resulting land use pressure are often ignored, leading to insufficient understanding of the driving mechanism of soil erosion, and further affecting the pertinence and effectiveness of restoration plans. The urban expansion pressure revealed by night light data is often ignored, causing the restoration plan to deviate from the actual driving mechanism. Second, although GIS platforms (such as ArcGIS) can achieve spatial overlay analysis, they rely on manually setting weight thresholds and are difficult to dynamically couple the non-linear relationships between ecological factors. For example, the minimum cost path model is mostly used in biological corridor planning, and technologies such as neural networks are not integrated to optimize the topological structure, resulting in low connectivity efficiency of the planned corridors and the actual economic cost of the final plan far exceeding expectations. Third, existing plans mostly statically output restoration zoning maps, lacking quantitative prediction of the long-term evolution of key processes of the ecosystem (such as carbon sequestration potential, water conservation, and biodiversity maintenance) in a dynamic environment after the implementation of measures. For example, if a forest vegetation restoration project is only laid out based on the current soil conditions and ignores the evaluation of key dynamic indicators such as its long-term carbon sequestration potential by combining future climate scenario simulations, it is difficult to ensure the sustainability of the selected vegetation and configuration plan under future climate conditions. This may lead to instability and insufficient adaptability of the core ecological functions of the restoration project.

[0003] Current research attempts to introduce machine learning, such as CNN for land cover classification, but has not yet formed a closed-loop full-chain technology system of "multi-source data fusion → intelligent diagnosis → adaptive plan generation → dynamic effect prediction". Therefore, there is an urgent need to develop an ecological protection planning method integrating geographic information fusion intelligent environmental data processing technology to achieve scientific decision-making and dynamic optimization of ecological protection and ecological restoration. Summary of the Invention

[0004] This application provides an ecological protection and restoration planning method and system to solve the above problems.

[0005] On the one hand, this application provides an ecological protection and restoration planning method, and the method includes the following steps: Step S1: Obtain the ecological multi-source data of the target area, and perform spatial registration and format standardization processing through a geographic information system; Step S2: Construct a three-dimensional ecological background model, and generate grid-based geographical units by integrating the multi-source ecological data; Step S3: Perform dynamic coupling analysis on the multi-source ecological data using a convolutional neural network to identify hotspots of ecological degradation, and output a spatial probability distribution map; Step S4: Based on the spatial probability distribution of the ecological degradation hotspots and the ecological background model, generate a multi-level ecological restoration plan through a self-organizing feature mapping neural network, and output a visual planning map.

[0006] In an implementation manner of the present application, the multi-source ecological data includes, but is not limited to: high-resolution remote sensing images, digital elevation models, soil type distribution maps, vegetation cover index time series, meteorological monitoring data, and spatial data of human activity intensity.

[0007] In an implementation manner of the present application, in the step S3, the dynamic coupling analysis uses an entropy weight-TOPSIS comprehensive evaluation model to calculate the ecological vulnerability index , and the process is as follows:

[0008] Among them, i is the geographical unit number; w j is the entropy weight of the j-th evaluation index, calculated through the index variability; represents the distance between unit j and the positive ideal solution; represents the distance between unit j and the negative ideal solution; the evaluation index set includes: vegetation cover attenuation rate, soil organic matter loss, topographic resistance factor γ, and rainstorm erosivity R.

[0009] In an implementation manner of the present application, the calculation formula of the topographic resistance factor γ is as follows: γ = ΔH / ΔL Among them, ΔH is the elevation range within the grid, ΔL is the length of the grid diagonal, and the surface roughness is extracted through the texture features of the remote sensing image.

[0010] In an implementation manner of the present application, the input layer of the self-organizing feature mapping neural network includes the ecological vulnerability index, the resistance cost of species migration paths, the economic cost of restoration measures, and the climate suitability score, and the topological structure of the output layer neurons corresponds to the spatio-temporal configuration matrix of the restoration plan.

[0011] In an implementation manner of the present application, in the step S4, the multi-level ecological restoration plan includes the vegetation restoration priority zoning, the layout of engineering measures, and the topological structure of ecological corridors. The topological structure of the ecological corridors is optimized and generated through an improved ant colony algorithm, and a simulated annealing mechanism is added during the algorithm iteration to avoid local optimal solutions. Among them, the path fitness function is:

[0012] Among them, C k is the ecological resistance value of the k-th section of the corridor, which is determined by land use type and slope. The ecological resistance value C k introduces the vegetation cover buffering effect in its calculation: when the corridor passes through the forest area, C k is multiplied by a preset attenuation coefficient; when passing through the urban built-up area, C k is multiplied by a preset amplification coefficient; Connectivity is the number of corridor-connected biological habitats; Construction_cost is the construction cost of the ecological corridor; α , β , δ is the weight coefficient, and α + β + δ = 1.

[0013] In an implementation manner of the present application, the spatial data of human activity intensity is obtained through the spatio-temporal overlay analysis of nighttime light remote sensing data, traffic road network density, and land use change trajectories, and its spatial resolution matches that of the high-resolution remote sensing image.

[0014] In an implementation manner of the present application, in the step S2, the size of the grid-shaped geographical unit is dynamically classified according to the geomorphic features, different grid scales are adopted for areas with different terrain complexities, and the grid attribute data is spatially filled by the Kriging interpolation method.

[0015] In an implementation manner of the present application, the method further includes: generating a dynamic prediction model for the implementation effect of the restoration plan, specifically: Inputting the planning scheme into a long short-term memory network, and combining with climate scenario simulation data, outputting the change curve of ecosystem stability and the spatial distribution map of carbon sink increment in the next 10 years.

[0016] On the other hand, the present application also provides an ecological protection and restoration planning system, and the system includes: A data preprocessing module, which is used to obtain the ecological multi-source data of the target area and perform spatial registration and format standardization processing through a geographic information system; A multi-source data fusion module, which is used to construct a three-dimensional ecological background model and generate grid-shaped geographical units by fusing the ecological multi-source data; A coupling analysis module, which is used to perform dynamic coupling analysis on the ecological multi-source data by using a convolutional neural network, identify hot spots of ecological degradation, and output a spatial probability distribution map; The scheme output module is used to generate a multi-level ecological restoration scheme through a self-organizing feature mapping neural network based on the spatial probability distribution of the ecological degradation hotspot area and the ecological background model, and output a visual planning map.

[0017] The present application provides an ecological protection and restoration planning method and system, which has the following beneficial effects: (1) The accuracy and efficiency of ecological assessment have been improved. Through the fusion of multi-source data and the construction of gridded geographic units, a three-dimensional coupled analysis of factors such as vegetation, terrain, and human activities has been achieved, which solves the one-sided shortcomings of traditional single-source data assessment. Combined with CNN dynamic coupling to identify degraded areas, the data processing efficiency has been improved, and more accurate predictions have been achieved. (2) Generate multi-level schemes through SOFM neural network, optimize ecological corridor topology based on improved ant colony algorithm, and dynamically balance ecological connectivity and construction cost through path fitness function, so as to reduce the economic cost of ecological corridor planning and construction; (3) The LSTM prediction model is introduced to simulate the long-term effects of ecological restoration plans under climate scenarios, output stability curves and carbon sink increments, and support dynamic adjustment strategies. Compared with traditional static planning, it improves the ecosystem recovery rate in the next 10 years and enhances the ability to respond to extreme climate events. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings: Figure 1 A flow chart of an ecological protection and restoration planning method provided in an embodiment of the present application; Figure 2 A composition diagram of an ecological protection and restoration planning system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0020] The embodiment of the present application provides an ecological protection and restoration planning method and system. The technical solution proposed in the embodiment of the present application is described in detail below with the help of the accompanying drawings.

[0021] Figure 1A flowchart of an ecological protection and restoration planning method provided by an embodiment of the present application. As Figure 1 shown, the method mainly includes the following steps: Step S1: Obtain multi-source ecological data of the target area, and perform spatial registration and format standardization processing through a geographic information system; Step S2: Construct a three-dimensional ecological background model, and generate grid-based geographical units by integrating the multi-source ecological data; Step S3: Perform dynamic coupling analysis on the multi-source ecological data by using a convolutional neural network, identify hotspots of ecological degradation, and output a spatial probability distribution map; Step S4: Based on the spatial probability distribution of the ecological degradation hotspots and the ecological background model, generate a multi-level ecological restoration plan through a self-organizing feature mapping neural network, and output a visual planning atlas.

[0022] In an embodiment of the present application, the multi-source ecological data includes but is not limited to: high-resolution remote sensing images, digital elevation models, soil type distribution maps, vegetation cover index time series, meteorological monitoring data, and spatial data of human activity intensity.

[0023] In an embodiment of the present application, in step S3, the dynamic coupling analysis uses an entropy weight-TOPSIS comprehensive evaluation model to calculate the ecological vulnerability index , and the process is as follows:

[0024] Among them, i is the geographical unit number; w j is the entropy weight of the jth evaluation index, calculated through the index variability; represents the distance between unit j and the positive ideal solution; represents the distance between unit j and the negative ideal solution; the evaluation index set includes: vegetation cover attenuation rate, soil organic matter loss, terrain resistance factor γ, and rainstorm erosivity R.

[0025] In an embodiment of the present application, the calculation formula of the terrain resistance factor γ is as follows: γ = ΔH / ΔL Among them, ΔH is the elevation range within the grid, ΔL is the length of the grid diagonal, and the surface roughness is extracted through the texture features of the remote sensing image.

[0026] In an embodiment of the present application, the input layer of the self-organizing feature mapping neural network includes the ecological vulnerability index, the resistance cost of species migration paths, the economic cost of restoration measures, and the climate suitability score, and the topological structure of the output layer neurons corresponds to the spatio-temporal configuration matrix of the restoration plan.

[0027] In the embodiment of the present application, in step S4, the multi-level ecological restoration plan includes a vegetation restoration priority partition, an engineering measure layout, and an ecological corridor topological structure. The ecological corridor topological structure is optimized and generated by improving the ant colony algorithm, and a simulated annealing mechanism is added during the algorithm iteration to avoid local optimal solutions. Among them, the path fitness function is:

[0028] where C k is the ecological resistance value of the k-th corridor section, which is determined by the land use type and slope. The calculation of the ecological resistance value C k introduces the vegetation cover buffering effect: when the corridor crosses the forest area, C k is multiplied by a preset attenuation coefficient; when crossing the urban built-up area, C k is multiplied by a preset amplification coefficient; Connectivity is the number of corridor-connected biological habitats; Construction_cost is the construction cost of the ecological corridor; α , β , δ are weight coefficients, and α + β + δ = 1.

[0029] In the embodiment of the present application, the spatial data of human activity intensity is obtained through the spatio-temporal overlay analysis of nighttime light remote sensing data, traffic road network density, and land use change trajectories, and its spatial resolution matches that of the high-resolution remote sensing image.

[0030] In the embodiment of the present application, in step S2, the size of the grid-shaped geographical unit is dynamically classified and divided according to the geomorphic features, different grid scales are adopted for areas with different terrain complexities, and the grid attribute data is spatially filled by the Kriging interpolation method.

[0031] In the embodiment of the present application, the method further includes: generating a dynamic prediction model for the implementation effect of the restoration plan, specifically: Input the planning scheme into the long short-term memory network, and combine with the climate scenario simulation data to output the ecological system stability change curve and the spatial distribution map of carbon sink increment in the next 10 years.

[0032] The above is an ecological protection and restoration planning method provided by the embodiment of the present application. Based on the same inventive concept, the embodiment of the present application also provides an ecological protection and restoration planning system. Figure 2 As shown in Figure 2 is a composition diagram of an ecological protection and restoration planning system provided by the embodiment of the present application. The system mainly includes: a data preprocessing module 201, which is used to obtain the ecological multi-source data of the target area and perform spatial registration and format standardization processing through a geographic information system; The multi-source data fusion module 202 is used to construct a three-dimensional ecological background model, and fuse the ecological multi-source data to generate grid-like geographical units; The coupling analysis module 203 is used to perform dynamic coupling analysis on the ecological multi-source data by using a convolutional neural network, identify hotspots of ecological degradation, and output a spatial probability distribution map; The solution output module 204 is used to generate a multi-level ecological restoration solution based on the spatial probability distribution of the ecological degradation hotspots and the ecological background model through a self-organizing feature mapping neural network, and output a visualized planning map.

[0033] The following specifically shows an example of the method provided by this application in a specific application scenario.

[0034] Dongming County is the first county where the Yellow River enters Shandong. The beach area reaches 317 square kilometers, making it the county with the largest beach area in the province. It belongs to the national key prevention area for soil and water loss in the sandy area of the Yellow River flood plain. The following problems have existed in previous ecological restoration and protection: (1) Severe soil and water loss: The sandy soil in the beach area accounts for 68%, and the annual average sediment loss reaches 12,000 tons per square kilometer; (2) Disruption of biological connectivity: Human activities (cultivated land / urban expansion) have led to habitat fragmentation, and the wetland bird habitat has shrunk by 40%; (3) Static restoration solutions: Traditional GIS planning has not coupled climate and economic factors, and the prediction error of carbon sink increment after the implementation of projects such as polygonum cuspidatum planting is >35%. Through the specific solution provided by this application, good ecological protection and restoration effects have been achieved. The specific planning process is as follows: 1. Multi-source data fusion and ecological background modeling Data acquisition: Collect remote sensing images with a resolution of 0.5 meters for the Dongming section of the Yellow River corridor, 30-meter DEM, soil salinity distribution maps, NDVI time series from 2015 to 2024, near 10-year heavy rain intensity data (maximum daily rainfall of 120 mm), and human activity data such as the night light index (average annual increase of 8.3%) and traffic road network density (expected to reach 4.2 km / km² in 2025).

[0035] Spatial registration: Uniformly project in the GIS platform into the CGCS2000 coordinate system, and the grid cell size is classified according to the terrain. Among them, the sandy land in the beach area: 50m×50m (high erosion risk area); the dike area: 100m×100m (low undulating area). Some grid attributes are shown in Table 1 below: Table 1 Partial grid attribute table

[0036] 2. CNN degradation identification and entropy weight-TOPSIS evaluation Dynamic coupling analysis: Input grid attributes into the ResNet-50 model to identify hotspots of soil and water loss (concentrated in the middle of the beach area, with an area proportion of 32%) and biodiversity loss areas (at the wetland edge, connectivity index < 0.4).

[0037] Calculation of the ecological vulnerability index (EVI): EVI i = 0.32 × (Vegetation attenuation) + 0.28 × (Soil loss) + 0.25 × γ (Topographic resistance) + 0.15 × R (Rainstorm erosion), where the topographic resistance factor γ is corrected by the root coverage rate of Reynoutria japonica in the beach area (when the measured root density > 15,000 plants / mu, γ is 0.6), and the output shows that the high vulnerability areas (EVI > 0.7) account for 41% of the whole region.

[0038] 3. Generation of the self-organizing feature map neural network SOFM scheme Input layer parameters: Ecological resistance: Resistance value of sandy land in the beach area = 8.0, forest land = 2.5; Economic cost: Planting Reynoutria japonica costs 1,200 yuan / mu, and arbor forest belts cost 3,500 yuan / mu; Climate suitability: Scored based on the RCP4.5 scenario (0 - 1).

[0039] Output layer optimization scheme: (1) Vegetation restoration zoning: Priority area I (EVI > 0.7): Plant Reynoutria japonica (sand fixation) + salt-tolerant shrubs (Tamarix chinensis); Priority area II (0.5 < EVI ≤ 0.7): Mixed forest (Fraxinus chinensis + Sabina chinensis).

[0040] Ecological corridor topology: Based on the improved ant colony algorithm, connect 6 wetland patches: Weights of the path fitness function: α = 0.5 (ecological connectivity), β = 0.3 (number of habitats), δ = 0.2 (construction cost); Crossing the urban section: C K = 2.5 (preset amplification factor), detour to add a wetland buffer zone, crossing the forest section: C K = 0.6 (preset attenuation factor), shortening the path by 12 km.

[0041] 4. Long short-term memory network LSTM dynamic prediction Input the restoration plan into the LSTM model to simulate the effects in the next ten years from 2026 to 2035. Climate data: IPCC RCP4.5 scenario (annual average temperature +1.2 °C, rainfall coefficient of variation +15%). The output shows that the carbon sink increment in 2035 is 86,000 tons. The ecosystem stability index rises from 0.62 to 0.81.

[0042] In this example, the multi-modal data fusion module uses the data interface of the M300 RTK drone + Sentinel-2 satellite. The GPU computing engine uses NVIDIA A100×4 (which can process 500,000 grid cells in parallel), and the visualization terminal uses a three-dimensional electronic sand table (scale 1:1000).

[0043] Summary: This example proves that the technical solution provided by this application has achieved good application value in complex ecological scenarios. First, multi-source data grid + CNN coupled analysis solves the problem of one-sided ecological assessment (e.g., the accuracy of the polygonum cuspidatum planting area location is increased to 92%). Second, the improved ant colony algorithm balances economic costs and ecological benefits (the corridor construction cost is reduced by 28.6%). Third, LSTM integrates climate scenarios to support long-term restoration decisions (the carbon sink prediction error < 10%). The solution system provided by this application can be extended to each county (city, district) along the Yellow River, providing a standardized technical paradigm for the construction of the ecological corridor along the Yellow River.

[0044] Each embodiment in this application is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple. For the relevant parts, reference can be made to the partial description of the method embodiment.

[0045] It should also be noted that the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. Without more limitations, the element defined by the statement "including one..." does not exclude the existence of other identical elements in the process, method, commodity or device including the said element.

[0046] The above are only the embodiments of this application and are not used to limit this application. For those skilled in the art, various changes and modifications can be made to this application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this application shall be included within the scope of the claims of this application.

Claims

1. An ecological protection and restoration planning method, characterized in that, The method includes the following steps: Step S1: Obtain the ecological multi-source data of the target area, and perform spatial registration and format standardization processing through a geographic information system; Step S2: Construct a three-dimensional ecological background model, and generate grid-shaped geographical units by fusing the ecological multi-source data; Step S3: Use a convolutional neural network to perform dynamic coupling analysis on the ecological multi-source data, identify ecological degradation hotspots, and output a spatial probability distribution map; Step S4: Based on the spatial probability distribution of the ecological degradation hotspots and the ecological background model, generate a multi-level ecological restoration plan through a self-organizing feature mapping neural network, and output a visual planning map.

2. The ecological protection and restoration planning method according to claim 1, wherein The ecological multi-source data includes but is not limited to: high-resolution remote sensing images, digital elevation models, soil type distribution maps, vegetation cover index time series, meteorological monitoring data, and human activity intensity spatial data.

3. The ecological protection and restoration planning method according to claim 1, characterized in that, In the step S3, the dynamic coupling analysis uses the entropy weight-TOPSIS comprehensive evaluation model to calculate the ecological vulnerability index , and the process is as follows: Among them, i is the geographical unit number; w j is the entropy weight of the j-th evaluation index, calculated through the index variability; represents the distance between unit j and the positive ideal solution; represents the distance between unit j and the negative ideal solution; The evaluation index set includes: vegetation cover attenuation rate, soil organic matter loss, topographic resistance factor γ, and rainstorm erosivity R.

4. An ecological protection and restoration planning method according to claim 3, characterized in that, The calculation formula of the terrain resistance factor γ is as follows: γ = ΔH / ΔL; where ΔH is the elevation range within the grid, ΔL is the length of the grid diagonal, and the surface roughness is extracted through the texture features of remote sensing images.

5. The ecological protection and restoration planning method according to claim 1, wherein The input layer of the self-organizing feature mapping neural network includes an ecological vulnerability index, a species migration path resistance cost, a restoration measure economic cost, and a climate suitability score, and the output layer neuron topological structure corresponds to the spatio-temporal configuration matrix of the restoration plan.

6. The ecological protection and restoration planning method according to claim 1, wherein In the step S4, the multi-level ecological restoration plan includes a vegetation restoration priority partition, an engineering measure layout, and an ecological corridor topological structure. The ecological corridor topological structure is optimized and generated through an improved ant colony algorithm, and a simulated annealing mechanism is added during the algorithm iteration to avoid local optimal solutions. Among them, the path fitness function is: Among them, C k is the ecological resistance value of the k-th corridor section, which is determined by land use type and slope. The ecological resistance value C k introduces the vegetation cover buffering effect in its calculation: when the corridor crosses the forest area, C k is multiplied by a preset attenuation coefficient; when crossing the urban built-up area, C k is multiplied by a preset amplification coefficient; Connectivity is the number of biological habitats connected by the corridor; Construction_cost is the construction cost of the ecological corridor; α , β , δ are weight coefficients, and α + β + δ = 1.

7. The ecological protection and restoration planning method according to claim 2, characterized in that The human activity intensity spatial data is obtained through the spatio-temporal overlay analysis of nighttime light remote sensing data, traffic road network density, and land use change trajectories, and its spatial resolution matches that of the high-resolution remote sensing images.

8. The ecological protection and restoration planning method according to claim 1, characterized in that In the step S2, the size of the grid-shaped geographical unit is dynamically graded and divided according to the geomorphic features. Different terrain complexity regions adopt different grid scales, and the grid attribute data is spatially filled through Kriging interpolation.

9. The ecological protection and restoration planning method according to claim 1, wherein, The method further includes: generating a dynamic prediction model for the implementation effect of the restoration plan, specifically: Input the planning plan into a long short-term memory network, and combine it with climate scenario simulation data to output a curve of the change in ecosystem stability and a spatial distribution map of carbon sink increment in the next 10 years.

10. An ecological protection and restoration planning system, characterized in that, The system includes: A data preprocessing module for obtaining the ecological multi-source data of the target area and performing spatial registration and format standardization processing through a geographic information system; A multi-source data fusion module for constructing a three-dimensional ecological background model and generating grid-shaped geographical units by fusing the ecological multi-source data; A coupling analysis module for using a convolutional neural network to perform dynamic coupling analysis on the ecological multi-source data, identifying ecological degradation hotspots, and outputting a spatial probability distribution map; A plan output module for generating a multi-level ecological restoration plan through a self-organizing feature mapping neural network based on the spatial probability distribution of the ecological degradation hotspots and the ecological background model, and outputting a visual planning map.

Images