Submarine cable construction analog simulation method under complex seabed geological conditions

Abstract

The invention relates to the technical field of seabed engineering, and discloses a submarine cable construction simulation method under complex seabed geological conditions. The method comprises the following steps: slicing and recombining multi-source investigation time sequence data, and constructing a geologic feature space based on a multi-dimensional time sequence window and time-space synchronization. Through feature fusion and dimension mapping, three-dimensional geological semantic body units with consistent time and space are generated, and attributes of the three-dimensional geological semantic body units and construction process parameters are associated and coded. And a recursive segmentation algorithm is adopted, and layered and nested dynamic construction decision units are automatically divided according to an attribute parameter mutation threshold. On the basis, a simulation agent is initialized for each decision-making unit, and dynamic re-evaluation of attribute parameters is driven through an agent interaction network. According to the method, the fidelity of the simulation model to the space-time evolution characteristics of the complex geological conditions is improved, the construction decision can adaptively respond to the local mutation of the geological parameters, and the simulation accuracy and the engineering practicability are enhanced.

Description

Technical Field

[0001] This invention relates to the field of submarine engineering technology, specifically to a simulation method for submarine cable construction under complex geological conditions on the seabed. Background Technology

[0002] Current simulations of submarine cable construction typically rely on static overlay or simple interpolation of multi-source survey data when constructing geological models. These data may originate from survey equipment of varying time and precision, making it difficult to guarantee strict synchronization of multiple geological attribute parameters across time series and spatial coordinates using conventional methods. Consequently, the simulation models developed suffer from spatiotemporal inconsistencies, weakening the dynamic correlation and continuous evolution of geological attributes, resulting in limited fidelity in representing the complex geological conditions of the real seabed.

[0003] In the construction planning and decision-making simulation phase, traditional methods often use fixed-size grid cells to divide the construction area, or rely on engineers' experience for manual partitioning. Static, homogeneous spatial segmentation methods cannot accurately capture and respond to the nonlinear abrupt changes and heterogeneous distribution of seabed geological parameters in space. The boundary delineation of construction decision units is often disconnected from actual changes in geological conditions, making it difficult to dynamically adapt to the specific impact of local geological abrupt changes on construction technology selection. This limits the simulation's ability to realistically simulate the dynamic evolution of construction interactions and decision chains under complex conditions. Summary of the Invention

[0004] The purpose of this invention is to provide a simulation method for submarine cable construction under complex geological conditions on the seabed, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, this invention provides a simulation method for submarine cable construction under complex seabed geological conditions, the method comprising: A multidimensional geological feature space based on a multidimensional time series window is constructed. The data of the multidimensional geological feature space comes from the recombination of slices of multi-source survey time series data streams. Each time series window contains a continuous sampling sequence of multiple seabed topographic attribute parameters synchronized with geographic coordinates. For the continuous sampling sequence within each time window in the multidimensional geological feature space, feature fusion and dimension mapping are performed to generate a three-dimensional geological semantic volume unit with spatiotemporal consistency, and the attribute parameters of the three-dimensional geological semantic volume unit are correlated with the preset submarine cable construction process parameters. The complete simulation domain composed of three-dimensional geological semantic volume units is spatially segmented by a recursive segmentation algorithm to generate a series of dynamic construction decision units with hierarchical nested structures. The boundary of each dynamic construction decision unit is automatically defined by the mutation threshold of the attribute parameters of the three-dimensional geological semantic volume unit. The process suitability of each dynamic construction decision unit is marked using the correlation coding results, and decision labels containing potential construction constraints are generated. The initial state of the simulation agent corresponding to each dynamic construction decision unit is initialized based on the decision labels. The simulation agent with initial state is embedded into the discrete event simulation engine. An interaction network is constructed based on the spatiotemporal proximity between the simulation agents. The decision label update instruction is transmitted through the interaction network to drive the re-evaluation of the attribute parameters of the three-dimensional geological semantic volume unit among the simulation agents.

[0006] Preferably, the spatial segmentation of the complete simulation domain composed of three-dimensional geological semantic volume units using a recursive segmentation algorithm generates a series of dynamic construction decision units with a hierarchical nested structure, including: The gradient change of the attribute parameters of the three-dimensional geological semantic volume unit in space is used as the trigger condition for recursive segmentation. When the gradient value of the attribute parameter exceeds the preset global mutation threshold, the initial spatial segmentation is performed at the gradient mutation location. Within each sub-region formed by the initial spatial segmentation, the local gradient distribution of the attribute parameters of the three-dimensional geological semantic volume unit is recalculated, and the sub-region is recursively subdivided according to the local mutation threshold until the gradient change of the attribute parameters within the sub-region stabilizes within the preset smoothing range. Assign a unique hierarchical identifier to each new spatial region generated by recursive segmentation, and establish a subordinate link between the hierarchical identifier and the parent region identifier to form a dynamic construction decision unit tree with a hierarchical nested structure. Based on the geometric center coordinates and hierarchical identifier of each dynamic construction decision unit, calculate its spatial topological relationship with adjacent units, and record the spatial topological relationship in the additional attributes of the dynamic construction decision unit tree; The dynamic construction decision unit tree is registered and aligned with the multidimensional geological feature space to ensure that each dynamic construction decision unit can be back-indexed to one or more three-dimensional geological semantic units that constitute it.

[0007] Preferably, the step of calculating the spatial topological relationship between each dynamic construction decision unit and its adjacent units based on the geometric center coordinates and its hierarchical identifier includes: Extract the geometric center coordinates of all leaf nodes in the dynamic construction decision unit tree, and construct a spatial adjacency graph of all leaf nodes on a two-dimensional plane based on the Delaunay triangulation algorithm. In the spatial adjacency diagram, the adjacency relationships of leaf nodes are aggregated layer by layer from bottom to top according to the subordinate link relationship of the hierarchical identifier, and the spatial topology relationship of each non-leaf node dynamic construction decision unit is derived at the corresponding level. The derived spatial topological relationships are encoded as adjacency matrices. The rows and columns of the matrix correspond to all units at the same level in the dynamic construction decision unit tree. The values ​​of the matrix elements indicate whether there is a shared boundary or corner contact between the corresponding two units. Based on the adjacency matrix, each dynamic construction decision unit is traversed, all its directly adjacent units are identified, and a list of adjacent unit identifiers is generated. The list of adjacent unit identifiers is used as an additional attribute of the dynamic construction decision unit tree and is bound and stored with the hierarchical identifier of the corresponding unit.

[0008] Preferably, the step of binding and storing the list of adjacent unit identifiers as an additional attribute of the dynamic construction decision unit tree with the hierarchical identifiers of the corresponding units includes: In response to the simulation agent receiving the decision label update instruction, the simulation agent retrieves the list of adjacent unit identifiers bound to the storage based on the hierarchical identifier of the dynamic construction decision unit to which it belongs. Based on the list of adjacent unit identifiers, the simulation agent initiates a status query request to the simulation agents corresponding to all adjacent dynamic construction decision units in order to obtain the current decision labels and process suitability labels of the adjacent units. The simulation agent integrates the process suitability label of its own unit with the current decision label obtained from neighboring units, executes conflict resolution logic based on the integration result, and generates updated process suitability labels. The updated process suitability label is used as the new decision label, and it is determined whether the difference between it and the original decision label exceeds the tolerance threshold. If it does, a new decision label update instruction is sent to the simulation agent corresponding to the adjacent unit through the interactive network.

[0009] Preferably, the step of using the correlation coding results to mark the process suitability of each dynamic construction decision unit includes: The correlation encoding between the attribute parameters of three-dimensional geological semantic units and the submarine cable construction process parameters is analyzed, and the key geological parameters that constrain construction decisions and their threshold ranges are extracted. The real-time values ​​of key geological parameters are compared with the corresponding threshold ranges. Based on the comparison results, an initial set of process constraints for the dynamic construction decision unit is generated. The set of process constraints includes the allowed construction process types and the corresponding parameter adjustment ranges. Based on the spatial topology of the dynamic construction decision unit, check the compatibility of the process constraint sets between adjacent units. If there are directly conflicting process type requirements, negotiate and prune the process constraint sets to remove the process options that cause the conflict. The negotiated and pruned set of process constraints is matched and verified with the real-time sea state data stream extracted from the multi-source survey time series data stream. Process options that are not feasible under the current sea state conditions are eliminated to form the final process suitability label. The final process suitability label is bound to the hierarchical identifier of the corresponding dynamic construction decision unit and injected into the initial state of the simulation agent corresponding to the dynamic construction decision unit.

[0010] Preferably, after binding the final process suitability label with the hierarchical identifier of the corresponding dynamic construction decision unit and injecting it into the initial state of the simulation agent corresponding to the dynamic construction decision unit, the process includes: Throughout its lifecycle, the simulation agent continuously monitors data update events from the multidimensional geological feature space. When it detects an update in the attribute parameters of a three-dimensional geological semantic unit associated with its own unit, it triggers a re-evaluation of the process suitability label. The simulation agent obtains the updated attribute parameters and re-executes the comparison with the threshold range based on the latest key geological parameter values ​​to generate an updated set of process constraints. The simulation agent performs a secondary matching and verification between the updated set of process constraints and the current sea state data stream, and integrates its current decision labels to form a temporary process suitability label. The simulation agent pre-negotiates the temporary process suitability label with the adjacent cells pointed to by the bound storage adjacent cell identifier list. If the pre-negotiation is successful, the temporary process suitability label is confirmed as the new decision label; otherwise, the original decision label remains unchanged.

[0011] Preferably, the construction of a multidimensional geological feature space based on a multidimensional time-series window includes: It receives multi-source time-series survey data streams from side-scan sonar, shallow seismic profiler and multibeam echo sounder, each data stream having a unified timestamp and high-precision geographic coordinates; Using a fixed duration or fixed geographic displacement as a sliding window, each multi-source survey time-series data stream is synchronously sliced ​​to generate multiple time-aligned data slices. Spatial interpolation and gridding are performed on the data slices within each time window to unify terrain and geological attribute data from different sources and at different resolutions into a grid with the same spatial resolution. The multiple geological attribute values ​​of each grid point are organized into a feature vector in a multidimensional geological feature space according to its geographical coordinates and timestamp; The system receives construction feedback data streams from a high-precision DGPS positioning system and a buried plow attitude device in real time, and uses these data streams as a dynamic attribute layer to be integrated into the corresponding spatiotemporal location in a multidimensional geological feature space in real time.

[0012] Preferably, the real-time reception of construction feedback data streams from the high-precision DGPS positioning system and the buried plow attitude device, and the real-time fusion of these construction feedback data streams as a dynamic attribute layer into the corresponding spatiotemporal location in the multidimensional geological feature space, includes: The construction feedback data stream was analyzed to extract the real-time spatial coordinates of the submarine cable, the real-time attitude angle of the burying plow, the real-time tension of the submarine cable, and the real-time burial depth data. Centered on the real-time spatial coordinates of the submarine cable, a dynamic fusion influence area is established. The attribute correction weights of all grid points in the multi-dimensional geological feature space within the dynamic fusion influence area are calculated. The attribute correction weights are inversely proportional to the distance from the grid point to the real-time spatial coordinates of the submarine cable. Based on real-time attitude angle, real-time cable tension and real-time burial depth data, the real-time disturbance estimate of the seabed topography is calculated, and the real-time disturbance estimate is distributed to each grid point in the dynamic fusion influence area according to the attribute correction weight. The allocated real-time disturbance estimate is weighted and superimposed with the original attribute values ​​of the corresponding grid points in the multidimensional geological feature space to update the attribute values ​​of the grid points, thereby completing the real-time fusion of the construction feedback data stream.

[0013] Preferably, the simulation agent integrates the process suitability label of its own unit with the current decision label obtained from adjacent units, and executes conflict resolution logic based on the integration result, including: The simulation agent establishes a conflict detection rule base, which defines the possible mutual exclusion relationships, order dependencies, and parameter compatibility ranges between different process suitability labels. The process suitability label of the unit itself is matched with the current decision label obtained from all neighboring units and input into the conflict detection rule base for matching and detection to identify all existing conflict relationships and their conflict types. For each identified conflict relationship, a preset conflict resolution strategy is invoked, which includes adjusting process parameters, rearranging construction sequence, or replacing process type. Following the priority order of conflict resolution strategies, each strategy is tried in turn, and it is verified whether the identified conflict relationships are eliminated after application, until the first effective strategy combination that can eliminate all identified conflict relationships is found. Based on the effective strategy combination, modify the process suitability label of its own unit to generate an updated process suitability label.

[0014] Preferably, the simulation agent pre-negotiates the temporary process suitability label with the adjacent cells pointed to by the bound and stored list of adjacent cell identifiers, including: The simulation agent generates a pre-negotiation request message, which contains temporary process suitability markings and details of the updated key geological parameters that led to the marking changes; The simulation agent sends a pre-negotiation request message to the simulation agents corresponding to all adjacent cells in the adjacent cell identifier list and starts the pre-negotiation timer; After receiving the pre-negotiation request message, the simulation agent corresponding to the adjacent unit evaluates the potential impact of the temporary process suitability label on it based on its current state and conflict detection rule base, and generates a pre-negotiation response message of agreement or disagreement. Before the pre-negotiation timer expires, the simulation agent that initiates the pre-negotiation collects pre-negotiation response messages from all adjacent units. If all responses are in agreement, the pre-negotiation is considered successful; if any opposition response is received, the pre-negotiation is considered unsuccessful. The pre-negotiation results are fed back to the simulation agent as the basis for its confirmation or rejection of the temporary process suitability label.

[0015] Compared with the prior art, the beneficial effects of the present invention are: For multi-source survey time-series data streams, a multi-dimensional time-series window is used for slicing and reorganization to ensure that each time-series window captures a continuous sampling sequence of multiple seabed topographic attribute parameters under strict geographic coordinate synchronization. By performing feature fusion and dimensional mapping on the sequences within the window, the generated three-dimensional geological semantic volume units have inherent spatiotemporal consistency. This allows the basic building blocks of the simulation model to directly integrate multi-dimensional, time-series original survey data stream information, rather than isolated, static data snapshots. The simulated geological environment constructed in this way can more realistically reflect the continuous state and correlation of seabed attributes under spatiotemporal coupling, providing a high-fidelity dynamic geological scene foundation for subsequent simulations.

[0016] A recursive segmentation algorithm is used to spatially segment the simulation domain, which consists of three-dimensional geological semantic units. The segmentation boundaries are not pre-defined but are automatically and recursively determined by the mutation thresholds of the attribute parameters of the three-dimensional geological semantic units themselves. The spatial segmentation process is directly driven by the objective changes in geological features. The generated dynamic construction decision units have a hierarchical nested structure, and their shape and scale are determined by the actual distribution of geological heterogeneity and abrupt interfaces, achieving adaptive matching between the construction decision units and complex geological structures. The decision logic in the construction simulation can respond more precisely to abrupt changes in local geological conditions based on these dynamic units, thereby simulating a more realistic, heterogeneous construction decision-making and interaction process. Attached Figure Description

[0017] Figure 1 This is a schematic diagram illustrating the working principle of the simulation method for submarine cable construction under complex geological conditions on the seabed, as described in this invention. Figure 2A flowchart for calculating and storing spatial topological relationships; Figure 3 This is a flowchart illustrating the state update negotiation process between simulated agents based on an adjacency list. Figure 4 A comparison chart of the comprehensive differences in decision labels for dynamic construction decision-making units; Figure 5 A heatmap of attribute correction weight distribution for dynamic fusion of feedback from submarine cable construction. Detailed Implementation

[0018] 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, and 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.

[0019] Please see Figure 1 This invention provides a simulation method for submarine cable construction under complex seabed geological conditions. The method includes: receiving multi-source survey time-series data streams from multi-source survey equipment; slicing and recombining the data streams using a fixed duration or geographic displacement as a sliding window to construct a multi-dimensional geological feature space based on multi-dimensional time-series windows. Each time-series window in this space contains a continuous sampling sequence of multiple seabed topographic attribute parameters with strictly synchronized geographic coordinates. For the continuous sampling sequence under each time-series window in the multi-dimensional geological feature space, feature fusion and dimensionality reduction operations are performed to map it into a three-dimensional geological semantic volume unit with spatiotemporal consistency. The attribute parameters of this unit are then correlated with a preset submarine cable construction process parameter system. A recursive segmentation algorithm is used to spatially partition the complete simulation domain aggregated from the three-dimensional geological semantic volume units. This algorithm automatically triggers segmentation based on the spatial gradient changes of the attribute parameters, generating a series of dynamic construction decision units with a hierarchical nested structure. Using the results of the above correlation encoding, process suitability calculation and conflict checking are performed on each dynamic construction decision unit to generate a decision label containing potential construction constraints. Based on this, the initial state of the simulation agent corresponding to the unit is initialized. All initialized simulation agents are embedded into the discrete event simulation engine. An interactive network is constructed based on the spatial proximity relationship between the dynamic construction decision units represented by the simulation agents. The simulation agents transmit decision label update instructions through this interactive network, driving the re-evaluation and state evolution of the attribute parameters of the three-dimensional geological semantic volume units among the simulation agents.

[0020] Example 1: See Figure 2The algorithm uses the spatial gradient changes of attribute parameters of 3D geological semantic units as the trigger condition for recursive segmentation, and the global mutation threshold is set based on historical geological data and construction experience. When the gradient value of the attribute parameter exceeds the preset global mutation threshold, the recursive segmentation algorithm performs the initial spatial segmentation at the location of the gradient mutation, dividing the complete simulation domain into several sub-regions. Within each sub-region formed by the initial segmentation, the algorithm recalculates the local gradient distribution of the attribute parameters of the internal 3D geological semantic units and performs recursive secondary segmentation of the sub-region based on a more refined local mutation threshold. The recursive process continues until the gradient changes of attribute parameters in each final sub-region stabilize within a preset smoothing interval. A unique hierarchical identifier is assigned to each new spatial region generated by each recursive segmentation, and a subordinate link relationship is established between this identifier and its parent region identifier, thereby forming a dynamic construction decision unit tree with a clear hierarchical structure. After the dynamic construction decision unit tree is constructed, the geometric center coordinates of all leaf nodes in the tree are extracted, and a spatial adjacency graph of these leaf nodes on a two-dimensional plane is constructed based on the Delaunay triangulation algorithm. Based on the hierarchical link relationships of the hierarchical identifiers, the adjacency relationships of leaf nodes are aggregated layer by layer from bottom to top on this spatial adjacency graph, thereby deriving the spatial topological relationship at the corresponding level for each non-leaf node dynamic construction decision unit. The derived spatial topological relationships at each level are encoded into an adjacency matrix, where the rows and columns correspond to all units at the same level in the dynamic construction decision unit tree, and the values ​​of the matrix elements indicate whether there is a shared boundary or only corner contact between two corresponding units. Based on the generated adjacency matrix, each dynamic construction decision unit is traversed, identifying all its directly adjacent units and generating a list of adjacent unit identifiers. This list of adjacent unit identifiers serves as a key additional attribute of the dynamic construction decision unit tree and is bound and stored with the hierarchical identifier of the corresponding unit.

[0021] Finally, the dynamic construction decision unit tree is registered and aligned with the multidimensional geological feature space to ensure that each dynamic construction decision unit can be accurately back-indexed to one or more three-dimensional geological semantic units that constitute it, thus realizing the association between the spatial segmentation results and the original geological data. In the specific implementation, for the simulation of a rectangular seabed area, its multidimensional geological feature space contains a processed and generated regular three-dimensional geological semantic unit mesh. Each three-dimensional geological semantic unit has three attribute parameters: seabed elevation, sediment shear strength, and gravel content percentage. The global mutation threshold of the recursive segmentation algorithm is set as follows: seabed elevation gradient exceeding 0.15 meters per meter, sediment shear strength gradient exceeding 5 kPa per meter, and gravel content percentage gradient exceeding 2% per meter. In the specific implementation, the algorithm traverses all three-dimensional geological semantic units and calculates the gradients of their attribute parameters in the east-west and north-south directions. When it is detected that the gradient value of any attribute parameter in a set of three-dimensional geological semantic units at a certain spatial location exceeds the corresponding global mutation threshold, for example, near the coordinate region (X100, Y200), the calculated value of the seabed elevation gradient is 0.22 meters per meter, the algorithm performs the initial spatial segmentation at this gradient mutation location, dividing the complete rectangular simulation domain into two sub-regions A and B along the mutation boundary.

[0022] In some embodiments, within sub-region A formed by the initial segmentation, the algorithm recalculates the local gradient distribution of all three-dimensional geological semantic units within that region and makes judgments based on a more refined local mutation threshold, which can be set to 60% of the corresponding global threshold. For example, within a local area of ​​sub-region A, the local gradient threshold for sediment shear strength is 3 kPa / m. When a sediment shear strength gradient of 3.5 kPa / m is detected at a local location in sub-region A, the algorithm performs recursive secondary segmentation on sub-region A at that location, generating sub-regions A1 and A2. The recursive segmentation process continues, iteratively calculating the internal gradient of each newly generated sub-region and determining whether further segmentation is needed based on its corresponding local mutation threshold, until the gradient change amplitude of any attribute parameter within all sub-regions stabilizes within a preset smoothing range, such as all absolute gradient values ​​being less than 0.05 m / m, 1 kPa / m, and 0.5% per m.

[0023] In practice, each recursive partitioning generates a unique hierarchical identifier for a new spatial region. For example, the identifier for the complete simulation domain is "R0", and the identifiers for the two sub-regions generated by the initial partitioning are "R0-1" and "R0-2". Sub-regions generated by secondary partitioning of "R0-1" are identified as "R0-1-1" and "R0-1-2". Simultaneously, subordinate links between sub-region identifiers and parent region identifiers are established, such as "R0-1-1" having "R0-1" as its parent identifier and "R0-1" having "R0" as its parent identifier, thus forming a dynamic construction decision unit tree with a hierarchical nested structure. In practice, based on the geometric center coordinates and hierarchical identifier of each dynamic construction decision unit, its spatial topological relationship with adjacent units is calculated. This process first extracts all leaf nodes in the dynamic construction decision unit tree, i.e., the geometric center coordinates of the smallest dynamic construction decision unit that is no longer to be partitioned, and constructs a triangulation network of these leaf nodes on a two-dimensional plane based on the Delaunay triangulation algorithm, thereby forming a spatial adjacency graph at the leaf node level.

[0024] In some embodiments, after obtaining the spatial adjacency graph of the leaf nodes, the algorithm aggregates layer by layer from bottom to top according to the subordinate links of the hierarchical identifiers. For example, leaf nodes "R0-1-1-1" and "R0-1-1-2" share a common parent node "R0-1-1". If these two leaf nodes are determined to be adjacent in the adjacency graph, the algorithm derives the spatial topological relationship between the parent node "R0-1-1" and other parent nodes at the same level, based on the adjacency relationship of its child nodes. The spatial topological relationship derived at each level is encoded into an adjacency matrix, where the rows and columns of the matrix correspond to all units at the same level in the dynamic construction decision unit tree. The values ​​of the matrix elements are used to represent the spatial relationship between the corresponding two units; for example, a value of 1 indicates a shared boundary, a value of 0.5 indicates only corner contact, and a value of 0 indicates no adjacency.

[0025] In practical implementation, based on the generated adjacency matrix, the algorithm traverses each dynamic construction decision unit at the current level, identifies all its directly adjacent units, and generates a list of adjacent unit identifiers. This adjacent unit identifier list can be understood as a key additional attribute of the dynamic construction decision unit tree, bound and stored with the corresponding unit's hierarchical identifier. In practical implementation, the dynamic construction decision unit tree is registered and aligned with the multidimensional geological feature space by establishing a mapping relationship between the spatial polygonal boundaries of the dynamic construction decision unit and the grid coordinate system of the multidimensional geological feature space. This ensures that each dynamic construction decision unit can, through its spatial extent, back-index one or more three-dimensional geological semantic volume units and all their attribute parameters located within that extent.

[0026] Example 2: See Figure 3During its operation, the simulation agent responds to the receipt of decision label update instructions. At this time, the simulation agent retrieves a list of neighboring unit identifiers bound to its own hierarchical identifier, based on the hierarchical identifier of its own dynamic construction decision unit. The simulation agent then initiates status query requests to the simulation agents corresponding to all adjacent dynamic construction decision units listed in the list to obtain the current decision labels and process suitability labels of these neighboring units. After integrating its own unit's process suitability labels with the current decision labels obtained from all neighboring units, the simulation agent executes conflict resolution logic. This process involves the simulation agent maintaining a conflict detection rule base, which explicitly defines the possible mutual exclusion relationships, sequence dependencies, and parameter compatibility ranges between different process suitability labels. The integration process inputs its own label and neighboring labels into this conflict detection rule base for matching and detection, identifying all existing conflict relationships and their specific conflict types. For each identified conflict relationship, a preset conflict resolution strategy is invoked, including process parameter adjustment, construction sequence rearrangement, or process type replacement. Following a pre-defined priority order for conflict resolution strategies, each strategy is applied sequentially, and the effectiveness of each strategy in eliminating all identified conflicts is verified until the first effective strategy combination that eliminates all identified conflicts is found. Based on the determined effective strategy combination, the process suitability label of the simulation agent's own unit is modified, thereby generating an updated process suitability label. The simulation agent uses this updated process suitability label as a new decision label and determines whether its difference from the original decision label exceeds a pre-defined tolerance threshold. If the difference exceeds the tolerance threshold, the simulation agent sends a new decision label update instruction to all its neighboring unit's corresponding simulation agents through the interactive network to drive a chain response in the neighboring units.

[0027] In practice, the simulation agent "Agent_R0-1-2" receives a decision label update instruction from the interactive network. The instruction is triggered because the geological parameters of adjacent areas have been revised in the simulation. Based on the hierarchical identifier of its dynamic construction decision unit "R0-1-2", the simulation agent "Agent_R0-1-2" retrieves a list of adjacent unit identifiers from the bound storage data structure. The list contains ["R0-1-1", "R0-2-1", "R0-2-2"]. Using this list of adjacent unit identifiers, the simulation agent "Agent_R0-1-2" initiates a synchronization status query request to the simulation agents corresponding to the identifiers "R0-1-1", "R0-2-1", and "R0-2-2" through the message passing mechanism built into the discrete event simulation engine. After the status query request is sent, the simulation agent "Agent_R0-1-2" waits for and obtains the simulation agent responses from the three adjacent dynamic construction decision units. The responses contain the current decision label and process suitability label of each unit, such as obtaining "Process A - Parameter Range P1", "Process B", and "Process C - Parameter Range P2".

[0028] The simulation agent "Agent_R0-1-2" integrates its own unit's process suitability label "Process A - Parameter Range P3" with the current decision labels obtained from adjacent units, and executes conflict resolution logic based on the integration result. The execution of this conflict resolution logic is based on a conflict detection rule base established and maintained internally by the simulation agent. This rule base explicitly defines the mutual exclusion relationships, sequential dependencies, and parameter compatibility ranges between different process suitability labels. For example, one rule defines: if two adjacent units use "Process A" and "Process B" respectively, and the parameter range of "Process A" overlaps with the construction window of "Process B," then it is marked as "parameter conflict"; another rule defines: process types "Process X" and "Process Y" must not be used simultaneously in adjacent units under any circumstances, and this is marked as "mutually exclusive conflict." In practice, the process suitability label "Process A - Parameter Range P3" of its own unit and the current decision labels ["Process A - Parameter Range P1", "Process B", "Process C - Parameter Range P2"] obtained from all adjacent units are input into the conflict detection rule base for matching and detection. The matching and detection process compares each rule one by one to identify all existing conflict relationships and their specific conflict types. For example, it identifies a "parameter range incompatibility" conflict with unit "R0-1-1" and a "process mutual exclusion" conflict with unit "R0-2-1".

[0029] For each identified conflict, the simulation agent "Agent_R0-1-2" invokes a preset conflict resolution strategy, which includes process parameter adjustment, construction sequence rescheduling, or process type replacement. In some embodiments, the strategy priority order is preset as follows: first attempt to try "process parameter adjustment," then attempt "construction sequence rescheduling" if unsuccessful, and finally attempt "process type replacement." The simulation agent "Agent_R0-1-2" attempts to apply each strategy sequentially according to its priority order. For example, it first attempts to adjust the parameter range in its own "Process A - Parameter Range P3" to make it compatible with "Process A - Parameter Range P1," and verifies whether the adjustment still results in mutual exclusion with "Process B." If adjusting the parameters cannot resolve the "process mutual exclusion" conflict, it then attempts to propose a construction sequence rescheduling scheme with unit "R0-2-1." It can be understood that the simulation agent "Agent_R0-1-2" will iteratively verify until it finds the first effective strategy combination that can eliminate all identified conflict relationships. Based on the found effective strategy combination, modify the process suitability label of the simulation agent "Agent_R0-1-2" itself, for example, change the label to "Process A-Parameter Range P1-adjust", thereby generating the updated process suitability label.

[0030] In practice, the simulation agent "Agent_R0-1-2" uses the updated process suitability label "Process A - Parameter Range P1 - adjust" as the new decision label. Whether the difference between the new decision label and the original decision label "Process A - Parameter Range P3" exceeds the tolerance threshold can be determined using a difference quantification function. In specific implementation, the difference quantization function is implemented as follows: The simulation agent first extracts all associated process parameters from the new and existing decision labels, and calculates the numerical difference between the new and old values ​​for each parameter. Then, it divides this difference by the range of change of the corresponding process parameter within its feasible range to normalize the difference, eliminating differences in dimensions and scales between different parameters. The normalized difference is squared and multiplied by a pre-assigned weight coefficient for that parameter, keeping the sum of all weight coefficients equal to one. Next, the weighted squares of all parameters are summed, and the square root operation is performed on the sum to generate a normalized comprehensive difference value. Finally, this comprehensive difference value is directly compared with a preset tolerance threshold. If the threshold is exceeded, the difference is considered significant, and an update command is sent to neighboring simulation agents via the interactive network. The difference quantization function calculates the normalized change of the new and old values ​​for each process parameter and synthesizes the changes of all parameters. Optionally, the preset tolerance threshold is quantized as a value Θ between 0 and 1. If the calculated comprehensive difference value of the new and old labels... If the value exceeds the tolerance threshold Θ, the simulation agent "Agent_R0-1-2" sends new decision label update instructions to the simulation agents corresponding to all its neighboring units through the interactive network, thereby driving the cascading responses and re-evaluations of the neighboring units. It can be understood that if the overall difference value... If the value is less than or equal to the tolerance threshold Θ, the simulation agent "Agent_R0-1-2" will maintain its current updated state and will not trigger further instruction propagation. The formula for calculating the comprehensive difference value of the decision label is: in: This represents the normalized overall difference value of the decision labels. Indicates the number of associated process parameters. Represents the weight coefficient of the i-th process parameter and all The sum is 1. This represents the value of the i-th parameter in the new decision label. This represents the value of the i-th parameter in the original decision label. This represents the range of variation of the i-th process parameter within its feasible range, and is used to normalize the parameter difference.

[0031] See Figure 4 This figure visualizes the simulation agent's decision label update and conflict resolution process: the horizontal axis represents the IDs of five dynamic construction decision units, the vertical axis represents the comprehensive difference value of decision labels, and the dashed line represents the preset tolerance threshold. This figure intuitively quantifies the degree of change in the decision labels of each unit and serves as the core visualization carrier for the difference value calculation → threshold judgment → instruction propagation logic. It clearly presents the quantitative control standards for the simulation agent's dynamic adjustment of construction decisions, providing data reference for adapting construction decisions under complex geological conditions.

[0032] Example 3: The process of performing process suitability labeling on dynamic construction decision units begins by parsing the correlation encoding between the attribute parameters of the 3D geological semantic volume unit and the submarine cable construction process parameters. From the encoding, key geological parameters that constrain construction decisions and their corresponding threshold ranges are extracted. The real-time values ​​of these key geological parameters within the dynamic construction decision unit are compared one by one with the extracted threshold ranges. Based on the comparison results, an initial process constraint set for the unit is generated. This set includes the types of construction processes allowed under the geological conditions of the unit and the parameter adjustment ranges corresponding to each process. Based on the spatial topology relationships already calculated for the dynamic construction decision unit, the compatibility of the initial process constraint sets between the unit and all its adjacent units is checked. If directly conflicting process type requirements are found, the process constraint sets of the relevant units are negotiated and pruned, removing the conflicting process options. The process constraint set after the adjacent unit compatibility negotiation and pruning is matched and verified with the real-time sea state data stream extracted from the real-time multi-source survey time-series data stream. Process options deemed unexecutable under the current real-time sea state conditions are eliminated, thus forming the final process suitability label for the dynamic construction decision unit. This final process suitability label is bound to the hierarchical identifier of the corresponding dynamic construction decision unit and injected into the initial state of the simulation agent corresponding to that unit.

[0033] During subsequent simulation runs, the simulation agent continuously monitors data update events from the multidimensional geological feature space throughout its lifecycle. When it detects an update in the attribute parameters of a 3D geological semantic unit associated with its own unit, it triggers a re-evaluation process for the process suitability label. The simulation agent obtains the updated attribute parameters and re-compares them with preset threshold ranges based on the latest key geological parameter values ​​to generate an updated set of process constraints. The simulation agent performs a secondary matching and verification of this updated set of process constraints with the latest sea state data stream and integrates it with its currently held decision labels to form a temporary process suitability label. The simulation agent pre-negotiates this temporary process suitability label with all adjacent units pointed to by the bound and stored list of adjacent unit identifiers. If the pre-negotiation is successful, the temporary process suitability label is confirmed as the new decision label; otherwise, the original decision label remains unchanged.

[0034] In practice, the process of labeling the suitability of a dynamic construction decision unit "R0-2-1" begins by parsing the correlation codes between the attribute parameters of the three-dimensional geological semantic units constituting the dynamic construction decision unit "R0-2-1" and the preset submarine cable construction process parameters. These correlation codes are stored in rule-based form, such as "Shear strength greater than 50 kPa allows process A" and "Gravel content less than 10% allows process B". In practice, key geological parameters that constrain construction decisions and their corresponding threshold ranges are extracted from the correlation codes. For the dynamic construction decision unit "R0-2-1", the extracted key geological parameters include the average shear strength S_threshold and the maximum gravel content G_threshold. The average shear strength value of 52 kPa and the maximum gravel content value of 8% of the three-dimensional geological semantic units within the dynamic construction decision unit "R0-2-1" are compared with the extracted threshold ranges. Based on the comparison results, an initial set of process constraints is generated for the dynamic construction decision unit "R0-2-1". The initial set of process constraints includes "allowed process A, parameter adjustment range of [50,70] kPa" and "allowed process B, parameter adjustment range of [5%,10%] gravel content".

[0035] In practice, based on the spatial topology relationships already calculated for the dynamic construction decision unit "R0-2-1", the list of adjacent units of "R0-2-1" is retrieved, for example, ["R0-1-2", "R0-2-2"]. The compatibility of the initial process constraint sets of the dynamic construction decision unit "R0-2-1" with these adjacent units is checked. If the initial process constraint set of the dynamic construction decision unit "R0-2-1" contains "Process A", while the current decision label of the adjacent unit "R0-2-2" is "Process C", and the conflict detection rule base defines "Process A" and "Process C" as mutually exclusive, then a directly conflicting process type requirement is identified. After identifying the conflict, the process constraint set of the dynamic construction decision unit "R0-2-1" is negotiated and pruned, removing the "Process A" option that causes the conflict. The process constraint set after the adjacent unit compatibility negotiation and pruning, for example, now only containing "Process B", is matched and verified with the real-time sea state data stream extracted from the real-time multi-source survey time-series data stream. The matching verification rules might include "Process B requires an ocean current speed below 1.5 knots." If the current real-time sea state data stream shows an ocean current speed of 1.8 knots, then the "Process B" option is removed. It can be understood that if the current sea state meets the requirements, the final process suitability label for the dynamic construction decision unit "R0-2-1" is formed, for example, "Process B - Parameter range [5%, 10%]". In specific implementation, the final process suitability label of the dynamic construction decision unit "R0-2-1" is bound to the hierarchical identifier of the dynamic construction decision unit "R0-2-1" and injected into the initial state of the simulation agent corresponding to the dynamic construction decision unit "R0-2-1".

[0036] In some embodiments, the simulation agent "Agent_R0-2-1" continuously monitors data update events from the multidimensional geological feature space throughout its lifecycle. These data update events are triggered by external geological models or real-time sensor data streams. When the simulation agent "Agent_R0-2-1" detects an update in the attribute parameters of a 3D geological semantic volume element associated with its own element, such as an update in the average shear strength from 52 kPa to 48 kPa, a re-evaluation process for process suitability labeling is triggered. The simulation agent "Agent_R0-2-1" obtains the updated attribute parameters and re-performs a comparison with preset threshold ranges based on the latest key geological parameter values. For example, an average shear strength of 48 kPa no longer meets the requirement of "greater than 50 kPa" for process A. Therefore, an updated set of process constraints is generated, removing the "process A" option. The updated set of process constraints might be "process B is allowed, parameter adjustment range is [5%, 10%] gravel content" and "process D is allowed, parameter adjustment range is [40, 55] kPa".

[0037] The simulation agent "Agent_R0-2-1" performs a secondary matching verification between the updated set of process constraints and the latest sea state data stream, and integrates the current decision label "Process B - Parameter Range [5%, 10%]" of the simulation agent "Agent_R0-2-1" to form a temporary process suitability label. The temporary process suitability label may be "Process D - Parameter Range [40, 48] kPa". The simulation agent "Agent_R0-2-1" pre-negotiates the temporary process suitability label with all adjacent units pointed to by the bound and stored list of adjacent unit identifiers. It can be understood that the pre-negotiation process involves communication and status confirmation with adjacent simulation agents "Agent_R0-1-2" and "Agent_R0-2-2". If the pre-negotiation is successful, the simulation agent "Agent_R0-2-1" confirms the temporary process suitability label "Process D - Parameter Range [40, 48] kPa" as the new decision label. If pre-negotiation fails, for example, if the current decision of the adjacent unit "R0-2-2" is a process mutually exclusive with "Process D", then the simulation agent "Agent_R0-2-1" will maintain its original decision label "Process B - Parameter Range [5%, 10%]". The difference between the old and new decision labels can be quantified by a comprehensive normalized difference value, and the formula for calculating the comprehensive normalized difference value M is: in: This represents the overall normalized difference value, which is a dimensionless scalar. This indicates the shear strength parameter value in the new label; This indicates the shear strength parameter value in the original label; This indicates the allowable range of variation for the shear strength parameter within the set of process constraints; This indicates the gravel content parameter value in the new label; This indicates the gravel content parameter value in the original label; This indicates the allowable range of variation for the gravel content parameter within the set of process constraints. and These are the normalized weighting coefficients for the shear strength parameter and the gravel content parameter, respectively, and satisfy the following conditions: + =1.

[0038] Example 4: Constructing a multi-dimensional geological feature space based on a multi-dimensional time-series window, the data of which originates from the processing of multi-source survey time-series data streams. The system receives multi-source survey time-series data streams from side-scan sonar, shallow seismic profilers, and multibeam echo sounders. Each data stream carries a unified timestamp and high-precision geographic coordinates. Using a fixed duration or fixed geographic displacement as a sliding window, synchronous slicing is performed on each multi-source survey time-series data stream to generate multiple time-aligned data slices. Spatial interpolation and gridding are performed on all data slices within each time-series window, uniformly resampling seafloor topography and geological attribute data from different sources and with different spatial resolutions into a regular grid with the same spatial resolution. The multiple geological attribute values ​​of each grid point, along with its geographic coordinates and timestamp information, are organized into a multi-dimensional feature vector in the multi-dimensional geological feature space. During the construction simulation, the system receives construction feedback data streams in real time from a high-precision DGPS positioning system and a buried plow attitude device, and integrates this data stream as a dynamic attribute layer into the multi-dimensional geological feature space.

[0039] The construction feedback data stream is analyzed to extract the real-time spatial coordinates of the submarine cable, the real-time attitude angle of the burying plow, the real-time tension of the submarine cable, and the real-time burial depth. A dynamic fusion influence area is established centered on the real-time spatial coordinates of the submarine cable. The attribute correction weights for all grid points in the multi-dimensional geological feature space within this influence area are calculated; these weights are inversely proportional to the Euclidean distance from the grid point to the real-time spatial coordinates of the submarine cable. Based on the real-time attitude angle, real-time tension of the submarine cable, and real-time burial depth data, the real-time disturbance estimate of the construction equipment on the seabed topography is calculated. According to the calculated attribute correction weights, the real-time disturbance estimate is allocated to each grid point within the dynamic fusion influence area. The real-time disturbance estimate allocated to each grid point is weighted and superimposed with the original attribute value of that grid point in the multi-dimensional geological feature space to update the attribute value of that grid point, thus completing the real-time fusion and correction of the geological feature space by the construction feedback data stream.

[0040] In practical implementation, a multi-dimensional geological feature space based on a multi-dimensional time-series window is constructed. The data originates from the processing of multi-source survey time-series data streams, receiving data from side-scan sonar, shallow seismic profilometers, and multibeam echo sounders. Each data stream carries a unified timestamp and high-precision geographic coordinates. For example, in the survey of a predetermined sea area, the multibeam echo sounder data stream provides water depth and seabed topography, the side-scan sonar data stream provides acoustic backscatter intensity to infer seabed type, and the shallow seismic profilometer data stream provides acoustic impedance to infer shallow surface geological structure. See Table 1 for the specific data structure of the multi-source survey time-series data streams.

[0041] Table 1: Time Series Data Flow Table of Multi-Source Survey In practice, a sliding window of fixed duration or fixed geographic displacement is used to synchronously slice each multi-source survey time-series data stream. For example, a fixed duration window of 10 seconds is used to synchronously slice the three data streams, generating multiple data slices that are strictly aligned in time. Spatial interpolation and gridding are performed on all data slices within each time-series window. The spatial interpolation uses the Kriging interpolation algorithm to uniformly resample seabed topography and geological attribute data from different sources and with different spatial resolutions into a regular grid with the same spatial resolution, such as generating a regular grid with a spatial resolution of 2 meters. The multiple geological attribute values ​​of each grid point, along with its geographic coordinates and timestamp information, are organized into a multi-dimensional feature vector in a multi-dimensional geological feature space. For example, the feature vector of a grid point location can be represented as (longitude, latitude, timestamp, water depth, backscattering intensity, acoustic impedance).

[0042] In some embodiments, during the construction simulation process, the system receives real-time construction feedback data streams from a high-precision DGPS positioning system and a burying plow attitude device. This data stream is then fused into the multi-dimensional geological feature space as a dynamic attribute layer. The construction feedback data stream is analyzed to extract the real-time spatial coordinates of the submarine cable, the real-time attitude angle of the burying plow, the real-time tension of the submarine cable, and the real-time burial depth data. For example, a data packet might contain: location (E123.500001, N12.400001), attitude angle (roll 2.5 degrees, pitch -1.2 degrees), tension 85 kN, and burial depth 1.8 meters. A dynamic fusion influence area is established centered on the real-time spatial coordinates of the submarine cable. For example, a circular area with a radius of 50 meters is set as the dynamic fusion influence area. The attribute correction weights of all grid points in the multi-dimensional geological feature space within the dynamic fusion influence area are calculated. The attribute correction weights are inversely proportional to the Euclidean distance from the grid point to the real-time spatial coordinates of the submarine cable. A specific formula for calculating the attribute correction weights is as follows: in: This represents the attribute correction weight of the grid point in the i-th row and j-th column within the area affected by dynamic fusion. Indicates the radius of the area affected by dynamic fusion. This represents the Euclidean distance from the grid point in the i-th row and j-th column to the real-time spatial coordinates of the submarine cable.

[0043] Based on real-time attitude angle, real-time cable tension, and real-time burial depth data, the real-time disturbance estimate of the seabed topography caused by the construction equipment is calculated. For example, based on the burial angle, tension, and velocity of the plough, the changes in the width and depth of the soil being plowed are estimated using a mechanical model, forming a topographic disturbance estimate δ. The calculated attribute correction weights are then used. The real-time perturbation estimate δ is assigned to each grid point within the dynamically fused influence area, with each grid point receiving a perturbation value of δ*. It's understandable that grid points closer to the construction site have higher weights. The closer the value is to 1, the larger the perturbation value assigned; the farther the grid point is, the smaller the weight and the smaller the perturbation value assigned. The real-time perturbation estimate assigned to each grid point is weighted and superimposed with the original attribute value of that grid point in the multidimensional geological feature space, thereby completing the real-time fusion and correction of the geological feature space by the construction feedback data stream.

[0044] See Figure 5 This figure visualizes the real-time fusion of construction feedback within a multi-dimensional geological feature space. The horizontal and vertical axes represent the X / Y coordinates of the seabed region, and the color bars on the right represent attribute correction weights. The stars in the figure correspond to the real-time location of the submarine cable. The figure shows the impact range of the construction feedback: centered on the real-time location of the submarine cable, the attribute correction weight is inversely proportional to the distance from the grid point to the cable, with the highest weight in the central area, gradually decreasing towards the edges, and a weight of 0 in areas exceeding the influence radius. This weight distribution is designed to transform construction feedback data, such as the burying plow posture and cable tension, into real-time disturbance estimates of the seabed topography, which are then superimposed onto the multi-dimensional geological feature space according to their weights. This figure intuitively presents the intensity and range of the impact of construction feedback on geological data, serving as the core visualization carrier of dynamic attribute layer fusion technology. It ensures the real-time nature of the geological feature space and makes the disturbance effects of construction on the geological environment more precise and controllable, providing a high-fidelity geological data foundation for subsequent dynamic construction decisions.

[0045] Example 5: The pre-negotiation process performed by the simulation agent before attempting to update the decision label has clearly defined steps. The simulation agent initiating the pre-negotiation generates a pre-negotiation request message containing its calculated temporary process suitability label and update details of the key geological parameters that triggered the label change. The simulation agent sends this pre-negotiation request message to all simulation agents corresponding to neighboring cells in its bound and stored list of neighboring cell identifiers, and starts a pre-negotiation timer after sending. Upon receiving the pre-negotiation request message, the simulation agents corresponding to the neighboring cells assess the potential impact of the temporary process suitability label in the message on their own state based on their current decision state and the internally maintained conflict detection rule base, and generate a pre-negotiation response message indicating agreement or disagreement, which is then returned to the initiator. Before the pre-negotiation timer expires, the simulation agent initiating the pre-negotiation collects all pre-negotiation response messages returned by neighboring cells. If all collected response messages are in agreement, the pre-negotiation is considered successful; if any disagreement response is received, the pre-negotiation is considered unsuccessful. The final result of the pre-negotiation will be fed back to the simulation agent that initiated the pre-negotiation, and this result will serve as the direct basis for it to confirm whether to adopt or abandon the temporary process suitability label.

[0046] In practice, before attempting to update the decision label, the simulation agent "Agent_R0-1-1" performs a pre-negotiation process. The agent generates a pre-negotiation request message containing the temporary process suitability label "Process D-parameter range [40,48] kPa" calculated by the agent, as well as details of the key geological parameter update that triggered this label change: "Shear strength parameter decreased from 52 kPa to 48 kPa." The agent then sends the generated pre-negotiation request message through the message channel of the discrete event simulation engine to all simulation agents corresponding to adjacent units in its bound and stored list of adjacent unit identifiers ["R0-1-2", "R0-2-1"], namely, agents "Agent_R0-1-2" and "Agent_R0-2-1". After sending, a pre-negotiation timer is started, with a timeout set to 5 seconds.

[0047] After receiving the pre-negotiation request message, the simulation agent "Agent_R0-1-2" corresponding to the adjacent unit evaluates the potential impact of the temporary process suitability label "Process D - Parameter Range [40,48] kPa" in the pre-negotiation request message on the state of the simulation agent "Agent_R0-1-2" itself, based on its current state "Process A - Parameter Range P1" and the conflict detection rule base maintained internally by the simulation agent "Agent_R0-1-2". In some embodiments, the conflict detection rule base includes the rule "Process A and Process D can coexist when their parameter ranges overlap". After evaluation, an agreed pre-negotiation response message is generated. After receiving the same pre-negotiation request message, the simulation agent "Agent_R0-2-1" corresponding to the adjacent unit evaluates the potential impact of the temporary process suitability label "Process D-parameter range [40,48] kPa" in the message based on the simulation agent "Agent_R0-2-1"'s current state "Process C - Parameter range P2" and the conflict detection rule base maintained internally by the simulation agent "Agent_R0-2-1". The conflict detection rule base contains the rule "Process C and Process D are mutually exclusive". After evaluation, an opposing pre-negotiation response message is generated.

[0048] Before the pre-negotiation timer expires, the simulation agent "Agent_R0-1-1" that initiates the pre-negotiation collects pre-negotiation response messages from all adjacent units. In a specific implementation, simulation agent "Agent_R0-1-1" receives an agreement message from simulation agent "Agent_R0-1-2" and an objection message from simulation agent "Agent_R0-2-1" after 4 seconds. Because of the objection message among the collected response messages, simulation agent "Agent_R0-1-1" determines that the pre-negotiation fails. It can be understood that the pre-negotiation result of "failure" will serve as the direct basis for simulation agent "Agent_R0-1-1" to confirm the adoption or abandonment of the temporary process suitability label. Based on this result, simulation agent "Agent_R0-1-1" abandons the temporary process suitability label "Process D-parameter range [40,48] kPa" and maintains the original state. In some embodiments, the time limit of the pre-negotiation timer... It is not a fixed value; its setting can be adjusted based on the estimated neighborhood complexity and network latency. One calculation method is as follows: in: Indicates the time limit of the pre-negotiation timer. This represents the basic processing time constant. This represents the additional processing time coefficient for each adjacent unit. Indicates the number of cells in the adjacent cell identifier list. Optional, It can be set to 2 seconds. The timeout can be set to 0.5 seconds per unit. If a simulation agent has 3 adjacent units, the timeout period for its pre-negotiation timer will be set accordingly. The calculation time is 3.5 seconds. If the simulation agent "Agent_R0-1-1" does not collect all responses before the pre-negotiation timer expires (e.g., only one agreeing response is received while another response is not returned after the time limit), the simulation agent "Agent_R0-1-1" determines that the pre-negotiation fails due to timeout. The final result of the pre-negotiation, whether "passed," "failed," or "failed due to timeout," will be fed back to the simulation agent that initiated the pre-negotiation. This result serves as the direct basis for its confirmation of adopting or abandoning the temporary process suitability label.

[0049] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0050] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A simulation method for submarine cable construction under complex seabed geological conditions, characterized in that, The method includes: A multidimensional geological feature space based on a multidimensional time series window is constructed. The data of the multidimensional geological feature space comes from the recombination of slices of multi-source survey time series data streams. Each time series window contains a continuous sampling sequence of multiple seabed topographic attribute parameters synchronized with geographic coordinates. For the continuous sampling sequence within each time window in the multidimensional geological feature space, feature fusion and dimension mapping are performed to generate a three-dimensional geological semantic volume unit with spatiotemporal consistency, and the attribute parameters of the three-dimensional geological semantic volume unit are correlated with the preset submarine cable construction process parameters. The complete simulation domain composed of three-dimensional geological semantic volume units is spatially segmented by a recursive segmentation algorithm to generate a series of dynamic construction decision units with hierarchical nested structures. The boundary of each dynamic construction decision unit is automatically defined by the mutation threshold of the attribute parameters of the three-dimensional geological semantic volume unit. The process suitability of each dynamic construction decision unit is marked using the correlation coding results, and decision labels containing potential construction constraints are generated. The initial state of the simulation agent corresponding to each dynamic construction decision unit is initialized based on the decision labels. The simulation agent with initial state is embedded into the discrete event simulation engine. An interaction network is constructed based on the spatiotemporal proximity between the simulation agents. The decision label update instruction is transmitted through the interaction network to drive the re-evaluation of the attribute parameters of the three-dimensional geological semantic volume unit among the simulation agents.

2. The simulation method for submarine cable construction under complex seabed geological conditions according to claim 1, characterized in that, The method involves spatially segmenting the complete simulation domain composed of three-dimensional geological semantic volume units using a recursive segmentation algorithm, generating a series of dynamic construction decision units with a hierarchical nested structure, including: The gradient change of the attribute parameters of the three-dimensional geological semantic volume unit in space is used as the trigger condition for recursive segmentation. When the gradient value of the attribute parameter exceeds the preset global mutation threshold, the initial spatial segmentation is performed at the gradient mutation location. Within each sub-region formed by the initial spatial segmentation, the local gradient distribution of the attribute parameters of the three-dimensional geological semantic volume unit is recalculated, and the sub-region is recursively subdivided according to the local mutation threshold until the gradient change of the attribute parameters within the sub-region stabilizes within the preset smoothing range. Assign a unique hierarchical identifier to each new spatial region generated by recursive segmentation, and establish a subordinate link between the hierarchical identifier and the parent region identifier to form a dynamic construction decision unit tree with a hierarchical nested structure. Based on the geometric center coordinates and hierarchical identifier of each dynamic construction decision unit, calculate its spatial topological relationship with adjacent units, and record the spatial topological relationship in the additional attributes of the dynamic construction decision unit tree; The dynamic construction decision unit tree is registered and aligned with the multidimensional geological feature space to ensure that each dynamic construction decision unit can be back-indexed to one or more three-dimensional geological semantic units that constitute it.

3. The simulation method for submarine cable construction under complex seabed geological conditions according to claim 2, characterized in that, The calculation of the spatial topological relationship between each dynamic construction decision unit and its adjacent units, based on the geometric center coordinates and hierarchical identifier of each unit, includes: Extract the geometric center coordinates of all leaf nodes in the dynamic construction decision unit tree, and construct a spatial adjacency graph of all leaf nodes on a two-dimensional plane based on the Delaunay triangulation algorithm. In the spatial adjacency diagram, the adjacency relationships of leaf nodes are aggregated layer by layer from bottom to top according to the subordinate link relationship of the hierarchical identifier, and the spatial topology relationship of each non-leaf node dynamic construction decision unit is derived at the corresponding level. The derived spatial topological relationships are encoded as adjacency matrices. The rows and columns of the matrix correspond to all units at the same level in the dynamic construction decision unit tree. The values ​​of the matrix elements indicate whether there is a shared boundary or corner contact between the corresponding two units. Based on the adjacency matrix, each dynamic construction decision unit is traversed, all its directly adjacent units are identified, and a list of adjacent unit identifiers is generated. The list of adjacent unit identifiers is used as an additional attribute of the dynamic construction decision unit tree and is bound and stored with the hierarchical identifier of the corresponding unit.

4. The simulation method for submarine cable construction under complex seabed geological conditions according to claim 3, characterized in that, The step of binding and storing the list of adjacent unit identifiers as an additional attribute of the dynamic construction decision unit tree with the hierarchical identifiers of the corresponding units includes: In response to the simulation agent receiving the decision label update instruction, the simulation agent retrieves the list of adjacent unit identifiers bound to the storage based on the hierarchical identifier of the dynamic construction decision unit to which it belongs. Based on the list of adjacent unit identifiers, the simulation agent initiates a status query request to the simulation agents corresponding to all adjacent dynamic construction decision units in order to obtain the current decision labels and process suitability labels of the adjacent units. The simulation agent integrates the process suitability label of its own unit with the current decision label obtained from neighboring units, executes conflict resolution logic based on the integration result, and generates updated process suitability labels. The updated process suitability label is used as the new decision label, and it is determined whether the difference between it and the original decision label exceeds the tolerance threshold. If it does, a new decision label update instruction is sent to the simulation agent corresponding to the adjacent unit through the interactive network.

5. The simulation method for submarine cable construction under complex seabed geological conditions according to claim 1, characterized in that, The process suitability labeling for each dynamic construction decision unit using the correlation coding results includes: The correlation encoding between the attribute parameters of three-dimensional geological semantic units and the submarine cable construction process parameters is analyzed, and the key geological parameters that constrain construction decisions and their threshold ranges are extracted. The real-time values ​​of key geological parameters are compared with the corresponding threshold ranges. Based on the comparison results, an initial set of process constraints for the dynamic construction decision unit is generated. The set of process constraints includes the allowed construction process types and the corresponding parameter adjustment ranges. Based on the spatial topology of the dynamic construction decision unit, check the compatibility of the process constraint sets between adjacent units. If there are directly conflicting process type requirements, negotiate and prune the process constraint sets to remove the process options that cause the conflict. The negotiated and pruned set of process constraints is matched and verified with the real-time sea state data stream extracted from the multi-source survey time series data stream. Process options that are not feasible under the current sea state conditions are eliminated to form the final process suitability label. The final process suitability label is bound to the hierarchical identifier of the corresponding dynamic construction decision unit and injected into the initial state of the simulation agent corresponding to the dynamic construction decision unit.

6. The simulation method for submarine cable construction under complex seabed geological conditions according to claim 5, characterized in that, The step of binding the final process suitability label with the hierarchical identifier of the corresponding dynamic construction decision unit and injecting it into the initial state of the simulation agent corresponding to the dynamic construction decision unit includes: Throughout its lifecycle, the simulation agent continuously monitors data update events from the multidimensional geological feature space. When it detects an update in the attribute parameters of a three-dimensional geological semantic unit associated with its own unit, it triggers a re-evaluation of the process suitability label. The simulation agent obtains the updated attribute parameters and re-executes the comparison with the threshold range based on the latest key geological parameter values ​​to generate an updated set of process constraints. The simulation agent performs a secondary matching and verification between the updated set of process constraints and the current sea state data stream, and integrates its current decision labels to form a temporary process suitability label. The simulation agent pre-negotiates the temporary process suitability label with the adjacent cells pointed to by the bound storage adjacent cell identifier list. If the pre-negotiation is successful, the temporary process suitability label is confirmed as the new decision label; otherwise, the original decision label remains unchanged.

7. The simulation method for submarine cable construction under complex seabed geological conditions according to claim 1, characterized in that, The construction of a multidimensional geological feature space based on a multidimensional time-series window includes: It receives multi-source time-series survey data streams from side-scan sonar, shallow seismic profiler and multibeam echo sounder, each data stream having a unified timestamp and high-precision geographic coordinates; Using a fixed duration or fixed geographic displacement as a sliding window, each multi-source survey time-series data stream is synchronously sliced ​​to generate multiple time-aligned data slices. Spatial interpolation and gridding are performed on the data slices within each time window to unify terrain and geological attribute data from different sources and at different resolutions into a grid with the same spatial resolution. The multiple geological attribute values ​​of each grid point are organized into a feature vector in a multidimensional geological feature space according to its geographical coordinates and timestamp; The system receives construction feedback data streams from a high-precision DGPS positioning system and a buried plow attitude device in real time, and uses these data streams as a dynamic attribute layer to be integrated into the corresponding spatiotemporal location in a multidimensional geological feature space in real time.

8. The simulation method for submarine cable construction under complex seabed geological conditions according to claim 7, characterized in that, The real-time reception of construction feedback data streams from the high-precision DGPS positioning system and the buried plow attitude device, and the real-time fusion of these construction feedback data streams into the corresponding spatiotemporal location in the multi-dimensional geological feature space, includes: The construction feedback data stream was analyzed to extract the real-time spatial coordinates of the submarine cable, the real-time attitude angle of the burying plow, the real-time tension of the submarine cable, and the real-time burial depth data. Centered on the real-time spatial coordinates of the submarine cable, a dynamic fusion influence area is established. The attribute correction weights of all grid points in the multi-dimensional geological feature space within the dynamic fusion influence area are calculated. The attribute correction weights are inversely proportional to the distance from the grid point to the real-time spatial coordinates of the submarine cable. Based on real-time attitude angle, real-time cable tension and real-time burial depth data, the real-time disturbance estimate of the seabed topography is calculated, and the real-time disturbance estimate is distributed to each grid point in the dynamic fusion influence area according to the attribute correction weight. The allocated real-time disturbance estimate is weighted and superimposed with the original attribute values ​​of the corresponding grid points in the multidimensional geological feature space to update the attribute values ​​of the grid points, thereby completing the real-time fusion of the construction feedback data stream.

9. The simulation method for submarine cable construction under complex seabed geological conditions according to claim 4, characterized in that, The simulation agent integrates the process suitability label of its own unit with the current decision label obtained from adjacent units, and executes conflict resolution logic based on the integration result, including: The simulation agent establishes a conflict detection rule base, which defines the possible mutual exclusion relationships, order dependencies, and parameter compatibility ranges between different process suitability labels. The process suitability label of the unit itself is matched with the current decision label obtained from all neighboring units and input into the conflict detection rule base for matching and detection to identify all existing conflict relationships and their conflict types. For each identified conflict relationship, a preset conflict resolution strategy is invoked, which includes adjusting process parameters, rearranging construction sequence, or replacing process type. Following the priority order of conflict resolution strategies, each strategy is tried in turn, and it is verified whether the identified conflict relationships are eliminated after application, until the first effective strategy combination that can eliminate all identified conflict relationships is found. Based on the effective strategy combination, modify the process suitability label of its own unit to generate an updated process suitability label.

10. The simulation method for submarine cable construction under complex geological conditions on the seabed according to claim 6, characterized in that, The simulation agent pre-negotiates the temporary process suitability labels with the adjacent cells pointed to by the bound and stored list of adjacent cell identifiers, including: The simulation agent generates a pre-negotiation request message, which contains temporary process suitability markings and details of the updated key geological parameters that led to the marking changes; The simulation agent sends a pre-negotiation request message to the simulation agents corresponding to all adjacent cells in the adjacent cell identifier list and starts the pre-negotiation timer; After receiving the pre-negotiation request message, the simulation agent corresponding to the adjacent unit evaluates the potential impact of the temporary process suitability label on it based on its current state and conflict detection rule base, and generates a pre-negotiation response message of agreement or disagreement. Before the pre-negotiation timer expires, the simulation agent that initiates the pre-negotiation collects pre-negotiation response messages from all adjacent units. If all responses are in agreement, the pre-negotiation is considered successful; if any opposition response is received, the pre-negotiation is considered unsuccessful. The pre-negotiation results are fed back to the simulation agent as the basis for its confirmation or rejection of the temporary process suitability label.

Images