Cellular automaton model-based urban ecological risk prediction system

Abstract

The invention provides an urban ecological risk prediction system based on a cellular automaton. The system integrates and obtains remote sensing images, meteorological data and human activity data through a data acquisition module, and extracts green land coverage, land utilization types and species distribution information of cities. Thirdly, a space grid is divided through a space grid division module by means of a dynamic quadtree algorithm, ecological units are generated, and therefore microcosmic space heterogeneity is captured; then, the system utilizes a cellular automaton model to optimize a state transition rule through deep reinforcement learning, and couples a soil erosion model and a hydrological model to simulate the dynamic change of urban ecology. And finally, quantizing parameter uncertainty through a risk assessment module by using a Bayesian belief network and Monte Carlo simulation. According to the system, urban ecological risks can be effectively simulated and predicted, the spatial precision and scientificity of prediction are improved, and the coupling influence of multiple ecological factors is considered at the same time.

Description

Technical Field

[0001] The present invention belongs to the cross-technical field of computer simulation and urban ecology, and more specifically relates to an urban ecological risk prediction system based on a cellular automaton model. Background Art

[0002] In the field of urban ecological risk prediction, traditional technical means mostly use static assessment methods, such as geographic information system (GIS) overlay analysis, or simplified dynamic models. These methods provide a basis for urban ecological risk assessment to a certain extent, but they have significant limitations.

[0003] First, static assessment methods such as GIS overlay analysis, although able to integrate multiple data sources, are incapable of simulating dynamic changes in urban ecology. They are usually based on fixed data snapshots for analysis, making it difficult to capture dynamic processes such as urban green space expansion and species migration that change over time, and thus are unable to fully reflect ecological changes in the temporal dimension.

[0004] Secondly, although some existing dynamic models attempt to introduce time factors, they are often over-simplified, resulting in large deviations between simulation results and actual ecological processes. These models are usually based on large-scale administrative divisions or grid divisions, ignoring the impact of micro-spatial heterogeneity on ecological risks. For example, the spatial distribution and changes of micro-ecological units such as green patches and species habitats within cities have an important impact on ecological risks, but traditional models often cannot accurately capture these changes.

[0005] In addition, most existing studies fail to comprehensively consider the interactive effects of multiple ecological factors when assessing urban ecological risks. The interactions between key ecological factors such as green space coverage, habitat connectivity, and species dispersal capacity are complex and dynamic, but existing models can often only consider these factors separately, ignoring the coupling relationship between them. The lack of this multi-factor coupling has led to doubts about the scientificity and accuracy of ecological risk assessment results.

[0006] Therefore, in view of the shortcomings of traditional urban ecological risk prediction methods in terms of dynamics, spatial accuracy and multi-factor coupling, the present invention proposes an urban ecological risk prediction system based on a cellular automaton model, in order to improve the scientificity and accuracy of urban ecological risk prediction. Summary of the invention

[0007] The purpose of the present invention is to provide an urban ecological risk prediction system based on cellular automata, which can effectively simulate and predict the dynamic changes of urban ecology, improve the spatial accuracy of ecological risk prediction and consider the coupling influence of multiple factors, thereby improving the scientificity and accuracy of urban ecological risk prediction.

[0008] In order to achieve the above object, the present invention is implemented by adopting the following technical scheme: the system comprises: The data acquisition module acquires multi-temporal remote sensing images through satellite ground stations and cloud platform interfaces, integrates meteorological data and human activity data, and uses a spatial-spectral joint classification model to extract green space coverage, land use type, and species distribution information; The spatial grid division module divides the spatial grid based on the dynamic quadtree algorithm and generates ecological units by combining the Voronoi diagram and the minimum cost path analysis; Cellular automaton model, which defines the cellular state space including ecological index, pressure index and resilience index, optimizes the state transition rules through deep reinforcement learning, and couples the soil erosion model and hydrological model; The risk assessment module uses Bayesian belief networks and Monte Carlo simulation to quantify parameter uncertainty, verifies the prediction results through a spatiotemporal cross-validation framework, and outputs a risk spatiotemporal distribution map and a warning layer.

[0009] In one embodiment, the remote sensing data processing of the data acquisition module includes: using the USGSESPACube architecture to perform atmospheric correction and radiometric calibration on Landsat data, and using the SNAP toolbox to complete interferometric measurement processing and vegetation index inversion on Sentinel-1 / 2 data.

[0010] In one embodiment, the spatial-spectral joint classification model consists of a random forest algorithm and a U-Net convolutional network, wherein the random forest processes multispectral feature bands, and the U-Net ResNet50 backbone network based on transfer learning performs pixel-level segmentation of high-resolution images.

[0011] In one embodiment, the acquisition of species distribution information includes: coupling the habitat suitability index inverted by remote sensing with the species occurrence point data from field surveys through the MaxEnt model, and deploying a soundprint monitoring array to collect bioacoustic characteristics.

[0012] In one solution, the dynamic quadtree algorithm sets the grid resolution according to the data density threshold, wherein the grid resolution in the high-density area is 50m and the grid resolution in the suburbs is 500m, and combines the DEM-assisted improved SIFT feature matching algorithm to achieve image registration.

[0013] In one embodiment, in the optimization of the state transition rules of the cellular automaton model, the state space of deep reinforcement learning includes ecological indicators, neighborhood influencing factors and external environmental variables, and the reward function dynamically adjusts the strategy based on the ecological value gain.

[0014] In one scheme, the Monte Carlo simulation uses Latin hypercube sampling to generate parameter perturbation combinations, covering species diffusion coefficients and policy enforcement intensity parameters, and accelerates simulation calculations through a fluid dynamics-inspired parallelization strategy.

[0015] In one approach, the spatiotemporal cross-validation framework divides historical data into training sets and validation sets according to time series, and aligns simulation results with real ecological event sequences through a dynamic time warping algorithm.

[0016] Beneficial effects of the present invention: 1. Improve the dynamics and accuracy of urban ecological risk prediction: Through the cellular automaton-based model, this system can simulate and predict the dynamic changes of urban ecological processes in real time. The optimization of spatial resolution and refined grid division also greatly improve the accuracy of micro-ecological unit simulation, effectively enhancing the accuracy of prediction.

[0017] 2. Multi-dimensional comprehensive assessment and risk level classification: This system integrates multiple ecological indicators such as biodiversity and habitat connectivity to comprehensively assess urban ecological risks and scientifically classify risk levels.

[0018] 3. Provide effective decision-making support: The ecological risk prediction results of this system can be intuitively displayed on a visual risk map, and scenario simulation can be performed to provide scientific and efficient decision-making support for urban planning departments to formulate ecological protection strategies.

[0019] 4. Comprehensive consideration of multiple ecological factors: This system analyzes the coupled impact of multiple ecological factors to make the prediction results more scientific, greatly improving the comprehensiveness and accuracy of urban ecological risk assessment.

[0020] In general, this invention will greatly help in scientifically predicting and responding to urban ecological risks, and will be of great significance for enhancing urban ecological protection and sustainable development. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a system block diagram of the present invention. DETAILED DESCRIPTION

[0022] In order to facilitate understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings. Typical embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described in the present invention. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

[0023] Unless otherwise defined, all technical and scientific terms used in the present invention have the same meaning as those understood by a person skilled in the art of the present invention. The terms used in the present invention in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. In order to facilitate the understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings. Typical embodiments of the present invention are shown in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described in the present invention. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

[0024] like Figure 1 As shown, the present invention proposes an urban ecological risk prediction system based on a cellular automaton model, which integrates multiple modules such as data collection, spatial grid division, cellular state transition, and risk assessment to achieve accurate prediction and dynamic management of urban ecological risks. The following is a specific implementation of each module: The data acquisition module is mainly responsible for acquiring multi-temporal high-resolution remote sensing images, such as Landsat, Sentinel and other data, and extracting information such as green space coverage, land use type and species distribution through professional image processing technology. At the same time, it integrates meteorological data (such as temperature and precipitation) and human activity data (such as the expansion rate of construction land) to provide comprehensive data support for subsequent model construction.

[0025] The implementation process of the data acquisition module begins with the collaborative acquisition and standardized processing of multi-source heterogeneous data. First, multi-temporal remote sensing data are downloaded in parallel through the satellite ground station and cloud platform interface. For the Landsat series data, the USGS ESPACube architecture is used for atmospheric correction and radiation calibration. For the Sentinel-1 / 2 data, the SNAP toolbox is used to complete the interferometric measurement processing and vegetation index inversion. In the temporal image registration link, the improved SIFT feature matching algorithm is combined with DEM assistance to achieve sub-pixel registration accuracy and ensure the geographic spatial consistency of data in different seasonal phases. For the feature extraction of high-resolution images, a spatial-spectral joint classification model is developed, combining the random forest algorithm with the U-Net convolutional network: the former processes multi-spectral feature bands and uses NDVI, NDBI and other indices to establish land use classification decision trees; the latter uses the ResNet50 backbone network pre-trained by transfer learning to achieve pixel-level segmentation of green space coverage boundaries on 0.5-meter resolution WorldView images. Experiments show that the overall accuracy of this method in complex urban environments reaches 92.7%.

[0026] The acquisition of species distribution data adopts a multimodal fusion strategy. The habitat suitability index inverted by remote sensing is coupled with the species occurrence point data of the field survey through the MaxEnt model. The dominant environmental variables are screened using the Jackknife test to construct a probability distribution surface. For hidden species that are difficult to observe directly, a soundprint monitoring array is deployed to collect bioacoustic characteristics, and a bird song soundprint library is constructed using a deep learning model (such as ResNeSt). Acoustic species identification is achieved through a time-frequency graph convolutional network. The processing of meteorological data focuses on solving the problem of differences in spatiotemporal scales. The improved Kriging interpolation algorithm is used to downscale the meteorological station data to a 30-meter grid, and the elevation-temperature regression model is introduced to correct the terrain effect. At the same time, the long short-term memory network (LSTM) is used to fill in the missing data and construct a spatiotemporal continuous meteorological element field.

[0027] The collection of human activity data integrates multi-dimensional information such as night light data (NPP-VIIRS), POI density, thermal and traffic flow monitoring, and quantifies the driving mechanism of construction land expansion through a spatial lag model. In view of the heterogeneity contradiction between administrative statistical data and real-time sensor data, a Bayesian spatiotemporal hierarchical model is designed: the prior distribution uses Markov random fields to express spatial autocorrelation, the likelihood function introduces a radial basis function network to fit the nonlinear relationship, and the posterior distribution is solved by Hamiltonian Monte Carlo sampling, and finally a construction land expansion rate surface with probabilistic significance is generated. After preprocessing, all data layers are integrated through a spatiotemporal adaptive fusion engine. The engine constructs a spatiotemporal weight matrix based on the improved STARFM algorithm, and uses cross-correlation analysis within a sliding time window to determine the optimal fusion weight to ensure the consistency of multi-source data in the spatiotemporal dimension. Finally, a standardized spatial data set with metadata quality identification is output, providing reliable input for subsequent cellular automaton models.

[0028] The spatial grid division module divides the research area into regular cell grids, such as 30m×30m, where each cell represents a micro-ecological unit. This module supports dynamic adjustment of grid resolution to meet research needs at different scales. The mapping between spatial coordinates and cell units is achieved through GIS tools to ensure that each cell can be accurately located in geographic space.

[0029] The implementation of the spatial grid division module begins with the spatial benchmark unification and dynamic resolution adaptation of multi-source data. After establishing the UTM projection coordinate system in the GIS environment, the heterogeneous data obtained by the data acquisition module are batch reprojected through the Warp interface of the GDAL library to eliminate the offset error caused by the coordinate system difference and ensure the consistency of the spatial benchmark of remote sensing images, meteorological grids and vector boundary data. For the generation of 30m basic grids, the improved Fishnet algorithm is used to construct a regular grid: first, the number of rows and columns is determined based on the minimum enclosing rectangle of the study area, and the calculation efficiency is optimized through the quadtree index. Then, the spatial topology verification tool is used to automatically correct the edge deformation caused by the curvature of the earth to ensure that the geographic positioning accuracy of each cell is controlled at the sub-meter level. To meet the needs of dynamic resolution adjustment, a multi-level grid coupling engine was developed, and an adaptive subdivision mechanism was constructed based on the Delaunay triangulation. When a local area requires 10m high-precision analysis, the system automatically calls the Voronoi subdivision algorithm to recursively subdivide the target cells while maintaining the topological connectivity of adjacent grids. Experiments have shown that this scheme can improve the recognition accuracy of morphological features of habitat boundaries by 18.7% while maintaining the computational efficiency of the 30m basic grid.

[0030] The process of assigning cell attributes adopts a technical path that combines spatial overlay analysis with raster calculation. The preprocessed green space coverage raster (derived from the U-Net segmentation result), land use classification map (random forest output) and meteorological element field are spatially registered through ArcPy script, and the attribute values ​​of each data layer are assigned to the corresponding cells in batches using the Zonal Statistics tool. For species distribution data, the probability distribution surface output by the MaxEnt model is discretized into cell units using the kernel density estimation method, and the habitat suitability index of each cell is calculated in combination with the moving window algorithm. The spatial processing of human activity data innovatively adopts the field model transformation method, maps the construction land expansion rate surface to the grid system through bilinear interpolation, and integrates the night light intensity and POI density to construct a composite human interference index.

[0031] Dynamic resolution management is achieved through a multi-level pyramid structure, and a LOD (Levels of Detail) data model is constructed: 12 levels of grid data from 1km to 1m are stored in the PostGIS database, and R-tree index is used to optimize spatial query efficiency. When the user selects a specific analysis scale, the system automatically matches the optimal grid level through an adaptive sampling algorithm, reducing the computational load while maintaining the accuracy of ecological process simulation. In order to verify the quality of grid division, a spatial topology consistency detection module was developed, and the connectivity analysis algorithm in graph theory was used to identify abnormal cells. Combined with Monte Carlo simulation, the spatial expression error of the grid system for ecological parameters was evaluated. The measured data showed that under the 30m benchmark grid, the spatial correlation coefficient retention rate of the temperature field reached 96.2%, and the distribution characteristic error of the NDVI value was controlled within ±3.5%. The final output cell matrix not only contains complete ecological attribute fields, but also embeds multi-scale association indexes, providing structured spatial calculation units for subsequent cell state transitions.

[0032] The core of the cellular state transition module is the spatiotemporal dynamic evolution mechanism of multi-agent collaboration. Its implementation process deeply integrates spatial explicit rules and random process simulation. In the initialization stage, based on the cellular attribute matrix generated by the previous module, a three-dimensional topological database containing geographic coordinates, ecological attributes and adjacency relationships is established through the PostGIS extension of PostgreSQL.

[0033] This module updates the cell state based on preset rules to simulate the dynamic changes of urban ecological processes. The specific rules include: Green space expansion rule: Combined with factors such as the state of neighboring cells, land use policies, and population density, the dynamic changes of green space are simulated. For example, when the surrounding cells are green space and the policy encourages greening, the cell has a greater probability of becoming a green space.

[0034] The green space expansion rule adopts the improved cellular automaton-Markov coupling model and defines the state transition probability as ,in It represents the proportion of green space cells in Moore's neighborhood, which is obtained by spatial convolution kernel calculation; It is the policy impact factor, which is quantitatively generated based on the TF-IDF of the document keywords; The weight factor is derived from the normalized value of the population density raster. The model was obtained through historical data training and the Bayesian optimization algorithm was used to search for the optimal combination in the parameter space. Experiments showed that the Kappa coefficient of the model reached 0.87 in the verification in Shanghai.

[0035] Species migration rules: Calculate the migration path of species based on factors such as habitat suitability, diffusion threshold, and obstacle distribution. By evaluating the habitat suitability of each cell, determine the possible migration direction of the species, and optimize the path in combination with diffusion threshold and obstacle distribution.

[0036] The implementation of species migration rules relies on an improved circuit theory model to construct a migration resistance surface

[0037] in is the habitat suitability index, is the terrain slope, represents obstacle factors such as road density. The migration path optimization uses the anisotropic diffusion equation

[0038] The diffusion coefficient , the species diffusion probability cloud map is obtained by discrete solution using the finite element method. When the inter-cell migration probability exceeds the threshold , d is the intercellular distance, k is the species diffusion capacity parameter), the system calls the A* algorithm to dynamically plan the optimal path, and integrates the Voronoi diagram to divide the sphere of influence to ensure that the migration process conforms to the niche theory.

[0039] Habitat connectivity rule: Use graph theory algorithms (such as minimum spanning tree) to evaluate the connectivity between cells. By constructing a connection network between cells, the connectivity of each cell is calculated to reflect the integrity and connectivity of the habitat.

[0040] Habitat connectivity assessment innovatively combines topological persistent homology theory with multi-layer graph neural networks. Constructing a cellular connectivity graph , a set of vertices Corresponding to the cell center, edge weight Reflects the quality of ecological corridors. When using the improved Kruskal algorithm to generate the minimum spanning tree, a dynamic programming strategy is introduced: when the newly added edges form a loop, the weight variance of the edges in the loop is compared.

[0041] Only the connection pattern with the smallest variance is retained. The connectivity quantification uses algebraic connectivity from spectral graph theory , where the Laplacian matrix , which can effectively characterize the vulnerability of the habitat network. Simulation experiments show that when The risk of species extinction will increase by 62%.

[0042] The system uses a discrete event-driven architecture when running, and performs three-stage updates in each time step: first, the state transition probability matrix of all cells is calculated in parallel, and the CUDA-accelerated Monte Carlo simulation is called to select state transition events; second, the cross-process cell state changes are synchronized through the message passing interface (MPI); finally, the global connection graph attributes are updated and written to the spatiotemporal database. In order to improve computing efficiency, a quadtree-based spatial index optimizer was developed to dynamically allocate the computing load to 256×256 block units. Actual measurements show that this scheme increases the iteration speed of the 10,000-level cellular system by 14.3 times. The model verification uses the cross-wavelet analysis method to test the phase coherence of the simulation results and Landsat time series data at multiple scales to ensure that the spatial heterogeneity and temporal continuity of the dynamic process are consistent with the actual observation law.

[0043] The risk assessment module calculates biodiversity index (such as Shannon-Wiener index), habitat fragmentation index and risk level based on the results of cell state transition. It quantifies uncertainty through Monte Carlo simulation and outputs a risk spatiotemporal distribution map. The specific implementation steps include: The construction of the risk assessment module focuses on the spatial coupling analysis of multi-dimensional ecological indicators and uncertainty propagation modeling. In the initialization stage, the spatiotemporal series data output by the cellular state transition module are loaded in parallel through the Spark cluster to establish a hybrid data cube containing species abundance matrix, habitat connectivity tensor and human activity intensity field.

[0044] The biodiversity index is calculated using the improved Shannon-Wiener index.

[0045] in is the number of individuals of the i-th species in the cell, is the total number of individuals, It represents the protection level coefficient of the species in the climate zone to which the cell belongs. By introducing a spatial heterogeneity correction factor, the index deviation caused by differences in sampling area is effectively eliminated. In order to improve computing efficiency, a species abundance histogram counter based on GPU acceleration is developed, and CUDA atomic operations are used to achieve parallel frequency statistics. The actual calculation speed is increased by 22 times at the scale of millions of cells.

[0046] The quantification of habitat fragmentation index innovatively integrates landscape ecology and complex network theory to construct a dual-channel evaluation system: the structural dimension adopts the effective grid size ,in is the green patch area, is the total area of ​​the study area; the functional dimension is based on the habitat connectivity map generated in module 3, defining the flow centrality ,in is the total number of the least resistance paths from species s to t, is the number of paths passing through cell i. Final fragmentation index

[0047] , a 500m resolution heat map was generated through spatial sliding window calculation. Experiments showed that the Pearson correlation coefficient between this indicator and the abandonment rate of bird nest sites reached 0.78.

[0048] The risk level classification uses a spatiotemporal constraint clustering algorithm to construct a feature vector (HA is human activity intensity), and multi-scale risk partitioning is achieved through the improved OPTICS clustering algorithm. The distance metric is defined as

[0049] in is the long-range correlation correction term based on the Hurst index, The spatial gradient of human activity intensity between cells is integrated. Moran's I test is performed simultaneously during the clustering process to ensure that the spatial autocorrelation threshold of each risk area .

[0050] Uncertainty quantification is achieved through a hybrid framework of Bayesian belief networks and Monte Carlo simulation. Establishing parameter perturbation model , covering 23 key parameters such as species diffusion coefficient and policy enforcement strength, and using Latin hypercube sampling to generate 1,000 sets of parameter combinations. During each simulation run, parameter perturbations are dynamically injected through the message queue, and a fluid dynamics-inspired parallelization strategy is used to achieve minute-level iterations on a 48-core server. The output stage uses non-parametric kernel density estimation to construct a risk probability surface

[0051] The Epanechnikov kernel function It effectively balances smoothness and detail preservation. The confidence interval map is generated by the quantile regression forest algorithm, and the average swing amplitude of the risk boundary is controlled within ±125m at a 95% confidence level.

[0052] The verification phase uses a spatiotemporal cross-validation framework: the observation data from 2010 to 2020 are divided into training sets and validation sets, and the morphological similarity between the simulated risk surface and the real ecological event sequence is calculated through the dynamic time warping algorithm (DTW). The results show that the average similarity is 82.4% at a 1km grid scale. The final output of the risk spatiotemporal distribution map is integrated into the WebGL visualization platform, supporting multi-dimensional slider interactive exploration, and embedding the ecological red line warning layer in GeoJSON format to provide decision-making information products for management departments. The 23TB process data generated during system operation is tamper-proofed through blockchain technology to ensure the traceability of the entire risk assessment process.

[0053] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM).

[0054] It should be understood that the detailed description of the technical solutions of the present invention by means of the preferred embodiments is illustrative rather than restrictive. A person skilled in the art may modify the technical solutions described in the embodiments, or replace some of the technical features by equivalents, based on reading the specification of the present invention; and these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An urban ecological risk prediction system based on cellular automata, characterized in that: The system comprises: The data acquisition module acquires multi-temporal remote sensing images through satellite ground stations and cloud platform interfaces, integrates meteorological data and human activity data, and uses a spatial-spectral joint classification model to extract green space coverage, land use type, and species distribution information; The spatial grid division module divides the spatial grid based on the dynamic quadtree algorithm and generates ecological units by combining the Voronoi diagram and the minimum cost path analysis; Cellular automaton model, which defines the cellular state space including ecological index, pressure index and resilience index, optimizes the state transition rules through deep reinforcement learning, and couples the soil erosion model and hydrological model; The risk assessment module uses Bayesian belief networks and Monte Carlo simulation to quantify parameter uncertainty, verifies the prediction results through a spatiotemporal cross-validation framework, and outputs a risk spatiotemporal distribution map and a warning layer.

2. The urban ecological risk prediction system based on cellular automata according to claim 1, characterized in that: The remote sensing data processing of the data acquisition module includes: using the USGS ESPACube architecture to perform atmospheric correction and radiometric calibration on Landsat data, and using the SNAP toolbox to complete interferometric measurement processing and vegetation index inversion on Sentinel-1 / 2 data.

3. The urban ecological risk prediction system based on cellular automata according to claim 1, characterized in that: The spatial-spectral joint classification model consists of a random forest algorithm and a U-Net convolutional network, wherein the random forest processes multispectral feature bands, and the U-Net ResNet50 backbone network based on transfer learning performs pixel-level segmentation of high-resolution images.

4. The urban ecological risk prediction system based on cellular automata according to claim 1, characterized in that: The acquisition of the species distribution information includes: coupling the habitat suitability index inverted by remote sensing with the species occurrence point data from field surveys through the MaxEnt model, and deploying a soundprint monitoring array to collect bioacoustic characteristics.

5. The urban ecological risk prediction system based on cellular automata according to claim 1, characterized in that: The dynamic quadtree algorithm sets the grid resolution according to the data density threshold, where the grid resolution in the high-density area is 50m and the grid resolution in the suburbs is 500m, and combines the DEM-assisted improved SIFT feature matching algorithm to achieve image registration.

6. The urban ecological risk prediction system based on cellular automata according to claim 1, characterized in that: In the state transition rule optimization of the cellular automaton model, the state space of deep reinforcement learning includes ecological indicators, neighborhood influencing factors and external environmental variables, and the reward function dynamically adjusts the strategy based on the ecological value gain.

7. The urban ecological risk prediction system based on cellular automata according to claim 1, characterized in that: The Monte Carlo simulation uses Latin hypercube sampling to generate parameter perturbation combinations, covering species diffusion coefficients and policy enforcement intensity parameters, and accelerates the simulation calculation through a parallelization strategy inspired by fluid dynamics.

8. The urban ecological risk prediction system based on cellular automata according to claim 1, characterized in that: The spatiotemporal cross-validation framework divides historical data into training sets and validation sets according to time series, and aligns simulation results with real ecological event sequences through a dynamic time warping algorithm.

Images