Ship berthing aid decision-making method and device, electronic equipment and storage medium

Abstract

The invention provides a ship berthing auxiliary decision-making method and device, electronic equipment and a storage medium, and belongs to the technical field of ship berthing route control, and the method comprises the steps: obtaining a multi-modal environment and ship navigation sensing data; the method comprises the following steps: extracting carbon emission factors on the basis of a multi-modal environment and ship navigation sensing data, then constructing a carbon emission scoring model according to a nonlinear collaborative amplification effect among the carbon emission factors, and generating a low-carbon semantic map; a path evaluation function fusing a ship energy efficiency model and carbon emission-berthing time window compliance constraint is utilized, path section carbon emission and berthing time window compliance serve as a search basis, a berthing path enabling the path evaluation function value to be minimum is searched in real time, and the optimal navigational speed and steering angle are generated for each path section; therefore, ship berthing aided decision-making under low-carbon compliance guidance is realized. According to the method, comprehensive optimization of low-carbon property, compliance and performability can be realized, and the port and navigation system is assisted to be upgraded to a green and intelligent direction.

Description

Technical Field

[0001] This invention relates to the field of ship berthing route control technology, and in particular to a ship berthing auxiliary decision-making method, device, electronic device and storage medium. Background Technology

[0002] Ports, as crucial hubs and energy supply nodes for shipping, are not only centers for high-frequency operations such as berthing, unberthing, and loading / unloading, but also bear the burden of carbon emissions from fuel consumption and equipment operation during ship stays in port. Statistics show that energy consumption during port operations accounts for approximately 20%-30% of a ship's total voyage. Among these, route planning and scheduling strategies during berthing directly impact engine operating conditions, sailing time, and fuel efficiency, becoming a key aspect of carbon emission control throughout the ship's lifecycle. Currently, how to optimize ship berthing scheduling and route selection to achieve the coordinated development of low-carbon and intelligent port operations has become a core issue that urgently needs to be addressed in the construction of a green port and shipping system. Summary of the Invention

[0003] To address the above issues, this invention provides a ship berthing auxiliary decision-making method, device, electronic device, and storage medium. It can integrate multimodal perception data, dynamically construct a low-carbon risk semantic graph, and optimize the path based on a joint path evaluation function of energy efficiency and compliance. This achieves comprehensive optimization that takes into account low carbon emissions, compliance, and feasibility, and helps the port and shipping system upgrade towards a green and intelligent direction.

[0004] This invention provides a ship berthing auxiliary decision-making method, comprising: Acquire multimodal environment and ship navigation perception data, and preprocess the multimodal environment and ship navigation perception data; Based on the preprocessed multimodal environment and ship navigation perception data, carbon emission factors are extracted, and then a carbon emission scoring model is constructed based on the nonlinear synergistic amplification effect among the carbon emission factors. Based on the carbon emission scoring model, a low-carbon semantic map is generated; Based on the aforementioned low-carbon semantic graph, a path evaluation function that integrates the ship energy efficiency model and carbon emission-berthing time window compliance constraints is used. The carbon emission of the path segment and the compliance of the berthing time window are used as search criteria to search for the berthing path that minimizes the path evaluation function value in real time. The optimal speed and turning angle are generated for each path segment to achieve ship berthing auxiliary decision-making under the guidance of low-carbon compliance.

[0005] Preferably, the acquired multimodal environment and ship navigation perception data includes AIS system data, hydrological sensor system data, radar / visual traffic monitoring system data, and historical ship carbon emission data. The AIS system data includes the ship's unique ID, location, speed, heading, ship type, size, and current load factor. The hydrological sensor system data includes ocean current speed and direction, tide level, and water depth. The radar / visual traffic monitoring system data includes regional traffic density.

[0006] Preferably, the preprocessing of the multimodal environment and ship navigation perception data includes aligning the multimodal environment and ship navigation perception data with timestamps using a sliding window mechanism to obtain time-aligned data with time series characteristics; projecting the time-aligned data onto a unified port area coordinate system and rasterizing the port area to generate several raster nodes; and finally performing multimodal attribute mapping on each raster node to output a structured raster node attribute set.

[0007] Preferably, the carbon emission factors include ocean current resistance factor, tidal resistance factor, regional traffic density factor, water depth pressure factor, and historical carbon hotspot factor. The carbon emission scoring model constructed based on the nonlinear synergistic amplification effect among the carbon emission factors is as follows:

[0008] In the formula, This represents the node's overall carbon emission score, used to measure the node's carbon emission risk level during the berthing route, based on expert experience. As a high carbon weight region, It is a medium-carbon risk area. It is a preferred low-carbon area; This represents the carbon emission factor index, with a value range of [value range missing]. These represent ocean current resistance factor, tidal resistance factor, regional traffic density factor, water depth pressure factor, and historical carbon hotspot factor, respectively. Indicates the first Carbon emission factor weighting coefficient; Indicates the first Carbon emission factor carbon emission score, with a value range of: ; Indicates the relationship with the first Another carbon emission factor number that constitutes a synergistic relationship among carbon emission factors, satisfying Used to form a synergistic combination of carbon emission factors ; Indicates the first Carbon emission factor carbon emission score, with a value range of: ; This indicates the intensity of the nonlinear combined effect of the two types of carbon emission factors at the node; Indicates the first The synergistic risk amplification coefficient of the carbon emission factor synergistic combination pair is used to control... The degree of rating amplification.

[0009] Preferably, generating a low-carbon semantic map based on the carbon emission scoring model includes labeling each grid node with a carbon emission semantic tag using graph construction technology based on the carbon emission scoring model, and organizing the node position, environmental attributes and carbon emission semantic tags into a structured graph model to form a low-carbon semantic map.

[0010] Preferably, the path evaluation function is:

[0011] In the formula, This represents the overall assessment value for the entire route, used to comprehensively evaluate the total carbon emissions of the route and compliance with the docking time window. Indicates the total carbon emissions along the pathway; Represents path segment Carbon emissions; Indicates the carbon emission factor; Indicates that the ship is on the route segment Maintain speed Required propulsion power; Indicates the time taken for a path segment; Indicates the weight of compliance penalties; This represents the total time taken for the path; This indicates the upper limit of the port berthing compliance time window; This means that the total path time is equal to or greater than the path time. Exceeding the upper limit of the berthing compliance time window If the value is zero, a penalty is applied; otherwise, the value is zero.

[0012] Preferably, a regression modeling method based on gradient boosting trees is used to construct the ship energy efficiency model, which is as follows:

[0013] In the formula, Indicates the ship's propulsion power; Indicates the total number of regression subtrees; Indicates the first The learning rate of each regression subtree; Indicates the first Regression subtree function; Indicates the speed of the route segment; This indicates the ship's current load factor, with a value range of [value range missing]. ; This indicates the overall carbon emission score; Indicates regional traffic density; This represents the state interaction between the ship's current load factor and its overall carbon emissions.

[0014] Preferably, a directed graph structure for path search is constructed based on the low-carbon semantic graph. ,in For a set of nodes, The total number of nodes; Let be a set of directed edges. These represent the start node and the end node, respectively. Represents the node index; each directed edge Indicates that the ship can depart from the node Navigation to the node The path segment. Preferably, the real-time search finds the berthing path with the minimum path evaluation function value and generates the optimal speed and turning angle for each path segment to achieve ship berthing auxiliary decision-making under the guidance of low-carbon compliance. This includes using a heuristic path search algorithm to traverse and evaluate paths in the directed graph G of the port area. Each traversal selects the path with the minimum H, and the optimal speed corresponding to each path segment is generated synchronously based on the propulsion power prediction model. The turning angle between the direction vectors of adjacent path segments is calculated using the inverse cosine function for berthing execution scheduling, thereby achieving ship berthing auxiliary decision-making under the guidance of low-carbon compliance.

[0015] Preferably, the heuristic path search algorithm is the A* search algorithm.

[0016] The present invention provides a ship berthing auxiliary decision-making device, comprising a data acquisition and preprocessing module, a carbon emission scoring model construction module, a low-carbon semantic graph generation module, and an auxiliary decision-making module; The data acquisition and preprocessing module is used to acquire multimodal environment and ship navigation perception data, and to preprocess the multimodal environment and ship navigation perception data. The carbon emission scoring model construction module is used to extract carbon emission factors based on the preprocessed multimodal environment and ship navigation perception data, and then construct a carbon emission scoring model based on the nonlinear synergistic amplification effect between the carbon emission factors. The low-carbon semantic graph generation module is used to generate a low-carbon semantic graph based on the carbon emission scoring model. The auxiliary decision-making module is used to search for the berthing path that minimizes the path evaluation function value in real time, based on the low-carbon semantic graph and the path evaluation function that integrates the ship energy efficiency model and the carbon emission-berthing time window compliance constraints. It uses the carbon emission of the path segment and the compliance of the berthing time window as the search criteria, and generates the optimal speed and turning angle for each path segment to achieve auxiliary decision-making for ship berthing under the guidance of low-carbon compliance.

[0017] The present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the ship berthing auxiliary decision-making method.

[0018] The present invention provides a storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the ship berthing auxiliary decision-making method.

[0019] This invention provides a ship berthing auxiliary decision-making method, device, electronic device, and storage medium, which has the following beneficial effects: By establishing a path evaluation function that combines carbon emission costs with berthing time window compliance penalties, this approach balances energy efficiency with port scheduling window constraints, achieving a multi-objective balance optimization of "low carbon and compliance" in path selection. This differs from traditional single-dimensional decision-making models that focus on the shortest path or shortest time, enabling ship berthing auxiliary decision-making guided by low carbon compliance. Carbon emission factors are extracted from environmental and ship navigation perception data such as flow velocity and direction, tide level, water depth, and regional traffic density. These factors are then converted into carbon emission scores based on the nonlinear synergistic amplification effect, and a structured low-carbon semantic graph is constructed to effectively reflect the distribution of path carbon emission risks in different regions, providing data support for green berthing path assessment. The constructed ship energy efficiency model integrates multiple factors such as speed, load factor, carbon emission score, regional traffic density, and the load factor-carbon emission score interaction term. Based on historical operating data, it trains regression to predict power demand, dynamically outputting high-precision propulsion power and accurately capturing the nonlinear variation of propulsion power along the path segment. This serves as the basis for path carbon emission calculation, improving the scenario adaptability and accuracy of the path evaluation function. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating the ship berthing auxiliary decision-making method according to an embodiment of the present invention.

[0021] Figure 2 This is a schematic diagram of the structure of the ship berthing auxiliary decision-making device according to an embodiment of the present invention.

[0022] Figure labeling: 10, Data acquisition and preprocessing module; 20, Carbon emission scoring model construction module; 30, Low-carbon semantic graph generation module; 40, Decision support module. Detailed Implementation

[0023] The following describes specific embodiments and appendices. Figure 1 - Appendix Figure 2 The invention is described in detail so that those skilled in the art can more fully understand its purpose, features and effects.

[0024] Unless otherwise specified, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. In the event of any discrepancy between the definitions of terms in this invention and their commonly understood meaning by one of ordinary skill in the art to which this invention pertains, the definitions provided in this invention shall prevail.

[0025] The existing ship berthing auxiliary decision-making methods mainly have the following problems: (1) At the data modeling level, they mainly rely on AIS information, static hydrological data or empirical rules, lack the ability to integrate multimodal perception data, such as ocean currents, tide levels, water depth, traffic density, historical emission hotspots, etc., and have not transformed these data into semantic inputs that are meaningful for carbon emission, resulting in a single dimension of path assessment; (2) They mostly take the shortest time, shortest path or obstacle avoidance priority strategy as the goal, lack a comprehensive assessment of the actual operating status of ships and environmental carbon emission characteristics, and are difficult to meet the current carbon emission reduction requirements of green port and shipping management.

[0026] This invention addresses the technical problems of insufficient utilization of multimodal sensing data, lack of carbon emission guidance capability in berthing path planning, and insufficient consideration of berthing time window compliance by providing a ship berthing auxiliary decision-making method, device, electronic device, and storage medium.

[0027] Example 1 As a specific embodiment of the present invention, this embodiment provides a ship berthing auxiliary decision-making method, referring to... Figure 1 The specific steps are as follows: S100: Acquire multimodal environment and ship navigation perception data, and preprocess the multimodal environment and ship navigation perception data; S200. Based on the preprocessed multimodal environment and ship navigation perception data, carbon emission factors are extracted, and then a carbon emission scoring model is constructed according to the nonlinear synergistic amplification effect between the carbon emission factors. S300. Based on the carbon emission scoring model, generate a low-carbon semantic map; S400. Based on the low-carbon semantic graph, using the path evaluation function that integrates the ship energy efficiency model and carbon emission-berthing time window compliance constraints, the carbon emission of the path segment and the compliance of the berthing time window are used as the search criteria to search in real time for the berthing path that minimizes the path evaluation function value, and to generate the optimal speed and turning angle for each path segment, so as to realize ship berthing auxiliary decision-making under the guidance of low-carbon compliance.

[0028] The ship berthing auxiliary decision-making method in this embodiment establishes a path evaluation function that combines carbon emission cost items and berthing time window compliance penalty items. It takes into account both energy efficiency optimization and port scheduling window constraints, and achieves multi-objective balance optimization of "low carbon and compliance" in path selection. This is different from the traditional single-dimensional decision-making model that takes the shortest path or the shortest time as the objective, and realizes ship berthing auxiliary decision-making under the guidance of low carbon compliance.

[0029] Specifically, in S100, the acquired multimodal environment and ship navigation perception data includes AIS (Automatic Identification System) data, hydrological sensor system data, radar / visual traffic monitoring system data, and historical ship carbon emission data. The AIS system data includes the ship's unique ID, location, speed, heading, ship type, size, and current load factor; the hydrological sensor system data includes ocean current speed and direction, tide level, and water depth; and the radar / visual traffic monitoring system data includes regional traffic density. Preferably, the AIS system updates data every 5 to 10 seconds, the hydrological sensing system updates data every 30 seconds to 1 minute, and the radar / visual traffic monitoring system updates data every 5 to 10 seconds.

[0030] Optionally, the multimodal environment and ship operation perception data can be collected by accessing the AIS system, hydrological sensing system, radar / visual traffic monitoring system, and port historical database through edge computing nodes deployed in the port area. The AIS system data comes from the AIS system, the hydrological sensing system data comes from the hydrological sensing system, the radar / visual traffic monitoring system data comes from the radar / visual traffic monitoring system, and the historical ship carbon emission data comes from the port historical database.

[0031] The preprocessing of the acquired multimodal environment and ship navigation perception data includes: aligning the multimodal environment and ship navigation perception data to timestamps using a sliding window mechanism to obtain time-aligned data with time-series characteristics; projecting the time-aligned data onto a unified port area coordinate system and rasterizing the port area to generate several raster nodes; finally, performing multimodal attribute mapping on each raster node to output a structured raster node attribute set. Optionally, the spatial resolution of the raster nodes is 30m × 30m.

[0032] Specifically, in S200, based on the preprocessed multimodal environment and ship navigation perception data, carbon emission factors are extracted. This includes extracting the carbon emission factors based on the structured grid node attribute set through attribute normalization, grading, and labeling. Preferably, the carbon emission factors include ocean current resistance factors, tidal resistance factors, regional traffic density factors, water depth pressure factors, and historical carbon hotspot factors. In this embodiment, five types of factors—ocean current resistance factor, tidal resistance factor, regional traffic density factor, water depth pressure factor, and historical carbon hotspot factor—are identified as carbon emission factors.

[0033] Furthermore, the carbon emission scoring model constructed based on the nonlinear synergistic amplification effect among the carbon emission factors is as follows:

[0034] In the formula, This represents the node's overall carbon emission score, used to measure the node's carbon emission risk level during the berthing route, based on expert experience. As a high carbon weight region, It is a medium-carbon risk area. It is a preferred low-carbon area; This represents the carbon emission factor index, with a value range of [value range missing]. These represent ocean current resistance factor, tidal resistance factor, regional traffic density factor, water depth pressure factor, and historical carbon hotspot factor, respectively. Indicates the first Based on expert experience, the weighting coefficients for carbon emission factors are preset as follows: ocean current resistance factor weighting coefficient is 0.3, tidal resistance factor weighting coefficient is 0.2, regional traffic density factor weighting coefficient is 0.2, water depth pressure factor weighting coefficient is 0.1, and historical carbon hotspot factor weighting coefficient is 0.2. Indicates the first Carbon emission factor carbon emission score, with a value range of: ; Indicates the relationship with the first The number of another carbon emission factor that constitutes a synergistic relationship among carbon emission factors usually satisfies the following conditions: Used to form a synergistic combination of carbon emission factors ; Indicates the first Carbon emission factor carbon emission score, with a value range of: Specific content and same; This indicates the intensity of the nonlinear combined effect of the two types of carbon emission factors at the node; Indicates the first The synergistic risk amplification coefficient of the carbon emission factor synergistic combination pair is used to control... The degree of rating amplification is determined by expert experience.

[0035] Optionally, for the first Carbon emission factor carbon emission score The scoring rules are as follows: The ocean current drag factor reflects the drag effect of ocean currents on the ship's forward direction. It is calculated from the angle between the current direction and the ship's course and assigned a grade based on the current speed: when the angle between the current direction and the ship's course is ∈ [150°, 180°] and the current speed is ≥1.2, it is marked as a high current with a carbon emission score of 3; when the angle is ∈ [90°, 150°] and the current speed is ≥0.8, it is marked as a medium current with a carbon emission score of 2; all other cases are marked as low currents with a carbon emission score of 1. The tidal drag factor reflects the impact of the current tide level on the ship's draft and propulsion resistance. A tide level ≤ 1.8 is marked as high tide and has a carbon emission score of 3; a tide level ∈ (1.8, 2.5) is marked as mid tide and has a carbon emission score of 2; a tide level > 2.5 is marked as low tide and has a carbon emission score of 1. The regional traffic density factor assesses whether ships face frequent operational disturbances such as turning and deceleration in the region. Regional traffic density ≥ 0.8 (normalized) is marked as high density with a carbon emission score of 3; regional traffic density ∈ [0.5, 0.8) is marked as medium density with a carbon emission score of 2; and regional traffic density < 0.5 is marked as low density with a carbon emission score of 1. The water depth pressure factor reflects the impact of the current water depth on the ship's propulsion efficiency and navigation risk. Water depth ≤ 9.5 is marked as high water depth with a carbon emission score of 3; water depth ∈ (9.5, 11] is marked as medium water depth with a carbon emission score of 2; and water depth > 11 is marked as low water depth with a carbon emission score of 1. Historical carbon hotspot factor is used to identify areas with abnormally high carbon emissions during historical berthing processes. The average carbon emissions of nodes over the past 12 months are calculated. Areas with an average carbon emission value ≥ 0.25 are marked as high-emission areas with a carbon emission score of 3; areas with an average carbon emission value ∈ [0.15, 0.25) are marked as medium-emission areas with a carbon emission score of 2; and areas with an average carbon emission value < 0.15 are marked as low-emission areas with a carbon emission score of 1.

[0036] Specifically, in S300, generating a low-carbon semantic map based on the carbon emission scoring model includes labeling each grid node with a carbon emission semantic tag using graph construction technology based on the carbon emission scoring model, and organizing the node position, environmental attributes, and carbon emission semantic tags into a structured graph model to form a low-carbon semantic map.

[0037] The carbon emission semantic tags include carbon emission factor level tags and comprehensive carbon emission score level tags. The carbon emission factor level tags include high ocean currents and high density, and the comprehensive carbon emission score level tags include high carbon weight areas and low carbon preferred areas.

[0038] Specifically, in S400, a directed graph structure for path search is constructed based on the low-carbon semantic graph. ,in The node set represents all 30m × 30m grid cells in the port area. The total number of nodes; Let be a set of directed edges. These represent the start node and the end node, respectively. Represents the node index; each directed edge Indicates that the ship can depart from the node Navigation to the node The path segment has the following attribute information: Speed ​​of the route segment This indicates that the ship departs from the node. Drive to the node Speed; Overall carbon emission score , indicating a directed edge Upper Endpoint Node Carbon emissions impact of the surrounding grid environment; regional traffic density This indicates that the ship departs from the node. To node The estimated degree of traffic disruption on the path segment, after normalization, ranges from [value missing]. Path time This indicates that the ship departs from the node. Drive to the node The time consumed; Represents a node To node The length of the path segment.

[0039] Furthermore, a gradient boosting tree-based regression modeling method, such as XGBoost, is employed to construct a ship energy efficiency model for predicting ship performance on path segments. The propulsion power required to maintain the ship's speed. The ship's energy efficiency model is as follows:

[0040] In the formula, Indicates the ship's propulsion power; This represents the total number of regression subtrees, which is automatically optimized during the training process. Indicates the first The learning rates of the regression subtrees were obtained experimentally. Indicates the first A set of regression subtree functions, the structure of which is automatically generated by the CART algorithm, is used to learn nonlinear mapping relationships in a high-dimensional feature space; Indicates the speed of the route segment; This indicates the ship's current load factor, with a value range of [value range missing]. ; This indicates the overall carbon emission score; Indicates regional traffic density; This represents the state interaction term between the ship's current load rate and overall carbon emissions, describing the amplifying effect of load on propulsion energy consumption in the carbon emissions environment.

[0041] This embodiment uses route speed, ship load factor, carbon emission score, regional traffic density, and the load factor-carbon emission score interaction term as input features to construct a ship energy efficiency model for predicting the propulsion power required for a ship to maintain speed on a route segment. Compared to existing static ship energy consumption models based solely on speed and load, this embodiment's ship energy efficiency model integrates multiple factors such as speed, load factor, carbon emission score, regional traffic density, and the load factor-carbon emission score interaction term. Based on historical operational data, it trains regression to predict power demand, dynamically outputting high-precision propulsion power and accurately capturing the nonlinear variation of propulsion power on the route segment. This serves as the basis for calculating route carbon emissions, improving the scenario adaptability and accuracy of the route evaluation function.

[0042] Furthermore, by integrating the ship energy efficiency model with carbon emission-berthing time window compliance constraints, the path evaluation function is constructed:

[0043] In the formula, This represents the overall assessment value for the entire route, used to comprehensively evaluate the total carbon emissions of the route and compliance with the docking time window. Indicates the total carbon emissions along the pathway; Represents path segment Carbon emissions; The carbon emission factor is based on the IPCC (Intergovernmental Panel on Climate Change) guidelines. Different energy sources have different carbon emission factors, such as coal (0.7476), gasoline (0.5532), electricity (2.2132), and natural gas (0.4479). Indicates that the ship is on the route segment Maintain speed Required propulsion power; Indicates the time taken for a path segment; The weighting of compliance penalties is determined by expert experience. This represents the total time taken for the path; This indicates the upper limit of the port berthing compliance time window, obtained from the port berthing system. This means that the total path time is equal to or greater than the path time. Exceeding the upper limit of the berthing compliance time window If the value is zero, a penalty is applied; otherwise, the value is zero.

[0044] This embodiment integrates a ship energy efficiency model with a path evaluation function that combines carbon emissions and berthing time window compliance constraints. It integrates carbon emissions and berthing compliance time windows into a berthing path evaluation function. By converting the product of the predicted propulsion power and the sailing time of each path segment into carbon emissions through a carbon emission coefficient, the functions are summed and the result is used to measure whether the current path exceeds the maximum berthing time allowed by port scheduling. At the same time, a compliance penalty weight coefficient is introduced to construct the entire path evaluation function, thereby achieving a dual-objective fusion decision of "carbon emissions and compliance". This avoids the problem of traditional algorithms that result in low carbon emissions but violations or compliance but high emissions.

[0045] Furthermore, the real-time search identifies the berthing path with the minimum path evaluation function value and generates optimal speed and turning angle for each path segment to achieve ship berthing auxiliary decision-making under low-carbon compliance guidance. This includes employing a heuristic path search algorithm to traverse and evaluate paths in the directed graph G of the port area. Each traversal selects the path with the minimum H value, and simultaneously generates the optimal speed corresponding to each path segment based on the propulsion power prediction model (i.e., the ship energy efficiency model). The turning angle between the direction vectors of adjacent path segments is calculated using an inverse cosine function for berthing execution scheduling, thereby achieving ship berthing auxiliary decision-making under low-carbon compliance guidance. Optionally, the heuristic path search algorithm is the A* search algorithm.

[0046] The ship berthing auxiliary decision-making method in this embodiment is the first to extract carbon emission factors from environmental and ship navigation perception data such as flow velocity and direction, tide level, water depth, and regional traffic density, and convert them into carbon emission scores based on the nonlinear synergistic amplification effect between factors. It also constructs a structured low-carbon semantic map, which effectively reflects the distribution of path carbon emission risks in different regions and provides data support for green berthing path assessment.

[0047] Example 2 As another specific embodiment of the present invention, this embodiment provides a ship berthing auxiliary decision-making device, referring to... Figure 2 It includes a data acquisition and preprocessing module 10, a carbon emission scoring model construction module 20, a low-carbon semantic graph generation module 30, and a decision support module 40.

[0048] The data acquisition and preprocessing module 10 is used to acquire multimodal environment and ship navigation perception data, and to preprocess the multimodal environment and ship navigation perception data.

[0049] The carbon emission scoring model construction module 20 is used to extract carbon emission factors based on the preprocessed multimodal environment and ship navigation perception data, and then construct a carbon emission scoring model based on the nonlinear synergistic amplification effect between the carbon emission factors.

[0050] The low-carbon semantic graph generation module 30 is used to generate a low-carbon semantic graph based on the carbon emission scoring model.

[0051] The auxiliary decision-making module 40 is used to search for the berthing path that minimizes the path evaluation function value in real time based on the low-carbon semantic graph and the path evaluation function that integrates the ship energy efficiency model and the carbon emission-berthing time window compliance constraints. It also generates the optimal speed and turning angle for each path segment to achieve ship berthing auxiliary decision-making under the guidance of low-carbon compliance.

[0052] Example 3 As another specific embodiment of the present invention, this embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory. The processor executes the computer program to implement the steps of the ship berthing auxiliary decision-making method described in Embodiment 1: S100: Acquire multimodal environment and ship navigation perception data, and preprocess the multimodal environment and ship navigation perception data; S200. Based on the preprocessed multimodal environment and ship navigation perception data, carbon emission factors are extracted, and then a carbon emission scoring model is constructed according to the nonlinear synergistic amplification effect between the carbon emission factors. S300. Based on the carbon emission scoring model, generate a low-carbon semantic map; S400. Based on the low-carbon semantic graph, using the path evaluation function that integrates the ship energy efficiency model and carbon emission-berthing time window compliance constraints, the carbon emission of the path segment and the compliance of the berthing time window are used as the search criteria to search in real time for the berthing path that minimizes the path evaluation function value, and to generate the optimal speed and turning angle for each path segment, so as to realize ship berthing auxiliary decision-making under the guidance of low-carbon compliance.

[0053] Example 4 As another specific embodiment of the present invention, this embodiment provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the ship berthing auxiliary decision-making method described in the embodiment: S100: Acquire multimodal environment and ship navigation perception data, and preprocess the multimodal environment and ship navigation perception data; S200. Based on the preprocessed multimodal environment and ship navigation perception data, carbon emission factors are extracted, and then a carbon emission scoring model is constructed according to the nonlinear synergistic amplification effect between the carbon emission factors. S300. Based on the carbon emission scoring model, generate a low-carbon semantic map; S400. Based on the low-carbon semantic graph, using the path evaluation function that integrates the ship energy efficiency model and carbon emission-berthing time window compliance constraints, the carbon emission of the path segment and the compliance of the berthing time window are used as the search criteria to search in real time for the berthing path that minimizes the path evaluation function value, and to generate the optimal speed and turning angle for each path segment, so as to realize ship berthing auxiliary decision-making under the guidance of low-carbon compliance.

[0054] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any other way. Any modifications or equivalent changes made based on the technical essence of the present invention shall still fall within the scope of protection claimed by the present invention.

Claims

1. A ship berthing auxiliary decision-making method, characterized in that, The method includes: Acquire multimodal environment and ship navigation perception data, and preprocess the multimodal environment and ship navigation perception data; Based on the preprocessed multimodal environment and ship navigation perception data, carbon emission factors are extracted, and then a carbon emission scoring model is constructed based on the nonlinear synergistic amplification effect among the carbon emission factors. Based on the carbon emission scoring model, a low-carbon semantic map is generated; Based on the aforementioned low-carbon semantic graph, a path evaluation function that integrates the ship energy efficiency model and carbon emission-berthing time window compliance constraints is used. The carbon emission of the path segment and the compliance of the berthing time window are used as search criteria to search for the berthing path that minimizes the path evaluation function value in real time. The optimal speed and turning angle are generated for each path segment to achieve ship berthing auxiliary decision-making under the guidance of low-carbon compliance.

2. The ship berthing auxiliary decision-making method according to claim 1, characterized in that, The acquired multimodal environment and ship navigation perception data includes AIS system data, hydrological sensor system data, radar / visual traffic monitoring system data, and historical ship carbon emission data. The AIS system data includes the ship's unique ID, location, speed, heading, ship type, size, and current load factor. The hydrological sensor system data includes ocean current speed and direction, tide level, and water depth. The radar / visual traffic monitoring system data includes regional traffic density.

3. The ship berthing auxiliary decision-making method according to claim 2, characterized in that, Preprocessing the multimodal environment and ship navigation perception data includes aligning the multimodal environment and ship navigation perception data with timestamps using a sliding window mechanism to obtain time-aligned data with time-series characteristics; The time-aligned data is projected onto a unified port area coordinate system, and the port area is rasterized to generate several raster nodes. Finally, multimodal attribute mapping is performed on each raster node to output a structured raster node attribute set.

4. The ship berthing auxiliary decision-making method according to any one of claims 1-3, characterized in that, The carbon emission factors include ocean current resistance factor, tidal resistance factor, regional traffic density factor, water depth pressure factor, and historical carbon hotspot factor. The carbon emission scoring model constructed based on the nonlinear synergistic amplification effect among these carbon emission factors is as follows: In the formula, This represents the node's overall carbon emission score, used to measure the node's carbon emission risk level during the berthing route, based on expert experience. As a high carbon weight region, It is a medium-carbon risk area. It is a preferred low-carbon area; This represents the carbon emission factor index, with a value range of [value range missing]. These represent ocean current resistance factor, tidal resistance factor, regional traffic density factor, water depth pressure factor, and historical carbon hotspot factor, respectively. Indicates the first Carbon emission factor weighting coefficient; Indicates the first Carbon emission factor carbon emission score, with a value range of: ; Indicates the relationship with the first Another carbon emission factor number that constitutes a synergistic relationship among carbon emission factors, satisfying Used to form a synergistic combination of carbon emission factors ; Indicates the first Carbon emission factor carbon emission score, with a value range of: ; This indicates the intensity of the nonlinear combined effect of the two types of carbon emission factors at the node; Indicates the first The synergistic risk amplification coefficient of the carbon emission factor synergistic combination pair is used to control... The degree of rating amplification.

5. The ship berthing auxiliary decision-making method according to claim 3, characterized in that, The process of generating a low-carbon semantic map based on the carbon emission scoring model includes: labeling each grid node with a carbon emission semantic tag using graph construction technology based on the carbon emission scoring model; and organizing the node location, environmental attributes, and carbon emission semantic tags into a structured graph model to form a low-carbon semantic map.

6. The ship berthing auxiliary decision-making method according to claim 1, characterized in that, The path evaluation function is: In the formula, This represents the overall assessment value for the entire route, used to comprehensively evaluate the total carbon emissions of the route and compliance with the docking time window. Indicates the total carbon emissions along the pathway; Represents path segment Carbon emissions; Indicates the carbon emission factor; Indicates that the ship is on the route segment Maintain speed Required propulsion power; Indicates the time taken for a path segment; Indicates the weight of compliance penalties; This represents the total time taken for the path; This indicates the upper limit of the port berthing compliance time window; This means that the total path time is equal to or greater than the path time. Exceeding the upper limit of the berthing compliance time window If the value is zero, a penalty is applied; otherwise, the value is zero.

7. The ship berthing auxiliary decision-making method according to claim 6, characterized in that, A ship energy efficiency model is constructed using a gradient boosting tree-based regression modeling method. The ship energy efficiency model is as follows: In the formula, Indicates the ship's propulsion power; Indicates the total number of regression subtrees; Indicates the first The learning rate of each regression subtree; Indicates the first Regression subtree function; Indicates the speed of the route segment; This indicates the ship's current load factor, with a value range of [value range missing]. ; This indicates the overall carbon emission score; Indicates regional traffic density; This represents the state interaction between the ship's current load factor and its overall carbon emissions.

8. A ship berthing auxiliary decision-making device, characterized in that, The device includes a data acquisition and preprocessing module (10), a carbon emission scoring model construction module (20), a low-carbon semantic graph generation module (30), and an auxiliary decision-making module (40). The data acquisition and preprocessing module (10) is used to acquire multimodal environment and ship navigation perception data, and to preprocess the multimodal environment and ship navigation perception data. The carbon emission scoring model construction module (20) is used to extract carbon emission factors based on the preprocessed multimodal environment and ship navigation perception data, and then construct a carbon emission scoring model based on the nonlinear synergistic amplification effect between the carbon emission factors. The low-carbon semantic graph generation module (30) is used to generate a low-carbon semantic graph based on the carbon emission scoring model; The auxiliary decision-making module (40) is used to search for the berthing path that minimizes the path evaluation function value in real time based on the low-carbon semantic graph and the path evaluation function that integrates the ship energy efficiency model and the carbon emission-berthing time window compliance constraint. It also generates the optimal speed and turning angle for each path segment to achieve ship berthing auxiliary decision-making under the guidance of low-carbon compliance.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the method according to any one of claims 1-7.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-7.

Images