Layout method for identifying high potential area of marine three-dimensional space based on suitability evaluation

Abstract

The application discloses a marine three-dimensional space high-potential area identification layout method based on suitability evaluation, relates to the technical field of marine space planning management, collects marine natural, social and ecological data and historical three-dimensional right confirmation project cases, carries out pretreatment to construct a marine three-dimensional space basic database; analyzes historical three-dimensional right confirmation cases based on a graph neural network, constructs a sea use activity-space layer relationship graph, and analyzes the compatibility of different sea use activity combinations in historical projects by using the sea use activity-space layer relationship graph. The application can identify high-potential areas that meet natural conditions, ecological constraints and social needs at the same time by constructing a multi-dimensional suitability evaluation system and a compatibility probability matrix, and avoids plane conflicts of traditional two-dimensional planning by using hierarchical weighted superposition analysis. The three-dimensional hierarchical utilization rules optimize the vertical space resource allocation, so that the same marine area can be compatible with multiple sea use activities, and inefficient or conflicting layout is reduced.

Description

Technical Field

[0001] This invention relates to the field of marine spatial planning and management technology, specifically to a method for identifying and distributing high-potential areas in marine three-dimensional space based on suitability assessment. Background Technology

[0002] With the continuous growth of global marine development and utilization, the scarcity of marine space resources is becoming increasingly prominent, and conflicts over marine use are intensifying. Nearshore waters are home to almost all types of marine activities, such as aquaculture, port shipping, and tourism and leisure, resulting in high intensity, complex types, numerous potential conflicts, and significant management challenges. For example, within the same vertical sea area, the surface may be used for shipping or offshore wind power construction, the middle layer of the water is suitable for three-dimensional aquaculture, and the bottom layer may be used for laying submarine pipelines or conducting mineral resource exploration. The mutual interference of multi-level marine activities has become the main bottleneck restricting the efficient use of marine space. However, by rationally dividing the vertical space of the sea area, a three-dimensional layered utilization model provides a mechanism for sharing resources and coordinating development for different types of marine activities, which helps to improve the efficiency of marine space utilization and reduce the difficulty and management costs of marine use demonstration and stakeholder coordination.

[0003] In existing technologies, overlapping areas are often identified as high-potential areas through overlay analysis, but this lacks consideration of the spatiotemporal characteristics of conflicts, which can easily lead to frequent contradictions during the implementation of planning schemes. Therefore, how to construct a marine activity-spatial layer relationship map based on graph neural networks, predict the compatibility probability of new marine use combinations, and recommend optimized layouts to ensure the accuracy of high-potential area identification and layout in marine three-dimensional space is the problem that this invention aims to solve. To this end, a method for identifying and laying out high-potential areas in marine three-dimensional space based on suitability evaluation is proposed. Summary of the Invention

[0004] The purpose of this invention is to provide a method for identifying and distributing high-potential areas in marine three-dimensional space based on suitability assessment, so as to solve the problems mentioned in the background art.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0006] A method for identifying and allocating high-potential marine three-dimensional spatial areas based on suitability assessment includes the following steps:

[0007] S1. Collect natural, social, and ecological data of the sea area and historical three-dimensional property rights confirmation project cases, clean, transform and standardize the preprocessing, unify the coordinate system and accuracy standards, and build a basic database of three-dimensional space of the sea area;

[0008] S2. Based on graph neural network analysis of historical three-dimensional rights confirmation cases, marine use activities and spatial layer are used as nodes, and the relationships in historical cases are used as edges to construct a marine use activity-spatial layer relationship graph and quantify the compatibility probability.

[0009] S3. Using the marine activity-spatial layer relationship map, analyze the compatibility of different marine activity combinations in historical projects and extract compatibility and conflict patterns;

[0010] S4. Combining historical patterns, use graph neural network models to predict the compatibility probability of new marine activity combinations and identify potential conflicts;

[0011] S5. Based on the analytic hierarchy process, determine the weights, conduct a stratified suitability evaluation of marine activities, generate scoring patches, and combine the compatibility probability prediction results to divide high, medium and low potential areas, and prioritize the recommendation of three-dimensional layout areas with low conflict probability.

[0012] S6. Based on high-potential areas, we recommend an optimized layout scheme to match the layout requirements of high-potential areas.

[0013] A further improvement to the technical solution of the present invention is that: S1 specifically includes:

[0014] Collect three categories of data for the target sea area: natural, social, and ecological. Natural data includes water depth, seabed type, and marine hazard risk. Social data covers transportation networks, port location conditions, and tourism facilities. Ecological data involves marine ecological protection red lines, seawater quality, and distribution of biological resources. At the same time, collect case data on historical three-dimensional rights confirmation projects, including types of sea use activities, spatial layer distribution, conflict records, and solutions.

[0015] The collected raw data is cleaned, including format standardization, outlier correction and missing value imputation. At the same time, data conversion is performed to unify data of different formats into a format that is easy to process, and standardization is carried out to unify the data measurement units and classification standards. This ensures that all spatial data are in the same coordinate system, the accuracy meets the project requirements, eliminates errors and conflicts caused by inconsistent data formats and standards, and improves data quality and consistency.

[0016] Based on the processed data, a basic database architecture for marine three-dimensional space was designed, and the core table structure and field definitions were determined. The core table structure of the database includes a marine basic information table, a natural data table, a social data table, an ecological data table, and a historical case table. The processed data was then imported into the database according to the designed architecture, and data entry verification was performed to ensure that the data was stored accurately. At the same time, a data indexing and query mechanism was established to form a basic database for marine three-dimensional space.

[0017] A further improvement to the technical solution of the present invention is that: S2 specifically includes:

[0018] Based on historical three-dimensional rights confirmation cases, the types of marine activities and spatial layers (water surface, water body, seabed, subsoil) are defined as two types of nodes in the map. The node attributes include the type of marine activity, spatial range, temporal characteristics, and physical and ecological attributes of the spatial layer. The interaction relationships in historical three-dimensional rights confirmation cases are used as edges. The edge attributes record the intensity of conflict. The attribute encoding format of nodes and edges is unified. The connectivity of the map is ensured through topological verification. Each marine activity must be associated with at least one spatial layer. Spatial layer nodes must be indirectly connected through activity nodes. Isolated nodes or redundant edges are eliminated to form a structured initial map.

[0019] A graph neural network (graph convolutional network) is used to train the initial graph to learn the latent feature representations of nodes and edges. The input is the initial features of nodes and graph structure information, and the output is the low-dimensional embedding vectors of nodes and edges. The node embeddings integrate the semantic features of activity type and spatial layer, and the edge embeddings encode the latent patterns of compatibility or conflict, thus obtaining the trained graph neural network model.

[0020] Based on a pre-trained graph neural network model, the compatibility of any activity-spatial layer combination in the graph is probabilistically quantified. By calculating the node embedding similarity and edge embedding weight, a compatibility probability matrix is ​​generated, where each element represents the compatibility score of a specific activity in a certain spatial layer. At the same time, a dynamic graph update mechanism is established to integrate the activity, spatial layer, and relationship data from new rights confirmation cases into the graph in real time, retrain and update the model parameters, and ensure that the compatibility quantification results reflect the latest requirements of marine use practices, outputting an interpretable marine use activity-spatial layer relationship graph.

[0021] A further improvement to the technical solution of the present invention is that: S3 specifically includes:

[0022] Based on the constructed relationship graph between marine activities and the spatial layer, we analyze its topological structure and node / edge features. We also identify high-frequency subgraph patterns using a graph traversal algorithm based on breadth-first search. Starting from high-frequency nodes, we traverse their neighborhoods within 3 hops to extract recurring subgraph structures, thereby determining the common connection methods between marine activities and the spatial layer. We use degree centrality to calculate the degree of nodes, quantify node importance, retain the top 10% of key nodes, and screen out key activities and spatial layer nodes.

[0023] The graph embedding technique (Node2Vec) is used to map nodes and edges to a low-dimensional vector space, preserving their structural and attribute information. The K-means clustering algorithm is used to group the embedded vectors to identify combinations of marine activities with similar compatibility patterns or conflict patterns. The compatibility pattern is characterized by close node embedding distance and high edge weight, while the conflict pattern corresponds to separate embeddings or extremely low edge weight. The edge weight distribution is analyzed by combining the graph attention mechanism to extract the conflict transmission rules across spatial layers and form a pattern rule base.

[0024] The universality of the patterns is verified by backtesting historical cases. The matching degree between the extracted patterns and real cases is calculated. Invalid patterns with overfitting or low coverage are eliminated. The significance of the patterns is evaluated by statistical tests to ensure their non-randomness. The verified patterns are classified according to conflict type to build a structured knowledge base.

[0025] A further improvement to the technical solution of the present invention is that: S4 specifically includes:

[0026] The historical marine activity-spatial layer relationship map is transformed into a heterogeneous graph structure, the node type and edge type are defined, and the conflict intensity attribute in the historical pattern is embedded as the edge weight. The heterogeneous interaction is modeled by graph neural network, the feature differences between the activity and spatial layer are captured by multi-type node encoder, and the neighborhood information is aggregated by attention mechanism to generate node embedding that integrates structure and attributes.

[0027] For newly submitted combinations of marine activities, they are mapped to a pre-trained heterogeneous graph as new nodes, and their embedding representations are updated through the message passing mechanism of graph neural networks. Based on the connection relationship between the new marine activity and the existing spatial layer, the interaction weights between the new marine activity and its neighboring nodes are dynamically calculated. Combined with the feature distribution learned from historical patterns, the compatibility probability of the combination is output. If the predicted probability is lower than the preset conflict threshold, it is marked as a potential conflict. If it is higher than the conflict threshold, it is judged as compatible.

[0028] For combinations of activities predicted as potential conflicts, the conflict types and transmission paths are further analyzed. By using the edge weight distribution of a graph neural network, the direct sources and indirect transmission chains of the conflicts are identified. The conflict patterns are extracted into structured rules and stored in a dynamic knowledge base.

[0029] A further improvement to the technical solution of this invention lies in the following: the process of extracting conflict patterns into structured rules is as follows:

[0030] Based on the edge weight distribution of graph neural networks, the direct source node and indirect influence path of the conflict are located, high-weight conflict edges directly connected to the new active node are extracted, the direct conflicting parties are identified, and the gradient contribution of the edge weights is calculated through backpropagation to trace the indirect transmission chain of hop count across nodes. At the same time, combined with the topology sorting algorithm, the nodes are sorted according to their dependencies to ensure that the path is acyclic and covers key nodes, and the complete transmission path graph of the conflict is output, with the edge weight and node type labeled.

[0031] Based on the node types, edge weights, and historical labels in the transmission path, conflict types are summarized, including spatial overlap conflicts, ecological chain conflicts, and temporal conflicts. The conflict intensity (weight mean) and the scope of influence (transmission path length) are quantified, and the identified transmission paths are transformed into structured rules. Then, the rules are encoded into a machine-readable format through logical programming.

[0032] Redundancy and conflict detection is performed between the new rules and the existing rules in the knowledge base. Similarity calculation based on graph embedding is used to merge duplicate rules, eliminate low-coverage rules, verify the matching degree between the new rule base and historical cases, ensure compatibility with historical data, retain the core rules covering the top 20%, add timestamps to the rules, and periodically eliminate outdated entries that have not been triggered for 5 years.

[0033] A further improvement to the technical solution of the present invention is that: S5 specifically includes:

[0034] Based on the spatial characteristics of the sea area and the needs of sea use activities, a hierarchical suitability evaluation index system is established, which includes three primary indicators: natural, social and ecological, with secondary indicators, and tertiary indicators. The analytic hierarchy process (AHP) is used to determine the weight of each indicator. By constructing a judgment matrix, a consistency test is performed, and the weight distribution is obtained by calculating the eigenvector.

[0035] Weighted overlay analysis of rasterized data based on weights is used to generate marine activity suitability score patches with a score range of 1-5 points, where 1 point is unsuitable and 5 points is optimal. The continuous scores are discretized into three levels using the natural breakpoint method: suitable (≥4 points), relatively suitable (3-4 points), and unsuitable (<3 points). Then, the marine functional zoning boundary is overlaid, prohibited development areas are removed, and a preliminary suitable area distribution map is formed.

[0036] The suitability score patch is multiplied by the compatibility probability predicted by the graph neural network to obtain the comprehensive potential value. Potential zones are divided into high-potential, medium-potential, and low-potential zones according to percentiles. High-potential zones are recommended for three-dimensional layout, as they simultaneously meet the requirements of high suitability and high compatibility. Conflict risk points in medium-potential zones that require manual review are automatically marked, while low-potential zones are directly excluded.

[0037] A further improvement to the technical solution of the present invention is that the expression for the marine activity suitability scoring patch is as follows:

[0038] ;

[0039] In the formula, For grid cells The appropriateness score (range 1-5 points). , , Weighting of natural, ecological, and social indicators. For the natural index k in The standardized score (1-5 points). , The standardized scores for the ecological and social indicators (1-5 points) are: n is the number of secondary and tertiary indicators under the natural indicators, m is the number of secondary and tertiary indicators under the ecological indicators, and q is the number of secondary and tertiary indicators under the social indicators.

[0040] The expression for the comprehensive potential value is as follows:

[0041] ;

[0042] In the formula, For grid The overall potential value (range 0-5). For compatibility probability, This is an indicator function; 1 is used for non-prohibited development zones, and 0 otherwise. High-potential zones have high adaptability. And highly compatible ( The medium potential zone is characterized by excellent performance in either adaptability or compatibility, while the low potential zone is characterized by failure in any one of the indicators.

[0043] A further improvement to the technical solution of the present invention is that: S6 specifically includes:

[0044] Based on the spatial characteristics of high-potential areas and the needs of marine activities, key parameters are extracted, including natural and social constraints such as water depth range, ecological sensitivity, and adjacent infrastructure. At the same time, in line with planning objectives, spatial overlay analysis is used to exclude areas that conflict with existing functional zoning, ensuring that the recommended scheme complies with legal boundaries. Combined with the compatibility probability matrix output by graph neural network, grid units that simultaneously meet high suitability and high compatibility are selected to form an initial candidate area set.

[0045] For candidate areas, based on the rules for three-dimensional layered use of marine areas, the vertical spatial ownership scope of water surface, water body, seabed, and subsoil is divided. The layered weighted method is used to allocate marine activities, prioritizing activities with strong spatial exclusivity in the dominant layer, while activities with high compatibility share the remaining space. For cross-layer activity combinations, a vertical safety distance suggestion is generated, and a three-dimensional spatial occupancy diagram is output to ensure that activities in each layer do not interfere with each other.

[0046] The initial layout plan is backtested and verified using a historical case library. The matching degree between the plan and existing successful cases is calculated. Options with deviations exceeding the preset deviation threshold are eliminated. The final optimized layout plan is determined and updated to the dynamic knowledge base, recording the layout parameters and decision-making basis.

[0047] Due to the adoption of the above technical solution, the technical progress achieved by this invention compared to the prior art is as follows:

[0048] This invention provides a method for identifying and planning high-potential marine three-dimensional spatial areas based on suitability assessment. By constructing a multi-dimensional suitability assessment system and a compatibility probability matrix, it can identify high-potential areas that simultaneously meet natural conditions, ecological constraints, and social needs. Furthermore, it avoids planar conflicts in traditional two-dimensional planning by using hierarchical weighted overlay analysis, and optimizes the vertical spatial resource allocation by using three-dimensional hierarchical rules, enabling multiple marine activities to be compatible in the same sea area and reducing inefficient or conflicting layouts.

[0049] This invention provides a method for identifying and deploying high-potential areas in a three-dimensional marine environment based on suitability assessment, using a compatibility prediction model driven by a graph neural network. By dynamically updating the relationship graph and performing real-time conflict detection, conflict propagation paths in medium- and low-potential areas can be marked in advance, and the potential risks of combinations of historical conflict patterns and new activities can be quantified. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0051] Figure 1 This is a schematic diagram of the workflow of the present invention;

[0052] Figure 2 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0054] Example 1, such as Figure 1 , Figure 2 As shown, this invention provides a method for identifying and distributing high-potential areas in marine three-dimensional space based on suitability assessment, comprising the following steps:

[0055] S1. Collect natural, social, and ecological data of the marine area, as well as historical three-dimensional rights confirmation project cases. Perform cleaning, transformation, and standardization preprocessing to unify the coordinate system and accuracy standards, and construct a basic database of marine three-dimensional space. Collect three major categories of data for the target marine area: natural, social, and ecological. Natural data includes water depth, seabed type, and marine disaster risk, while social data covers transportation networks, port location conditions, and tourism facilities. Ecological data involves marine ecological protection red lines, seawater quality, and distribution of biological resources. At the same time, collect historical three-dimensional rights confirmation project case data, including types of marine use activities, spatial layer distribution, conflict records, and solutions. Data sources include publicly available data from government departments, remote sensing images, industry reports, and field surveys. After collection, conduct preliminary screening to remove duplicate, missing, or obviously erroneous data to ensure the integrity and reliability of the data foundation. Clean the collected raw data, including standardizing the format, correcting outliers, and filling in missing values.

[0056] Simultaneously, data conversion was performed to unify data from different formats into a more processable format and standardize the data, unifying the units of measurement and classification standards to ensure that all spatial data are under the same coordinate system and meet project requirements in terms of accuracy. This eliminated errors and conflicts caused by inconsistent data formats and standards, improving data quality and consistency. Based on the processed data, a basic database architecture for marine three-dimensional spatial analysis was designed, determining the core table structure and field definitions. The core table structure of the database includes a marine basic information table, a natural data table, a social data table, an ecological data table, and a historical case table. Among them, the marine basic information table includes ID, coordinate range, and administrative division; the natural data table includes water depth, seabed sediment, and disaster level; the social data table includes transportation network and port location scores; the ecological data table includes red line range and water quality level; and the historical case table includes marine use combinations, conflict types, and solutions. The processed data was imported into the database according to the designed architecture, and data entry verification was performed to ensure accurate data storage. At the same time, a data index and query mechanism were established to form a basic database for marine three-dimensional spatial analysis.

[0057] S2. Based on graph neural network analysis of historical three-dimensional rights confirmation cases, marine activities and spatial layers are used as nodes, and relationships in historical cases are used as edges to construct a marine activity-spatial layer relationship graph, quantifying compatibility probabilities. Based on historical three-dimensional rights confirmation cases, marine activity types and spatial layers (water surface, water body, seabed, subsoil) are defined as two types of nodes in the graph. Node attributes include the type of marine activity, spatial range, temporal characteristics, and physical and ecological attributes of the spatial layer. Interactions in historical three-dimensional rights confirmation cases are used as edges, and edge attributes record conflict intensity. The attribute encoding format of nodes and edges is unified, and graph connectivity is ensured through topological verification. Each marine activity must be associated with at least one spatial layer, and spatial layer nodes must be indirectly connected through activity nodes. Isolated nodes or redundant edges are removed to form a structured initial graph. A graph neural network (graph convolutional network) is used to train the initial graph to learn the latent feature representations of nodes and edges. The input is the initial features of nodes and graph structure information, and the output is the low-dimensional embedding vectors of nodes and edges. Among them, node embeddings integrate the semantic features of activity types and spatial layers, and edge embeddings encode the latent patterns of compatibility or conflict.

[0058] During training, known compatibility labels from historical cases are used as supervision signals. The model parameters are optimized through the cross-entropy loss function, so that the embedding vectors can accurately reflect the compatibility probability distribution between different activities and spatial layers, improving the generalization ability of relationship prediction. The trained graph neural network model is obtained. Based on the trained graph neural network model, the compatibility of any activity-spatial layer combination in the graph is probabilistically quantified. By calculating the node embedding similarity and edge embedding weight, a compatibility probability matrix is ​​generated, where each element represents the compatibility score of a specific activity in a certain spatial layer. At the same time, a dynamic graph update mechanism is established to integrate the activity, spatial layer and relationship data in new rights confirmation cases into the graph in real time, retrain and update the model parameters to ensure that the compatibility quantification results reflect the latest requirements of marine use practices and output an interpretable marine use activity-spatial layer relationship graph.

[0059] The expression for node embedding similarity is as follows:

[0060] ;

[0061] In the formula, Embed similarity for nodes. For sea use activity nodes Embedding vectors of spatial layer nodes s (generated via GNN). For vector norm, A learnable weight matrix used for non-linear feature interactions. This is a vector concatenation operation. The balance coefficient is set to 0.5. A larger value indicates stronger semantic compatibility between the activity and the spatial layer;

[0062] The expression for edge embedding weights is as follows:

[0063] ;

[0064] In the formula, Embed weight values ​​for edges. For the edge The initial embedding, Let be the edge feature transformation matrix, and q be the attention weight vector. The activation function (slope = 0.2) helps alleviate the vanishing gradient problem. This is a learnable parameter vector (the transpose of a column vector, i.e., a row vector), used to map the concatenated high-order features to a scalar weight value. Edge features Embedded with both ends of the node , The higher-order representation of the splicing result after linear transformation. A value close to 1 indicates fewer historical conflicts or stronger synergistic effects;

[0065] The expression for the compatibility probability matrix is ​​as follows:

[0066] ;

[0067] In the formula, This is a compatibility probability matrix, representing marine activities. The compatibility probability in spatial layer s, The control coefficient controls the sharpness of the probability distribution. For strong compatibility, High conflict;

[0068] S3. Utilize the marine activity-spatial layer relationship graph to analyze the compatibility of different marine activity combinations in historical projects, extract compatibility and conflict patterns, and analyze its topological structure and node / edge features based on the constructed marine activity-spatial layer relationship graph. Then, identify high-frequency subgraph patterns using a breadth-first search-based graph traversal algorithm. Starting from high-frequency nodes, traverse their neighborhoods within 3 hops to extract recurring subgraph structures, thereby determining common connection methods between marine activities and the spatial layer. Calculate the degree of nodes using degree centrality to quantify node importance, retain the top 10% of key nodes, and filter out key activities and spatial layer nodes. Use graph embedding technology (Node2Vec) to map nodes and edges to a low-dimensional vector space, preserving their structural and attribute information. Group the embedded vectors using the K-means clustering algorithm to identify marine activity combinations with similar compatibility or conflict patterns.

[0069] Among them, the compatible mode is characterized by close node embedding distance and high edge weight, while the conflict mode corresponds to separate embedding or extremely low edge weight. The edge weight distribution is analyzed by combining the graph attention mechanism to extract the conflict transmission rules across spatial layers and form a pattern rule base. The universality of the pattern is verified by backtesting through historical cases. The matching degree between the extracted pattern and the real case is calculated to eliminate invalid patterns with overfitting or low coverage. The significance of the pattern is evaluated by statistical test to ensure its non-randomness. The verified patterns are classified according to conflict type to build a structured knowledge base.

[0070] The expression for degree centrality calculation is as follows:

[0071] ;

[0072] In the formula, Let v be the degree centrality of the nodes, v be the node to be computed, u be all other nodes in the graph (including activity and spatial layer nodes), V be the set of all nodes in the graph, and E be the set of all edges in the graph. This is an indicator function; it takes the value 1 if there is an edge between nodes v and u, and 0 otherwise. The larger the value, the more extensive the connections of the node in the graph;

[0073] The expression for embedding vector grouping is as follows:

[0074] ;

[0075] ;

[0076] ;

[0077] In the formula, J is the objective function of the K-means clustering algorithm, i.e., the intra-cluster sum of squares error, K is the preset number of clusters, and x is the node or edge embedding vector generated by Node2Vec. Let r be the set of vectors in the r-th cluster. Let J be the centroid of the r-th cluster. The smaller J is, the more similar the vectors within the cluster. The embedding distance of nodes within the cluster is small for compatible pattern clusters, while the nodes within the cluster are scattered for conflict pattern clusters.

[0078] The expression for the degree of matching between the pattern and the real case is as follows:

[0079] ;

[0080] In the formula, P represents the degree of matching between the pattern and real-world cases, and C represents the extracted subgraph pattern and the historical case set. For the graph substructure corresponding to the c-th case, For subgraph isomorphism relations, i.e., pattern P in It exists in The closer the value is to 1, the stronger the universality of the pattern.

[0081] The expression for the statistical test is as follows:

[0082] ;

[0083] In the formula, This is the chi-square statistic. The frequency of actual observations. The expected frequency when randomly distributed. The larger the value, the more significant the non-randomness of the pattern.

[0084] S4. Combining historical patterns, a graph neural network model is used to predict the compatibility probability of new marine activity combinations, identify potential conflicts, convert the historical marine activity-spatial layer relationship graph into a heterogeneous graph structure, clarify the node and edge types, and embed the conflict intensity attribute from the historical patterns as edge weights. A graph neural network is used to model heterogeneous interactions, and a multi-type node encoder is used to capture the feature differences between activities and the spatial layer. The attention mechanism is used to aggregate neighborhood information and generate node embeddings that fuse structure and attributes. During model training, the compatibility labels of historical activity combinations are used as supervision signals to optimize the cross-entropy loss function, enabling the network to learn the implicit feature distribution of compatibility and conflict patterns. For newly submitted marine activity combinations, they are mapped to the pre-trained heterogeneous graph as new nodes, and their embedding representations are updated through the message passing mechanism of the graph neural network.

[0085] Based on the connection between new marine activities and the existing spatial layer, the interaction weights between new marine activities and neighboring nodes are dynamically calculated. Combined with the feature distribution learned from historical patterns, the compatibility probability of the combination is output. If the predicted probability is lower than the preset conflict threshold, it is marked as a potential conflict. If it is higher than the conflict threshold, it is judged as compatible. For activity combinations predicted as potential conflicts, the conflict type and transmission path are further analyzed. Through the edge weight distribution of the graph neural network, the direct source and indirect transmission chain of the conflict are identified. The conflict pattern is extracted into structured rules and stored in the dynamic knowledge base. At the same time, newly confirmed compatibility / conflict cases are fed back to the model training set. The graph neural network parameters are updated using an incremental learning strategy to ensure that the model adapts to the emergence of new activities and continuously improves the prediction accuracy and pattern coverage.

[0086] Furthermore, the process of extracting conflict patterns into structured rules is as follows:

[0087] Based on the edge weight distribution of graph neural networks, the direct source node and indirect influence path of conflict are located. High-weight conflict edges (edge ​​weight > 0.7) directly connected to new active nodes are extracted to identify the direct conflicting parties. The gradient contribution of edge weights is calculated through backpropagation to trace the indirect transmission chain of cross-node hop count. At the same time, combined with the topology sorting algorithm, the nodes are sorted according to their dependencies to ensure that the path is acyclic and covers key nodes. The complete transmission path graph of conflict is output, and the edge weights and node types are labeled. Based on the node types, edge weights and historical labels in the transmission path, the conflict types are summarized, including spatial overlapping conflict, ecological chain conflict and temporal conflict.

[0088] Among them, spatial overlap conflict refers to physical occupancy at the same layer (overlapping area ratio), ecological chain conflict refers to cross-layer ecological impact, and temporal conflict refers to activity cycle contradiction. The conflict intensity (weight mean) and impact range (transmission path length) are quantified, and the identified transmission path is transformed into structured rules. Then, the rules are encoded into a machine-readable format through logical programming. Redundancy and conflict detection are performed between the new rules and the existing rules in the knowledge base. Similarity calculation based on graph embedding is used (cosine similarity > 0.9 is considered duplicate), duplicate rules are merged, low coverage rules are eliminated, and the matching degree between the new rule base and historical cases is verified (if the coverage decreases by > 5%, it is rolled back) to ensure compatibility with historical data. The core rules covering the top 20% are retained, timestamps are added to the rules, and outdated entries that have not been triggered for 5 years are periodically eliminated.

[0089] S5. Based on the analytic hierarchy process, determine the weights, conduct a stratified suitability evaluation of marine activities, generate scoring patches, and combine the compatibility probability prediction results to divide high, medium and low potential areas, and prioritize the recommendation of three-dimensional layout areas with low conflict probability.

[0090] S6. Based on high-potential areas, we recommend an optimized layout scheme to match the layout requirements of high-potential areas.

[0091] Example 2, as Figure 1 , Figure 2 As shown, based on Embodiment 1, the present invention provides a technical solution: Preferably, S5 specifically includes:

[0092] Based on the spatial characteristics of marine areas and the needs of marine use activities, a hierarchical suitability evaluation index system was established, including three primary indicators: natural, social, and ecological. These are further subdivided into secondary indicators, and then tertiary indicators. The Analytic Hierarchy Process (AHP) was used to determine the weights of each indicator. A judgment matrix was constructed, and consistency checks were performed. The weight distribution was obtained by calculating the eigenvectors, with natural conditions having higher weights (0.4-0.6), followed by ecological constraints (0.3-0.4), and social factors having lower weights (0.1-0.2). The weight results were used to standardize the data of each indicator, eliminating dimensional differences. Based on the weights, a weighted overlay analysis was performed on the rasterized data to generate a marine use activity suitability score map. The scoring range is 1-5 points, where 1 point is unsuitable and 5 points is optimal. The continuous scores are discretized into three levels using the natural breakpoint method: suitable (≥4 points), relatively suitable (3-4 points), and unsuitable (<3 points). Then, the marine functional zoning boundary is superimposed to remove prohibited development areas and form a preliminary suitable area distribution map. The suitability score patches are multiplied by the compatibility probability predicted by the graph neural network to obtain the comprehensive potential value. Potential areas are divided into high-potential areas, medium-potential areas, and low-potential areas according to percentiles. High-potential areas are given priority for three-dimensional layout, as they simultaneously meet the requirements of high suitability and high compatibility. Conflict risk points in medium-potential areas that require manual review are automatically marked, and low-potential areas are directly excluded.

[0093] The following is a classification of primary, secondary, and tertiary indicators, along with the data and data types for each indicator:

[0094]

[0095] Furthermore, the expression for the suitability scoring patch for marine activities is as follows:

[0096] ;

[0097] In the formula, For grid cells The appropriateness score (range 1-5 points). , , Weighting of natural, ecological, and social indicators. For the natural index k in The standardized score (1-5 points). , The standardized scores for the ecological and social indicators (1-5 points) are: n is the number of secondary and tertiary indicators under the natural indicators, m is the number of secondary and tertiary indicators under the ecological indicators, and q is the number of secondary and tertiary indicators under the social indicators.

[0098] The expression for the overall potential value is as follows:

[0099] ;

[0100] In the formula, For grid The overall potential value (range 0-5). For compatibility probability, This is an indicator function; 1 is used for non-prohibited development zones, and 0 otherwise. High-potential zones have high adaptability. And highly compatible ( The medium potential zone is characterized by excellent performance in either adaptability or compatibility, while the low potential zone is characterized by failure in any single indicator.

[0101] S6 specifically includes:

[0102] Based on the spatial characteristics and marine activity needs of high-potential areas, key parameters are extracted, including natural and social constraints such as water depth range, ecological sensitivity, and proximity to infrastructure. Simultaneously, in alignment with planning objectives, spatial overlay analysis is used to exclude areas conflicting with existing functional zoning, ensuring that recommended schemes comply with legal boundaries. Furthermore, by combining the compatibility probability matrix output by a graph neural network, grid units that simultaneously meet high suitability and high compatibility are selected to form an initial candidate area set. At this stage, it is crucial to avoid rigidly constrained areas such as ecological protection red lines and navigation channel anchorages. For candidate areas, based on the rules for three-dimensional stratified marine use, the water surface, The vertical spatial ownership of water bodies, seabed, and subsoil is allocated for marine activities using a layered weighted method. Activities with strong spatial exclusivity are prioritized for placement in the dominant layer, while highly compatible activities share the remaining space. For cross-layer activity combinations, a vertical safety distance suggestion is generated, and a three-dimensional spatial occupancy diagram is output to ensure that activities in each layer do not interfere with each other. The preliminary layout scheme is backtested and verified using a historical case library, and the matching degree between the scheme and existing successful cases is calculated. Options with deviations exceeding the preset deviation threshold are eliminated to determine the final optimized layout scheme. The optimized layout scheme is then synchronously updated to the dynamic knowledge base, recording layout parameters and decision-making basis.

[0103] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for identifying the layout of high-potential areas in the marine three-dimensional space based on suitability evaluation, characterized in that, The method comprises the following steps: S1, collecting natural, social, and ecological data of the sea area and historical stereoscopic right confirmation project cases, preprocessing to construct a stereoscopic spatial database of the sea area; S2, analyzing the historical stereoscopic right confirmation cases based on a graph neural network, and constructing a sea use activity-space layer relationship graph, specifically including: Based on the historical stereoscopic right confirmation cases, the sea use activity type and the space layer are defined as two types of nodes in the graph, the node attributes include the type of sea use activity, the spatial range, the time characteristics, and the physical and ecological attributes of the space layer, the interaction relationship in the historical stereoscopic right confirmation cases is taken as the edge, the edge attribute records the conflict intensity, and the attribute coding formats of the nodes and the edges are unified, the space layer nodes are indirectly connected through the activity nodes, the isolated nodes or redundant edges are removed, and a structured initial graph is formed; The initial graph is trained by using a graph neural network to learn the potential feature representation of the nodes and the edges, the input is the initial features of the nodes and the graph structure information, and the output is the low-dimensional embedding vectors of the nodes and the edges, wherein the node embedding fuses the semantic features of the activity type and the space layer, the edge embedding encodes the potential pattern of compatibility or conflict, and a trained graph neural network model is obtained; Based on the trained graph neural network model, the compatibility of any activity-space layer combination in the graph is quantified by probability, the node embedding similarity and the edge embedding weight value are calculated to generate a compatibility probability matrix, a dynamic updating mechanism of the graph is established, the activity, the space layer and the relationship data in the new right confirmation case are integrated into the graph in real time, the model parameters are retrained and updated, and an interpretable sea use activity-space layer relationship graph is output; The expression of the node embedding similarity is as follows: ； In the formula, is the node embedding similarity, is the marine activity node and the embedding vector of the spatial layer node s, is the vector norm, is a learnable weight matrix for nonlinear feature interaction, is a vector splicing operation, is a balance coefficient, set to 0.5; The expression of the edge embedding weight value is as follows: ； wherein, is an edge embedding weight value, is an edge is an initial embedding of an edge, is an edge feature transformation matrix, and q is an attention weight vector, is an activation function, is a learnable parameter vector, is an edge feature is a high-order representation of the concatenation result of the two end node embeddings after linear transformation. The expression of the compatibility probability matrix is as follows: ； wherein is a compatibility probability matrix, indicating the use of sea activities is a compatibility probability matrix, indicating the use of sea activities is a control coefficient, controlling the sharpness of the probability distribution, is a strong compatibility, is a high conflict; S3, using the sea use activity-space layer relationship graph to analyze the compatibility of different sea use activity combinations in the historical projects, and extracting the compatible and conflicting patterns; S4, combining the historical patterns, using the graph neural network model to predict the compatibility probability of new sea use activity combinations, and identifying potential conflicts; S5, determining the weight based on the analytic hierarchy process, performing hierarchical suitability evaluation on the sea use activities, generating a score plot, and combining the compatibility probability prediction results to divide the high, medium, and low potential areas; S6, based on the high potential area, recommending an optimized layout scheme to match the layout demand of the high potential area.

2. The layout method for identifying high-potential areas in the marine three-dimensional space based on suitability evaluation according to claim 1, characterized in that: The S1 specifically includes: Collecting natural, social, and ecological data of the target sea area, and collecting historical stereoscopic right confirmation project case data, including sea use activity types, space layer distribution, conflict records, and solutions; Cleaning the collected raw data, including format unification, abnormal value correction, and missing value filling, and performing data conversion and standardization processing to unify the measurement units and classification standards of the data; Based on the processed data, the three-dimensional spatial database framework of the sea area is designed, the core table structure and field definition are determined, the core table structure of the database includes sea area basic information table, natural data table, social data table, ecological data table and historical case table, the processed data is imported into the database according to the designed framework, data storage verification is carried out, and data index and query mechanism are established to form the three-dimensional spatial database of the sea area. 3.The method for identifying the layout of high-potential areas in marine stereoscopic space based on suitability evaluation according to claim 1, characterized in that: The S3 specifically includes: Based on the constructed sea use activity-space layer relationship graph, the topological structure and node / edge characteristics are analyzed, and the high-frequency subgraph mode is identified through the breadth-first search based graph traversal algorithm. Starting from the high-frequency node, the neighborhood within 3 hops is traversed, the repeatedly appearing subgraph structure is extracted, and the common connection mode of sea use activity and space layer is determined. The degree of the node is calculated by using the degree centrality, and the importance of the node is quantified. The Top10% key nodes are reserved, and the key activities and space layer nodes are screened out; The graph embedding technology is used to map the nodes and edges to the low-dimensional vector space, and the structure and attribute information are reserved. The embedded vectors are grouped by using the K-means clustering algorithm, the sea use activity combinations with similar compatible mode or conflict mode are identified, the edge weight distribution is analyzed combined with the graph attention mechanism, the conflict transmission rule across the space layer is extracted, and the mode rule library is formed; The universality of the mode is verified through historical case back testing, the matching degree of the extracted mode and the real case is calculated, the invalid mode with overfitting or low coverage is eliminated, and the significance of the mode is evaluated by using statistical test. The verified mode is classified according to the conflict type, and the structured knowledge base is constructed. 4.The method for identifying the layout of high-potential areas in marine stereoscopic space based on suitability evaluation according to claim 1, characterized in that: The S4 specifically includes: The historical sea use activity-space layer relationship graph is converted into a heterogeneous graph structure, the node type and edge type are determined, and the conflict intensity attribute in the historical mode is embedded as the edge weight. The heterogeneous interaction is modeled by using the graph neural network. The feature differences of activities and space layers are captured by using the multi-type node encoder. The neighborhood information is aggregated by using the attention mechanism, and the node embedding integrating structure and attribute is generated; For the newly submitted sea use activity combination, it is mapped to the pre-trained heterogeneous graph as a new node, and its embedding representation is updated by using the message passing mechanism of the graph neural network. According to the connection relationship between the new sea use activity and the existing space layer, the interaction weight between the new node and the neighborhood nodes is dynamically calculated. The compatible probability of the combination is output by combining the feature distribution learned in the historical mode. If the predicted probability is lower than the preset conflict threshold, it is marked as potential conflict. If it is higher than the conflict threshold, it is determined as compatible; For the activity combination predicted as potential conflict, the conflict type and transmission path are further analyzed, the direct source and indirect transmission chain of the conflict are identified by using the edge weight distribution of the graph neural network, the conflict mode is extracted as a structured rule, and is stored in the dynamic knowledge base.

5. The layout method for identifying high-potential areas in oceanic space based on suitability evaluation according to claim 4, characterized in that: The process of extracting the conflict mode into a structured rule is: Based on the edge weight distribution of the graph neural network, the direct source nodes and indirect influence paths of conflicts are located, high-weight conflict edges directly connected to the new activity node are extracted, the direct conflict parties are identified, the gradient contribution of the edge weight is calculated through back propagation, the indirect transmission chain across node hops is tracked, and meanwhile, the topological sorting algorithm is combined to sort the node dependency relationship and output the complete transmission path graph of the conflict, with the edge weight and node type labeled; Based on the node type, edge weight and historical label in the transmission path, the conflict type is summarized, including spatial overlap conflict, ecological chain conflict and time sequence conflict, the conflict intensity and influence range are quantified, and the identified transmission path is converted into a structured rule, which is then encoded into a machine-readable format through logical programming; The new rule and the existing rule in the knowledge base are detected for redundancy and conflict, the similarity calculation based on graph embedding is adopted to merge repeated rules, eliminate low-coverage rules, verify the matching degree of the new rule base and historical cases, retain the top 20% of core rules, add a timestamp to the rule, and periodically eliminate obsolete items that have not been triggered for 5 years. 6.The layout method for identifying high-potential areas in oceanic stereoscopic space based on suitability evaluation according to claim 1, characterized in that: The S5 specifically includes: Based on the spatial characteristics of the sea area and the demand for sea use activities, a hierarchical suitability evaluation index system is established, including three first-level indexes of nature, society and ecology, with second-level indexes and third-level indexes, and the weights of each index are determined by the analytic hierarchy process, the weight distribution is obtained by constructing a judgment matrix and conducting consistency test, and the characteristic vector is calculated; Based on the weight, the weighted overlay analysis is performed on the rasterized data to generate the sea use activity suitability scoring polygons, the scoring interval is 1-5 points, among which 1 point is not suitable and 5 points are optimal, and the natural breakpoint method is used to discretize the continuous score into three levels: suitable for ≥4 points, more suitable for 3-4 points, and not suitable for <3 points, and then the marine function zoning boundary is superimposed to exclude the prohibited development area to form a preliminary suitable area distribution map; The suitability scoring polygons are multiplied by the compatibility probability predicted by the graph neural network to obtain the comprehensive potential value, and the potential area is divided into high potential area, medium potential area and low potential area according to the percentile, and the high potential area is preferentially recommended for three-dimensional layout, which meets the high suitability and high compatibility at the same time, the medium potential area is automatically marked for manual review of conflict risk points, and the low potential area is directly excluded.

7. The layout method for identifying high-potential areas in oceanic space based on suitability evaluation according to claim 6, characterized in that: The expression of the sea use activity suitability scoring polygon is as follows: ； wherein, Suitability total score of the grid cell, Weight of natural, ecological and social indicators, Normalized score of natural indicator k in Normalized score of ecological and social indicators, n is the number of secondary indicators and tertiary indicators under the natural indicator, m is the number of secondary indicators and tertiary indicators under the ecological indicator, and q is the number of secondary indicators and tertiary indicators under the social indicator.​​​​​ The expression of the comprehensive potential value is as follows: ； In the formula, For grid The overall potential value, For compatibility probability, This is an indicator function; it is 1 for non-restricted development zones and 0 otherwise.

8. The layout method for identifying high-potential areas in oceanic space based on suitability evaluation according to claim 7, characterized in that: The S6 specifically includes: Based on the spatial characteristics of the high potential area and the demand for sea use activities, key parameters are extracted, including water depth range, ecological sensitivity and natural and social constraints of adjacent infrastructure, and the planning target is connected, through spatial overlay analysis, the area in conflict with the existing functional zoning is excluded, and combined with the compatibility probability matrix output by the graph neural network, the grid cells that meet the high suitability and high compatibility at the same time are selected to form an initial candidate area set; For the candidate area, according to the rules of sea stereoscopic layering, the vertical space ownership range of water surface, water body, seabed and bottom soil is divided, and the layered weighting method is used to allocate sea activities. The activities with strong spatial exclusivity are arranged in the dominant layer, and the activities with high compatibility share the remaining space. For the cross-layer activity combination, the vertical safety distance is suggested, and the three-dimensional space occupation diagram is output. The historical case library is used to backtest and verify the preliminary layout scheme, calculate the matching degree of the scheme and the existing successful cases, eliminate the options with deviation exceeding the preset deviation threshold, determine the final optimized layout scheme, and update the optimized layout scheme to the dynamic knowledge base, and record the layout parameters and decision basis.

Images