An intelligent navigation method and system based on surplus water depth management

Abstract

This invention discloses an intelligent navigation method and system based on surplus water depth management. First, a map model is generated for the target sea area. During the analysis period, gridded water depth and predicted tide level data are acquired, and feasible and infeasible areas are defined. Next, the sea area is divided into multiple unit regions, and water depth and tide level feature data are extracted. The Mean Shift algorithm is used to cluster these regions, resulting in multiple region groups. Then, the draft of the target vessel is dynamically analyzed on a region-by-region basis, and the average surplus water depth of each region group is predicted and classified. Finally, in the map model, path costs are set according to the region group level, and the A* algorithm is used to plan navigation paths, generating multiple optimized paths. This effectively improves navigation safety and efficiency, achieving precise navigation risk analysis and management based on surplus water depth.

Description

Technical Field

[0001] This invention relates to the field of intelligent navigation, and more specifically, to an intelligent navigation method and system based on excess depth management. Background Technology

[0002] Ample water depth is an important analytical parameter in navigation, which is of great significance for ocean navigation. It changes with the actual draft and actual water depth. Various factors cause changes in draft, such as the ship's trim and list, sinking during navigation, fuel and living supplies consumption, changes in draft difference, differences in seawater density, and the increase in ship draft caused by waves. Factors affecting actual water depth include tides, meteorological rise and fall, and sediment deposition. The accurate analysis of ample water depth directly affects the safety of navigation planning.

[0003] Traditional margin of safety analysis is often based on static monitoring data and static sea area data, lacking dynamic analysis capabilities. It is difficult to analyze potential risks in actual navigation, and navigation planning based on margin of safety often lacks dynamic regional feature analysis, resulting in low navigation safety and efficiency, difficulty in avoiding potential risk factors, and difficulty in comprehensively improving navigation efficiency. Summary of the Invention

[0004] This invention overcomes the shortcomings of the prior art and proposes an intelligent navigation method and system based on surplus water depth management.

[0005] The first aspect of this invention provides an intelligent navigation method based on excess depth management, comprising:

[0006] S101: Generate a map model based on the target sea area. In one analysis cycle, acquire water depth data and forecast tide data based on gridded data within the target sea area and set feasible and infeasible areas.

[0007] S102: In the map model, multiple unit areas are divided based on the target sea area. For each unit area, feature data is extracted from the two dimensions of water depth and tide level. The feature data is used as sample data, and regional clustering analysis is performed based on the Mean Shift algorithm to obtain multiple regional groups.

[0008] S103: Using regional groups as the unit of analysis, dynamically analyze the draft of the target ship and predict the average free water depth in each regional group, and classify the regional groups based on the average free water depth.

[0009] S104: In the map model, the path cost is set according to the region group level, the unit region is used as the movement unit, the navigation planning is performed based on the A* algorithm, and multiple optimized paths are generated.

[0010] In this solution, S101 specifically refers to:

[0011] Based on the size of the target sea area, a two-dimensional visualization map model is constructed;

[0012] Acquire water depth data and predicted tide level data based on gridded data within the target sea area;

[0013] Dynamic water depth assessments are conducted based on water depth data and predicted tide levels, and preliminary feasible and infeasible areas are determined by combining this with the fixed draft of the target vessel.

[0014] In this solution, S102 specifically refers to:

[0015] In the map model, multiple unit areas are divided based on the target sea area to ensure that the area and outline of each unit area are consistent.

[0016] For a given area, multi-dimensional information extraction and feature vectorization are performed on water depth data and predicted tide level data to obtain water depth features and tide level features.

[0017] The water depth and tidal level characteristics of each unit area are analyzed. Using the water depth and tidal level characteristics as sample data, the Mean Shift algorithm is used to perform clustering iterations with the sample data of each unit area as the initial center point. Based on the clustering results, multiple region groups are mapped and divided.

[0018] In this scheme, the region clustering analysis includes:

[0019] For each unit region, each sample data is treated as a data point, and a bandwidth H is set, with each data point serving as the initial center point.

[0020] The neighborhood points of the initial center point are obtained based on H. In the process of determining the neighborhood points, the standard Euclidean distance is introduced to calculate the distance between data points. The average distance between the water depth feature and the tidal level feature of two data points is used as the distance between the data points.

[0021] Based on the Mean Shift algorithm, the vector from the neighborhood point to the center point is calculated, and the vector is weighted and averaged using a Gaussian kernel function to obtain the mean shift vector;

[0022] The initial center point is moved and updated according to the mean drift vector. All initial center points are iterated until the number of iterations is reached or the distance moved by the initial center point is less than the first preset distance.

[0023] After the iteration is completed, points within a second preset distance range from the initial center point are identified as a cluster, and the clustering result is obtained.

[0024] Based on the clustering results, the unit region is mapped and divided to obtain multiple region groups.

[0025] In this solution, S103 specifically refers to:

[0026] Within a region group, the estimated transit time and speed of the target vessel are obtained. Combined with the water depth data and forecast tide data of each unit region, the predicted draft and free water depth of the target vessel are dynamically calculated.

[0027] Within a region group, the excess water depth of each unit region is averaged to obtain the average excess water depth.

[0028] The level of each zone group is determined by the size of the average excess water depth.

[0029] In this solution, S104 specifically refers to:

[0030] In the map model, the starting point and ending point of the target ship are set. Based on the A* algorithm, the unit area is used as the movement unit. Based on the level of the area group, the priority of each unit area is set. The path cost of each unit area is set based on the priority, and the shortest path search is performed to obtain multiple navigation paths.

[0031] By combining infeasible areas, multiple navigation paths are optimized to avoid them, generating multiple optimized paths.

[0032] In this scheme, each regional group includes one or more unit areas. Water depth data, predicted tide data, and optimized routes are all dynamically visualized using map models.

[0033] A second aspect of the present invention also provides an intelligent navigation system based on excess water depth management. The system includes a memory and a processor. The memory includes an intelligent navigation program based on excess water depth management. When executed by the processor, the intelligent navigation program based on excess water depth management performs the following steps:

[0034] S101: Generate a map model based on the target sea area. In one analysis cycle, acquire water depth data and forecast tide data based on gridded data within the target sea area and set feasible and infeasible areas.

[0035] S102: In the map model, multiple unit areas are divided based on the target sea area. For each unit area, feature data is extracted from the two dimensions of water depth and tide level. The feature data is used as sample data, and regional clustering analysis is performed based on the Mean Shift algorithm to obtain multiple regional groups.

[0036] S103: Using regional groups as the unit of analysis, dynamically analyze the draft of the target ship and predict the average free water depth in each regional group, and classify the regional groups based on the average free water depth.

[0037] S104: In the map model, the path cost is set according to the region group level, the unit region is used as the movement unit, the navigation planning is performed based on the A* algorithm, and multiple optimized paths are generated.

[0038] A third aspect of the present invention also provides a computer-readable storage medium comprising an intelligent navigation program based on excess water depth management, wherein when the intelligent navigation program based on excess water depth management is executed by a processor, it implements the steps of the intelligent navigation method based on excess water depth management as described in any of the preceding claims.

[0039] This invention discloses an intelligent navigation method and system based on surplus water depth management. First, a map model is generated for the target sea area. During the analysis period, gridded water depth and predicted tide level data are acquired, and feasible and infeasible areas are defined. Next, the sea area is divided into multiple unit regions, and water depth and tide level feature data are extracted. The Mean Shift algorithm is used to cluster these regions, resulting in multiple region groups. Then, the draft of the target vessel is dynamically analyzed on a region-by-region basis, and the average surplus water depth of each region group is predicted and classified. Finally, in the map model, path costs are set according to the region group level, and the A* algorithm is used to plan navigation paths, generating multiple optimized paths. This effectively improves navigation safety and efficiency, achieving precise navigation risk analysis and management based on surplus water depth. Attached Figure Description

[0040] Figure 1 A flowchart of an intelligent navigation method based on excess water depth management according to the present invention is shown;

[0041] Figure 2 A block diagram of an intelligent navigation system based on excess water depth management according to the present invention is shown. Detailed Implementation

[0042] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0043] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0044] Figure 1A flowchart of an intelligent navigation method based on excess water depth management according to the present invention is shown.

[0045] like Figure 1 As shown, the first aspect of the present invention provides an intelligent navigation method based on excess water depth management, comprising:

[0046] S101: Generate a map model based on the target sea area. In one analysis cycle, acquire water depth data and forecast tide data based on gridded data within the target sea area and set feasible and infeasible areas.

[0047] S102: In the map model, multiple unit areas are divided based on the target sea area. For each unit area, feature data is extracted from the two dimensions of water depth and tide level. The feature data is used as sample data, and regional clustering analysis is performed based on the Mean Shift algorithm to obtain multiple regional groups.

[0048] S103: Using regional groups as the unit of analysis, dynamically analyze the draft of the target ship and predict the average free water depth in each regional group, and classify the regional groups based on the average free water depth.

[0049] S104: In the map model, the path cost is set according to the region group level, the unit region is used as the movement unit, the navigation planning is performed based on the A* algorithm, and multiple optimized paths are generated.

[0050] It should be noted that the map model and related hydrological data are specifically provided based on the S-100 hydrological data service platform. The water depth data and predicted tide level data correspond to S-102 water depth data and S-104 predicted tide level data, both of which are online hydrological data. S-100 is a universal hydrographic measurement data model, specifically an international hydrographic measurement standard. S-102 and S-104 are both nautical data specifications.

[0051] According to an embodiment of the present invention, step S101 specifically includes:

[0052] Based on the size of the target sea area, a two-dimensional visualization map model is constructed;

[0053] Acquire water depth data and predicted tide level data based on gridded data within the target sea area;

[0054] Dynamic water depth assessments are conducted based on water depth data and predicted tide levels, and preliminary feasible and infeasible areas are determined by combining this with the fixed draft of the target vessel.

[0055] It should be noted that the analysis period is one navigation planning cycle. The map model is used to visualize marine data and plan navigation routes. Water depth data includes actual water depth, water depth deviation, seabed type number, uncertainty range, etc., and predicted tide data includes high tide level, low tide level, tidal time, meteorological information, etc. Each grid in the map model contains corresponding water depth data and predicted tide data, and the data volume is determined based on the network resolution.

[0056] According to an embodiment of the present invention, step S102 specifically includes:

[0057] In the map model, multiple unit areas are divided based on the target sea area to ensure that the area and outline of each unit area are consistent.

[0058] For a given area, multi-dimensional information extraction and feature vectorization are performed on water depth data and predicted tide level data to obtain water depth features and tide level features.

[0059] The water depth and tidal level characteristics of each unit area are analyzed. Using the water depth and tidal level characteristics as sample data, the Mean Shift algorithm is used to perform clustering iterations with the sample data of each unit area as the initial center point. Based on the clustering results, multiple region groups are mapped and divided.

[0060] It should be noted that the unit area is generally divided into grids to optimize the path planning process.

[0061] According to an embodiment of the present invention, the region clustering analysis includes:

[0062] For each unit region, each sample data is treated as a data point, and a bandwidth H is set, with each data point serving as the initial center point.

[0063] The neighborhood points of the initial center point are obtained based on H. In the process of determining the neighborhood points, the standard Euclidean distance is introduced to calculate the distance between data points. The average distance between the water depth feature and the tidal level feature of two data points is used as the distance between the data points.

[0064] Based on the Mean Shift algorithm, the vector from the neighborhood point to the center point is calculated, and the vector is weighted and averaged using a Gaussian kernel function to obtain the mean shift vector;

[0065] The initial center point is moved and updated according to the mean drift vector. All initial center points are iterated until the number of iterations is reached or the distance moved by the initial center point is less than the first preset distance.

[0066] After the iteration is completed, points within a second preset distance range from the initial center point are identified as a cluster, and the clustering result is obtained.

[0067] Based on the clustering results, the unit region is mapped and divided to obtain multiple region groups.

[0068] It should be noted that the Mean Shift algorithm is a density clustering method. In marine analysis, due to the high density of gridded data, the significant and complex differences between data from different marine areas, and the considerable uncertainty in cluster grouping patterns, this invention employs the Mean Shift clustering algorithm to perform a certain number of iterative analyses. This enables rapid classification of water depth and tidal characteristics across multiple marine areas, providing data support for subsequent intelligent channel planning. Both the first and second preset distances are user-defined distances.

[0069] According to an embodiment of the present invention, step S103 specifically includes:

[0070] Within a region group, the estimated transit time and speed of the target vessel are obtained. Combined with the water depth data and forecast tide data of each unit region, the predicted draft and free water depth of the target vessel are dynamically calculated.

[0071] Within a region group, the excess water depth of each unit region is averaged to obtain the average excess water depth.

[0072] The level of each zone group is determined by the size of the average excess water depth.

[0073] It should be noted that the surplus water depth calculation is obtained through dynamic analysis of the target vessel's transit time, speed, water depth, and predicted tide level. The level of each regional group is determined based on the average surplus water depth; the greater the average surplus water depth, the higher the level, meaning a higher priority for the regional group and a lower corresponding path cost.

[0074] According to an embodiment of the present invention, step S104 specifically includes:

[0075] In the map model, the starting point and ending point of the target ship are set. Based on the A* algorithm, the unit area is used as the movement unit. Based on the level of the area group, the priority of each unit area is set. The path cost of each unit area is set based on the priority, and the shortest path search is performed to obtain multiple navigation paths.

[0076] By combining infeasible areas, multiple navigation paths are optimized to avoid them, generating multiple optimized paths.

[0077] It should be noted that unit areas within a region group have the same priority. Multiple optimized paths are generated through real-time dynamic analysis, providing efficient routes for tidal navigation and offering significant guidance for navigation planning in complex sea areas.

[0078] According to an embodiment of the present invention, each region group includes one or more unit regions.

[0079] According to embodiments of the present invention, water depth data, predicted tide level data, and optimized routes are all dynamically displayed through a map model.

[0080] It should be noted that marine information such as water depth data, forecast tide level data, and optimized paths are dynamically displayed in the map model through multiple layers, and the resolution of the gridded data can be dynamically adjusted based on the data transmission status.

[0081] It is worth mentioning that the analysis of excess water depth is of great significance for ocean navigation. It changes with the actual draft and actual water depth. Various factors cause changes in draft, such as the ship's trim and list, sinking during navigation, fuel and living supplies consumption, changes in draft difference, differences in seawater density, and the increase in ship draft caused by waves. Factors affecting actual water depth include tides, meteorological rises and falls, and sediment deposition. Furthermore, the accurate analysis of excess water depth directly affects the safety of navigation planning.

[0082] Traditional methods for analyzing surplus water depth often rely on static ocean data, lacking dynamic analysis capabilities. This makes it difficult to identify potential risks during actual navigation, and navigation planning based on surplus water depth often lacks dynamic regional feature analysis, resulting in lower navigation safety and efficiency, difficulty in avoiding potential risk factors, and difficulty in comprehensively improving navigation efficiency.

[0083] Based on this, the present invention divides the target sea area into regions. For each unit region, the corresponding water depth and predicted tide level characteristics are dynamically acquired. Based on the two-dimensional features, cluster analysis is performed to divide the area into multiple regional groups. Within each regional group, the navigation characteristics and sea area characteristics are similar. Furthermore, by dividing the regional groups, unit region analysis is performed, and dynamic excess water depth calculation analysis is conducted for the same regional group. Priorities are set, and dynamic route planning is performed on the unit region based on the priorities. This improves the route efficiency for tidal navigation, enhances the dynamic route planning capability for complex sea areas, effectively avoids potential risk factors, and effectively performs regionalized dynamic classification planning of the target sea area. Furthermore, based on the classification form and priority of the regional groups, hydrological data of different regions can be dynamically prioritized and displayed, improving the accuracy of navigation status assessment of the target sea area, effectively responding to sudden changes in sea conditions, and reducing the lag in decision support.

[0084] According to an embodiment of the present invention, it further includes:

[0085] Choose a real-time route from multiple optimized paths;

[0086] In the map model, the corresponding transit unit areas are filtered out based on the real-time flight routes, and the real-time area groups are obtained by marking the area groups to which the transit unit areas belong.

[0087] At a historical point in time, based on the real-time route direction and sequence, actual water depth data is acquired for the real-time area group, and the data is serialized to form a sequence data;

[0088] Multiple sequence data are obtained based on selected historical time points;

[0089] An LSTM sequence network is constructed, and multiple sequence data are imported into the sequence network as training data for sequence feature learning and model training.

[0090] During the actual navigation of the target ship, real-time water depth and real-time draft data are extracted based on the waypoints, and the excess water depth at the waypoints is calculated. Based on the waypoint sequence, the excess water depth and real-time water depth data at the waypoints are serialized to obtain the first real-time sequence and the second real-time sequence.

[0091] The second real-time sequence is imported into the sequence network for prediction to obtain the predicted sequence. The excess water depth is calculated by combining the predicted sequence with the real-time draft to obtain the predicted excess water depth sequence.

[0092] Based on the first real-time sequence and the predicted excess water depth sequence, the target vessel is subjected to real-time early warning assessment of excess water depth dual sequence, and navigation risk control is carried out for the target vessel.

[0093] It should be noted that a real-time area group generally includes multiple groups. Path points can be set as unit areas or at preset distance intervals. The first real-time sequence is the real-time surplus water depth sequence, and the second real-time sequence is the real-time draft depth sequence.

[0094] It is worth mentioning that traditional navigation technologies struggle to effectively predict and provide early warnings regarding excess water depth. Judgments are often based on human experience combined with real-time data, resulting in low early warning decision-making capabilities and making it difficult for ship platforms to adapt to integrated route management for multiple vessels. Therefore, this invention acquires draft and water depth data of the target vessel in real-time along its route (based on real-time monitoring) and generates a real-time sequence based on the route. Furthermore, it trains an LSTM-based prediction network by learning sequence features from historical draft and water depth data obtained from the set real-time route. This network then performs real-time prediction of the target vessel's excess water depth, generating a predicted sequence. Combining the real-time excess water depth sequence and the predicted sequence provides dual early warning analysis, enabling safe and effective control of the target vessel's navigation status and effectively mitigating potential navigation risks. This invention is suitable for navigation platforms performing safe navigation analysis for multiple vessels.

[0095] Among the multiple historical time points, a preset number of time points can be selected as historical time points based on a selected historical navigation period.

[0096] According to embodiments of the present invention, the region clustering analysis further includes:

[0097] The grid resolution for acquiring water depth data and predicted tide level data is used to dynamically allocate weights Q1 and Q2 based on the grid resolution.

[0098] The neighborhood points of the initial center point are obtained based on H. In the process of determining the neighborhood points, the standard Euclidean distance is introduced to calculate the distance between data points.

[0099] Calculate the distance between the water depth feature and the tidal level feature of two data points and perform a weighted average based on Q1 and Q2 to obtain the distance between the data points.

[0100] It should be noted that grid resolution reflects data accuracy. In this invention, based on the real-time data transmission status of the navigation platform, the grid resolution of the current water depth data and the predicted tide level data is obtained. Weight matching is then performed on the two types of data, and this is applied to distance calculation in the clustering process. This effectively improves the feature classification effect and clustering efficiency of the clustered unit region. The higher the grid resolution, the higher the weight Q is set.

[0101] Figure 2 A block diagram of an intelligent navigation system based on excess water depth management according to the present invention is shown.

[0102] A second aspect of the present invention also provides an intelligent navigation system 2 based on surplus water depth management. The system includes a memory 21 and a processor 22. The memory 21 includes an intelligent navigation program based on surplus water depth management. When the processor 22 executes the intelligent navigation program based on surplus water depth management, it performs the following steps:

[0103] S101: Generate a map model based on the target sea area. In one analysis cycle, acquire water depth data and forecast tide data based on gridded data within the target sea area and set feasible and infeasible areas.

[0104] S102: In the map model, multiple unit areas are divided based on the target sea area. For each unit area, feature data is extracted from the two dimensions of water depth and tide level. The feature data is used as sample data, and regional clustering analysis is performed based on the Mean Shift algorithm to obtain multiple regional groups.

[0105] S103: Using regional groups as the unit of analysis, dynamically analyze the draft of the target ship and predict the average free water depth in each regional group, and classify the regional groups based on the average free water depth.

[0106] S104: In the map model, the path cost is set according to the region group level, the unit region is used as the movement unit, the navigation planning is performed based on the A* algorithm, and multiple optimized paths are generated.

[0107] It should be noted that the map model and related hydrological data are specifically provided based on the S-100 hydrological data service platform. The water depth data and predicted tide level data correspond to S-102 water depth data and S-104 predicted tide level data, both of which are online hydrological data. S-100 is a universal hydrographic measurement data model, specifically an international hydrographic measurement standard. S-102 and S-104 are both nautical data specifications.

[0108] According to an embodiment of the present invention, step S101 specifically includes:

[0109] Based on the size of the target sea area, a two-dimensional visualization map model is constructed;

[0110] Acquire water depth data and predicted tide level data based on gridded data within the target sea area;

[0111] Dynamic water depth assessments are conducted based on water depth data and predicted tide levels, and preliminary feasible and infeasible areas are determined by combining this with the fixed draft of the target vessel.

[0112] It should be noted that the analysis period is one navigation planning cycle. The map model is used to visualize marine data and plan navigation routes. Water depth data includes actual water depth, water depth deviation, seabed type number, uncertainty range, etc., and predicted tide data includes high tide level, low tide level, tidal time, meteorological information, etc. Each grid in the map model contains corresponding water depth data and predicted tide data, and the data volume is determined based on the network resolution.

[0113] According to an embodiment of the present invention, step S102 specifically includes:

[0114] In the map model, multiple unit areas are divided based on the target sea area to ensure that the area and outline of each unit area are consistent.

[0115] For a given area, multi-dimensional information extraction and feature vectorization are performed on water depth data and predicted tide level data to obtain water depth features and tide level features.

[0116] The water depth and tidal level characteristics of each unit area are analyzed. Using the water depth and tidal level characteristics as sample data, the Mean Shift algorithm is used to perform clustering iterations with the sample data of each unit area as the initial center point. Based on the clustering results, multiple region groups are mapped and divided.

[0117] It should be noted that the unit area is generally divided into grids to optimize the path planning process.

[0118] According to an embodiment of the present invention, the region clustering analysis includes:

[0119] For each unit region, each sample data is treated as a data point, and a bandwidth H is set, with each data point serving as the initial center point.

[0120] The neighborhood points of the initial center point are obtained based on H. In the process of determining the neighborhood points, the standard Euclidean distance is introduced to calculate the distance between data points. The average distance between the water depth feature and the tidal level feature of two data points is used as the distance between the data points.

[0121] Based on the Mean Shift algorithm, the vector from the neighborhood point to the center point is calculated, and the vector is weighted and averaged using a Gaussian kernel function to obtain the mean shift vector;

[0122] The initial center point is moved and updated according to the mean drift vector. All initial center points are iterated until the number of iterations is reached or the distance moved by the initial center point is less than the first preset distance.

[0123] After the iteration is completed, points within a second preset distance range from the initial center point are identified as a cluster, and the clustering result is obtained.

[0124] Based on the clustering results, the unit region is mapped and divided to obtain multiple region groups.

[0125] It should be noted that the Mean Shift algorithm is a density clustering method. In marine analysis, due to the high density of gridded data, the significant and complex differences between data from different marine areas, and the considerable uncertainty in cluster grouping patterns, this invention employs the Mean Shift clustering algorithm to perform a certain number of iterative analyses. This enables rapid classification of water depth and tidal characteristics across multiple marine areas, providing data support for subsequent intelligent channel planning. Both the first and second preset distances are user-defined distances.

[0126] According to an embodiment of the present invention, step S103 specifically includes:

[0127] Within a region group, the estimated transit time and speed of the target vessel are obtained. Combined with the water depth data and forecast tide data of each unit region, the predicted draft and free water depth of the target vessel are dynamically calculated.

[0128] Within a region group, the excess water depth of each unit region is averaged to obtain the average excess water depth.

[0129] The level of each zone group is determined by the size of the average excess water depth.

[0130] It should be noted that the surplus water depth calculation is obtained through dynamic analysis of the target vessel's transit time, speed, water depth, and predicted tide level. The level of each regional group is determined based on the average surplus water depth; the greater the average surplus water depth, the higher the level, meaning a higher priority for the regional group and a lower corresponding path cost.

[0131] According to an embodiment of the present invention, step S104 specifically includes:

[0132] In the map model, the starting point and ending point of the target ship are set. Based on the A* algorithm, the unit area is used as the movement unit. Based on the level of the area group, the priority of each unit area is set. The path cost of each unit area is set based on the priority, and the shortest path search is performed to obtain multiple navigation paths.

[0133] By combining infeasible areas, multiple navigation paths are optimized to avoid them, generating multiple optimized paths.

[0134] It should be noted that unit areas within a region group have the same priority. Multiple optimized paths are generated through real-time dynamic analysis, providing efficient routes for tidal navigation and offering significant guidance for navigation planning in complex sea areas.

[0135] According to an embodiment of the present invention, each region group includes one or more unit regions.

[0136] According to embodiments of the present invention, water depth data, predicted tide level data, and optimized routes are all dynamically displayed through a map model.

[0137] It should be noted that marine information such as water depth data, forecast tide level data, and optimized paths are dynamically displayed in the map model through multiple layers, and the resolution of the gridded data can be dynamically adjusted based on the data transmission status.

[0138] It is worth mentioning that the analysis of excess water depth is of great significance for ocean navigation. It changes with the actual draft and actual water depth. Various factors cause changes in draft, such as the ship's trim and list, sinking during navigation, fuel and living supplies consumption, changes in draft difference, differences in seawater density, and the increase in ship draft caused by waves. Factors affecting actual water depth include tides, meteorological rises and falls, and sediment deposition. Furthermore, the accurate analysis of excess water depth directly affects the safety of navigation planning.

[0139] Traditional methods for analyzing surplus water depth often rely on static ocean data, lacking dynamic analysis capabilities. This makes it difficult to identify potential risks during actual navigation, and navigation planning based on surplus water depth often lacks dynamic regional feature analysis, resulting in lower navigation safety and efficiency, difficulty in avoiding potential risk factors, and difficulty in comprehensively improving navigation efficiency.

[0140] Based on this, the present invention divides the target sea area into regions. For each unit region, the corresponding water depth and predicted tide level characteristics are dynamically acquired. Based on the two-dimensional features, cluster analysis is performed to divide the area into multiple regional groups. Within each regional group, the navigation characteristics and sea area characteristics are similar. Furthermore, by dividing the regional groups, unit region analysis is performed, and dynamic excess water depth calculation analysis is conducted for the same regional group. Priorities are set, and dynamic route planning is performed on the unit region based on the priorities. This improves the route efficiency for tidal navigation, enhances the dynamic route planning capability for complex sea areas, effectively avoids potential risk factors, and effectively performs regionalized dynamic classification planning of the target sea area. Furthermore, based on the classification form and priority of the regional groups, hydrological data of different regions can be dynamically prioritized and displayed, improving the accuracy of navigation status assessment of the target sea area, effectively responding to sudden changes in sea conditions, and reducing the lag in decision support.

[0141] A third aspect of the present invention also provides a computer-readable storage medium comprising an intelligent navigation program based on excess water depth management, wherein when the intelligent navigation program based on excess water depth management is executed by a processor, it implements the steps of the intelligent navigation method based on excess water depth management as described in any of the preceding claims.

[0142] This invention discloses an intelligent navigation method and system based on surplus water depth management. First, a map model is generated for the target sea area. During the analysis period, gridded water depth and predicted tide level data are acquired, and feasible and infeasible areas are defined. Next, the sea area is divided into multiple unit regions, and water depth and tide level feature data are extracted. The Mean Shift algorithm is used to cluster these regions, resulting in multiple region groups. Then, the draft of the target vessel is dynamically analyzed on a region-by-region basis, and the average surplus water depth of each region group is predicted and classified. Finally, in the map model, path costs are set according to the region group level, and the A* algorithm is used to plan navigation paths, generating multiple optimized paths. This effectively improves navigation safety and efficiency, achieving precise navigation risk analysis and management based on surplus water depth.

[0143] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0144] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0145] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0146] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0147] Alternatively, if the integrated units of this invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

[0148] The above description is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed by the present invention, which should be covered by the scope of protection of the present invention.

Claims

1. An intelligent navigation method based on excess water depth management, characterized in that, include: S101: Generate a map model based on the target sea area. In one analysis cycle, acquire water depth data and forecast tide data based on gridded data within the target sea area and set feasible and infeasible areas. S102: In the map model, multiple unit areas are divided based on the target sea area. For each unit area, feature data is extracted from the two dimensions of water depth and tide level. The feature data is used as sample data, and regional clustering analysis is performed based on the Mean Shift algorithm to obtain multiple regional groups. S103: Using regional groups as the unit of analysis, dynamically analyze the draft of the target ship and predict the average free water depth in each regional group, and classify the regional groups based on the average free water depth. S104: In the map model, the path cost is set according to the region group level, the unit region is used as the movement unit, the navigation planning is performed based on the A* algorithm, and multiple optimized paths are generated. Specifically, S102 is as follows: In the map model, multiple unit areas are divided based on the target sea area to ensure that the area and outline of each unit area are consistent. For a given area, multi-dimensional information extraction and feature vectorization are performed on water depth data and predicted tide level data to obtain water depth features and tide level features. The water depth and tidal level characteristics of each unit area were analyzed. Using the water depth and tidal level characteristics as sample data, the Mean Shift algorithm was used to perform clustering iterations with the sample data of each unit area as the initial centroid. Based on the clustering results, multiple region groups were mapped and divided. The region clustering analysis includes: For each unit region, each sample data is treated as a data point, and a bandwidth H is set, with each data point serving as the initial center point. The neighborhood points of the initial center point are obtained based on H. In the process of determining the neighborhood points, the standard Euclidean distance is introduced to calculate the distance between data points. The average distance between the water depth feature and the tidal level feature of two data points is used as the distance between the data points. Based on the Mean Shift algorithm, the vector from the neighborhood point to the center point is calculated, and the vector is weighted and averaged using a Gaussian kernel function to obtain the mean shift vector; The initial center point is moved and updated according to the mean drift vector. All initial center points are iterated until the number of iterations is reached or the distance moved by the initial center point is less than the first preset distance. After the iteration is completed, points within a second preset distance range from the initial center point are identified as a cluster, and the clustering result is obtained. Based on the clustering results, the unit regions are mapped and divided to obtain multiple region groups; Specifically, S103 is as follows: Within a region group, the estimated transit time and speed of the target vessel are obtained. Combined with the water depth data and forecast tide data of each unit region, the predicted draft and free water depth of the target vessel are dynamically calculated. Within a region group, the excess water depth of each unit region is averaged to obtain the average excess water depth. The grading of each zone group is determined by the magnitude of the average excess water depth; Specifically, S104 is as follows: In the map model, the starting point and ending point of the target ship are set. Based on the A* algorithm, the unit area is used as the movement unit. Based on the level of the area group, the priority of each unit area is set. The path cost of each unit area is set based on the priority, and the shortest path search is performed to obtain multiple navigation paths. By combining infeasible areas, multiple navigation paths are optimized to avoid them, generating multiple optimized paths; This also includes: Choose a real-time route from multiple optimized paths; In the map model, the corresponding transit unit areas are filtered out based on the real-time flight routes, and the real-time area groups are obtained by marking the area groups to which the transit unit areas belong. At a historical point in time, based on the real-time route direction and sequence, actual water depth data is acquired for the real-time area group, and the data is serialized to form a sequence data; Multiple sequence data are obtained based on selected historical time points; An LSTM sequence network is constructed, and multiple sequence data are imported into the sequence network as training data for sequence feature learning and model training. During the actual navigation of the target ship, real-time water depth and real-time draft data are extracted based on the waypoints, and the excess water depth at the waypoints is calculated. Based on the waypoint sequence, the excess water depth and real-time water depth data at the waypoints are serialized to obtain the first real-time sequence and the second real-time sequence. The second real-time sequence is imported into the sequence network for prediction to obtain the predicted sequence. The excess water depth is calculated by combining the predicted sequence with the real-time draft to obtain the predicted excess water depth sequence. Based on the first real-time sequence and the predicted excess water depth sequence, the target vessel is subjected to real-time early warning assessment of excess water depth dual sequence, and navigation risk control is carried out for the target vessel.

2. The intelligent navigation method based on excess water depth management according to claim 1, characterized in that, Specifically, S101 is as follows: Based on the size of the target sea area, a two-dimensional visualized map model is constructed; Acquire water depth data and predicted tide level data based on gridded data within the target sea area; Dynamic water depth assessments are conducted based on water depth data and predicted tide levels, and preliminary feasible and infeasible areas are determined by combining this with the fixed draft of the target vessel.

3. The intelligent navigation method based on excess water depth management according to claim 1, characterized in that, Each region group comprises one or more unit regions.

4. The intelligent navigation method based on excess water depth management according to claim 1, characterized in that, Water depth data, predicted tide level data, and optimized routes are all dynamically visualized using map models.

5. An intelligent navigation system based on excess depth management, characterized in that, The system includes a memory and a processor. The memory includes an intelligent navigation program based on excess water depth management. When the processor executes the intelligent navigation program based on excess water depth management, it implements the steps of the intelligent navigation method based on excess water depth management as described in claim 1.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes an intelligent navigation program based on excess water depth management, which, when executed by a processor, implements the steps of the intelligent navigation method based on excess water depth management as described in any one of claims 1 to 4.

Images