A method and system for planning a maritime transport route
By constructing a three-dimensional model of the sea area and the optimal route prediction model, and combining environmental and historical navigation information to correct the maritime transportation route, the problem of insufficient safety in the planning of maritime transportation routes in the existing technology is solved, and safer navigation paths and speed control is achieved .
Patent Information
- Application Number
- CN202411194383.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-29
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-08-29
AI Technical Summary
The existing maritime transportation route planning methods ignore the variability of the maritime environment, resulting in the impact of navigation safety and the lack of effective maritime transportation route planning methods and systems.
By acquiring the ship's environmental images, a three-dimensional model of the sea area is constructed, and the optimal route prediction model is trained in combination with the historical ship's navigation trajectory, and the initial path is corrected based on the path deviation score to obtain the optimal navigation speed to control the ship's navigation.
It improves the safety of ship navigation, avoids stranding or insufficient fuel caused by environmental or its own factors, and ensures that the ship can safely reach its target position.
Smart Images

Figure CN118710170B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of route planning, and in particular relates to a method and system for planning a maritime transport route. Background Art
[0002] With the rapid development of society, the transportation of materials between different regions is becoming more and more frequent. The inter-regional material transportation routes include railway transportation routes, road transportation routes, air transportation routes and sea transportation routes. Due to its high single cargo capacity and high transportation efficiency, sea transportation has gradually become a crucial mode of transportation in material transportation.
[0003] However, in maritime transportation, how to plan maritime transport routes has become a key issue. In the existing maritime transport route planning, it is usually based on detection technologies such as satellites and ship sonars, and new routes are planned based on previous routes. This planning method often ignores the changeable characteristics of the marine environment, which has a great adverse impact on the safety of maritime transport routes. Therefore, a new maritime transport route planning method and system is urgently needed to solve the defects of the existing technology. Summary of the invention
[0004] The present invention provides a method and system for planning a marine transport route, so as to solve the technical problem of how to improve the safety of ship navigation in marine transport.
[0005] In order to solve the above technical problems, an embodiment of the present invention provides a method for planning a maritime transport route, comprising:
[0006] Acquire a number of environmental images of the ship, obtain a number of first label sets and a number of first feature point sets according to a preset image recognition method, and construct a three-dimensional model of the sea area according to the first label sets and the first feature point sets;
[0007] Obtain the current position and target position of the ship, and obtain an initial path based on the three-dimensional model of the sea area and a preset path solving algorithm;
[0008] Obtaining historical ship navigation trajectories, and training a preset initial route prediction model in combination with the historical ship navigation trajectories according to a preset model training method to obtain an optimal route prediction model; inputting the current position and the target position into the optimal route prediction model to obtain a predicted path;
[0009] Based on a preset deviation algorithm, the initial path and the predicted path are combined to obtain a path deviation score, and according to the path deviation score and the three-dimensional model of the sea area, the initial path is corrected to obtain an optimal path;
[0010] Acquire environmental information and working information of the ship, and obtain the optimal sailing speed of the ship in combination with the optimal path;
[0011] The ship is controlled to navigate according to the optimal path and the optimal navigation speed.
[0012] It can be understood that, compared with the prior art, the present invention recognizes the environmental image of the ship through an image recognition method, thereby constructing a three-dimensional model of the sea area, and obtaining an initial path through the three-dimensional model of the sea area; constructing an optimal route prediction through the historical ship navigation trajectory to obtain a predicted path; correcting the initial path through the predicted path to obtain the optimal path, so that the optimal path not only takes into account the historical ship navigation information, but also takes into account the current sea area information, avoiding obstacles in the path and other factors that affect the navigation safety of the ship, thereby ensuring the safety of the ship's navigation in maritime transportation; through the environmental information and working information of the ship, combined with the optimal path, the optimal navigation speed of the ship is obtained, avoiding the occurrence of situations such as stranding or fuel shortage caused by environmental or self factors during navigation, ensuring that the ship can safely reach the target location, and improving the safety of ship navigation.
[0013] As a preferred solution, the method of acquiring a plurality of environmental images of the ship, obtaining a plurality of first label sets and a plurality of first feature point sets according to a preset image recognition method, and constructing a three-dimensional model of the sea area according to the first label sets and the first feature point sets, specifically includes:
[0014] According to the preset remote sensing image technology, environmental images of a preset range of the ship are collected to obtain several environmental images of the ship;
[0015] Performing image cleaning on each of the environment images to obtain a plurality of first environment images;
[0016] Inputting each of the first environment images into a preset labeler, processing an initial label set corresponding to each of the first environment images output by the labeler, and obtaining a first label set corresponding to each of the first environment images;
[0017] Extracting feature points from each of the first environment images to obtain an initial feature point set corresponding to each of the first environment images;
[0018] According to a preset edge feature extraction method, combining each of the first environment images, a plurality of edge features are obtained;
[0019] Merging the initial feature point set and the edge feature of each of the first environment images to obtain a first feature point set corresponding to each of the first environment images;
[0020] A three-dimensional model of the sea area is constructed according to the first feature point set and the first label set corresponding to each of the first environment images.
[0021] This preferred solution obtains a label set and a feature point set by labeling and extracting features from the environmental image, thereby being able to construct a more accurate three-dimensional model of the sea area.
[0022] As a preferred solution, the preset edge feature extraction method is combined with each of the first environment images to obtain a number of edge features, specifically including:
[0023] Constructing a filter factor according to a two-dimensional Gaussian function, and filtering each of the first environment images using Gaussian filtering in combination with the filter factor to obtain a plurality of second environment images;
[0024] Calculating the gradient magnitude and direction of each second environment image, and iterating the gradient magnitude of each second environment image to obtain initial edge feature data of each second environment image;
[0025] Threshold segmentation and data extraction are performed on the initial edge feature data of each second environment image to obtain a plurality of edge features.
[0026] This preferred solution obtains more accurate edge feature data by filtering the environmental image and iterating the gradient amplitude, and then can extract accurate edge features, ensuring that the subsequent feature point set is more accurate and the accuracy of the three-dimensional model of the sea area is guaranteed.
[0027] As a preferred solution, constructing a three-dimensional sea area model according to the first feature point set and the first label set corresponding to each of the first environment images specifically includes:
[0028] Matching the first feature point set and the first tag set corresponding to each of the first environment images to obtain some obstacle information; wherein the obstacle information includes: obstacle position information and obstacle geometry information;
[0029] Constructing a sea area obstacle location grid map based on the obstacle location information;
[0030] Constructing a sea area obstacle geometric view according to the obstacle geometric information;
[0031] A three-dimensional model of the sea area is constructed based on the sea area obstacle location grid map and the sea area obstacle geometric view.
[0032] This preferred solution obtains obstacle information by matching the feature point set and the label set of the image, thereby constructing a three-dimensional model of the sea area based on the obstacle information, so that the constructed three-dimensional model of the sea area is associated with obstacles closely related to maritime transportation safety.
[0033] As a preferred solution, the current position and target position of the ship are obtained, combined with the three-dimensional model of the sea area, and the initial path is obtained according to a preset path solving algorithm, which specifically includes:
[0034] Get the current position and target position of the ship;
[0035] According to the current position and target position of the ship, combined with the sea obstacle position grid map of the three-dimensional sea model, an incentive value function is constructed:
[0036]
[0037] Among them, τ(i,j) is the excitation value between the i-th grid and the j-th grid, d se is the Euclidean distance between the current position and the target position of the ship, d si is the Euclidean distance between the current position of the ship and the i-th grid, d ij is the Euclidean distance between the i-th grid and the j-th grid, d je is the Euclidean distance between the i-th grid and the target position of the ship;
[0038] According to the sea area obstacle position grid map of the three-dimensional sea area model, a steering function is constructed:
[0039]
[0040] Among them, μ i Indicates the steering value starting from the i-th grid, d ij is the Euclidean distance between the i-th grid and the j-th grid, The steering angle between the i-th grid and the j-th grid;
[0041] According to the geometric view of the sea obstacles in the three-dimensional model of the sea area, a collision function is constructed:
[0042]
[0043] Among them, D i,j is the collision risk from the i-th grid to the j-th grid, K is the number of obstacles between the i-th grid and the j-th grid, V k is the volume of the kth obstacle, H k is the highest height of the volume of the kth obstacle, and a is the weight value;
[0044] According to the incentive value function, the steering function and the collision function, a path search function is constructed:
[0045]
[0046] Where P is the path search function, α, β and γ are the weights of the incentive value, steering value and collision risk, τ(i,j) is the incentive value between the i-th grid and the j-th grid, μ i represents the steering value starting from the i-th grid, D i,j is the collision risk from the i-th grid to the j-th grid, n represents the number of grids in the historical path, j∈allowed i It means that the current i-th grid and j-th grid are accessible. Indicates that the i-th grid and the j-th grid are inaccessible;
[0047] According to the search function, an initial path is obtained in combination with the current position and the target position of the ship.
[0048] This preferred solution constructs an incentive value function, a steering function and a collision function through a three-dimensional model of the sea area, and then obtains a path search function. Through the path search function, an initial path based on the three-dimensional model of the sea area can be obtained, so that the obtained path fully considers multiple factors such as distance, ship steering, obstacle collision risk, etc., thereby improving the safety of the path.
[0049] As a preferred solution, the method of obtaining the historical ship navigation track, training the preset initial route prediction model in combination with the historical ship navigation track according to a preset model training method to obtain an optimal route prediction model; inputting the current position and the target position into the optimal route prediction model to obtain a predicted path, specifically includes:
[0050] Obtain the preset initial route prediction model and historical ship route trajectory;
[0051] According to the current position and the target position, a reward function and a loss function of an initial route prediction model are set;
[0052] Inputting the historical ship route trajectory into the initial route prediction model, combining the reward function and the loss function, iteratively training the initial route prediction model to obtain an optimal route prediction model;
[0053] The current position and the target position are input into the optimal route prediction model to obtain a predicted path.
[0054] This preferred solution trains the initial route prediction model through historical ship route trajectories to obtain the optimal route prediction model, and then can obtain the predicted path based on the optimal route prediction model, which provides a basis for subsequent correction of the initial path. Through a variety of path acquisition methods and corrections, the safety of the path can be guaranteed.
[0055] As a preferred solution, the preset deviation algorithm is combined with the initial path and the predicted path to obtain a path deviation score, and the initial path is corrected according to the path deviation score and combined with the three-dimensional model of the sea area to obtain the optimal path, which specifically includes:
[0056] According to the sea area obstacle position grid map of the three-dimensional sea area model, the initial path and the predicted path are divided respectively to obtain a first grid set corresponding to the initial path and a second grid set corresponding to the predicted path;
[0057] According to a preset first calculation method, the first grid set and the second grid set are combined to obtain a first path score of the initial path and a second path score of the predicted path respectively; wherein the preset first calculation method includes:
[0058]
[0059] Among them, S is the path score, x i and i are the horizontal and vertical coordinate values of the center point of the i-th grid, respectively, and n is the number of grids in the grid set;
[0060] According to the first path score and the second path score, combined with the current position and the target position of the ship, a path deviation score is calculated:
[0061]
[0062] Among them, Deviate is the path deviation score, S1 is the first path score, S2 is the second path score, x s and s are the horizontal and vertical coordinate values of the current position of the ship, respectively. e and e are the horizontal and vertical coordinate values of the ship’s target position respectively;
[0063] A preset deviation standard value is obtained, and the initial path is corrected in combination with the path deviation score and the sea area obstacle position grid map of the sea area three-dimensional model to obtain an optimal path.
[0064] This preferred solution calculates the deviation scores between different paths and then corrects the paths according to the deviation scores, thereby improving the safety of the paths and ensuring the safety of navigation of marine transport ships.
[0065] As a preferred solution, the acquiring of the environmental information and working information of the ship and combining the optimal path to obtain the optimal sailing speed of the ship specifically includes:
[0066] Obtaining environmental information and operating information of the ship, wherein the environmental information includes: water flow direction, water flow velocity, wind direction and wind speed; the operating information includes: total amount of ship fuel and ship weight;
[0067] constructing an environmental function of the ship according to the water flow direction, the water flow velocity, the wind direction and the wind speed;
[0068] Constructing a speed control function of the ship according to the total amount of fuel of the ship and the weight of the ship;
[0069] According to the optimal path, combined with the three-dimensional model of the sea area and the speed control function, the optimal sailing speed of the ship is obtained.
[0070] This preferred solution can calculate the optimal sailing speed based on the environmental information and working information of the ship in combination with the optimal path, thereby ensuring that the ship can safely reach the target location on the optimal path.
[0071] Accordingly, an embodiment of the present invention provides a marine transport route planning system, comprising: a sea area three-dimensional model building module, an initial path acquisition module, a predicted path acquisition module, an optimal path acquisition module, an optimal navigation speed acquisition module and a ship navigation module;
[0072] The three-dimensional model building module of the sea area is used to obtain a number of environmental images of the ship, obtain a number of first label sets and a number of first feature point sets according to a preset image recognition method, and build a three-dimensional model of the sea area according to the first label sets and the first feature point sets;
[0073] The initial path acquisition module is used to acquire the current position and target position of the ship, and obtain the initial path according to the preset path solving algorithm in combination with the three-dimensional model of the sea area;
[0074] The predicted path acquisition module is used to acquire the historical ship navigation track, and train the preset initial route prediction model in combination with the historical ship navigation track according to the preset model training method to obtain the optimal route prediction model; the current position and the target position are input into the optimal route prediction model to obtain the predicted path;
[0075] The optimal path acquisition module is used to obtain a path deviation score based on a preset deviation algorithm, in combination with the initial path and the predicted path, and to correct the initial path according to the path deviation score and in combination with the three-dimensional model of the sea area to obtain an optimal path;
[0076] The optimal sailing speed acquisition module is used to acquire the environmental information and working information of the ship, and obtain the optimal sailing speed of the ship in combination with the optimal path;
[0077] The ship navigation module is used to control the ship to navigate according to the optimal path and the optimal navigation speed.
[0078] As a preferred solution, the initial path acquisition module includes: an initial path acquisition unit;
[0079] The initial path acquisition unit is used to acquire the current position and target position of the ship;
[0080] According to the current position and target position of the ship, combined with the sea obstacle position grid map of the three-dimensional sea model, an incentive value function is constructed:
[0081]
[0082] Among them, τ(i,j) is the excitation value between the i-th grid and the j-th grid, d se is the Euclidean distance between the current position and the target position of the ship, d si is the Euclidean distance between the current position of the ship and the i-th grid, d ij is the Euclidean distance between the i-th grid and the j-th grid, d je is the Euclidean distance between the i-th grid and the target position of the ship;
[0083] According to the sea area obstacle position grid map of the three-dimensional sea area model, a steering function is constructed:
[0084]
[0085] Among them, μ i Indicates the steering value starting from the i-th grid, d ij is the Euclidean distance between the i-th grid and the j-th grid, The steering angle between the i-th grid and the j-th grid;
[0086] According to the geometric view of the sea obstacles in the three-dimensional model of the sea area, a collision function is constructed:
[0087]
[0088] Among them, D i,j is the collision risk from the i-th grid to the j-th grid, K is the number of obstacles between the i-th grid and the j-th grid, V k is the volume of the kth obstacle, H k is the highest height of the volume of the kth obstacle, and a is the weight value;
[0089] According to the incentive value function, the steering function and the collision function, a path search function is constructed:
[0090]
[0091] Where P is the path search function, α, β and γ are the weights of the incentive value, steering value and collision risk, τ(i,j) is the incentive value between the i-th grid and the j-th grid, μ i represents the steering value starting from the i-th grid, D i,j is the collision risk from the i-th grid to the j-th grid, n represents the number of grids in the historical path, j∈allowed i It means that the current i-th grid and j-th grid are accessible. Indicates that the i-th grid and the j-th grid are inaccessible;
[0092] According to the search function, an initial path is obtained in combination with the current position and the target position of the ship.
[0093] It can be understood that, compared with the prior art, the system recognizes the environmental image of the ship through an image recognition method, thereby constructing a three-dimensional model of the sea area, and obtaining an initial path through the three-dimensional model of the sea area; constructing an optimal route prediction through the historical ship navigation trajectory to obtain a predicted path; correcting the initial path through the predicted path to obtain the optimal path, so that the optimal path not only takes into account the historical ship navigation information, but also takes into account the current sea area information, avoiding obstacles in the path and other factors that affect the navigation safety of the ship, thereby ensuring the safety of the ship's navigation in maritime transportation; through the ship's environmental information and working information, combined with the optimal path, the optimal navigation speed of the ship is obtained, avoiding the occurrence of ship grounding or fuel shortage due to environmental or self-factors during navigation, ensuring that the ship can safely reach the target location, and improving the safety of ship navigation. BRIEF DESCRIPTION OF THE DRAWINGS
[0094] Figure 1 : A flow chart of the steps of a method for planning a maritime transport route provided by an embodiment of the present invention;
[0095] Figure 2 : A structural diagram of a marine transport route planning system provided by an embodiment of the present invention;
[0096] The reference numerals of the drawings in the specification are as follows:
[0097] 201: Sea area three-dimensional model building module; 202: Initial path acquisition module; 203: Prediction path acquisition module; 204: Optimal path acquisition module; 205: Optimal navigation speed acquisition module; 206: Ship navigation module. DETAILED DESCRIPTION
[0098] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0099] Embodiment 1
[0100] Please refer to Figure 1 , which is a flow chart of the steps of a method for planning a maritime transport route provided by an embodiment of the present invention, including steps S101-S106.
[0101] Step S101: Acquire a number of environmental images of the ship, obtain a number of first label sets and a number of first feature point sets according to a preset image recognition method, and construct a three-dimensional model of the sea area according to the first label sets and the first feature point sets.
[0102] In this embodiment, the step of acquiring a plurality of environmental images of the ship, obtaining a plurality of first tag sets and a plurality of first feature point sets according to a preset image recognition method, and constructing a three-dimensional model of the sea area according to the first tag sets and the first feature point sets specifically includes:
[0103] According to the preset remote sensing image technology, environmental images of a preset range of the ship are collected to obtain several environmental images of the ship;
[0104] Performing image cleaning on each of the environment images to obtain a plurality of first environment images;
[0105] Inputting each of the first environment images into a preset labeler, processing an initial label set corresponding to each of the first environment images output by the labeler, and obtaining a first label set corresponding to each of the first environment images;
[0106] Extracting feature points from each of the first environment images to obtain an initial feature point set corresponding to each of the first environment images;
[0107] According to a preset edge feature extraction method, combining each of the first environment images, a plurality of edge features are obtained;
[0108] Merging the initial feature point set and the edge feature of each of the first environment images to obtain a first feature point set corresponding to each of the first environment images;
[0109] A three-dimensional model of the sea area is constructed according to the first feature point set and the first label set corresponding to each of the first environment images.
[0110] It should be noted that the preset remote sensing image technology refers to a technology for collecting environmental images through remote sensing platforms and sensors; the preset range of the ship can be determined according to factors such as the actual size of the ship; and the processing of the initial label set refers to data preprocessing of the labels in the label set.
[0111] In this embodiment, the preset edge feature extraction method is combined with each of the first environment images to obtain a number of edge features, specifically including:
[0112] Constructing a filter factor according to a two-dimensional Gaussian function, and filtering each of the first environment images using Gaussian filtering in combination with the filter factor to obtain a plurality of second environment images;
[0113] Calculating the gradient magnitude and direction of each second environment image, and iterating the gradient magnitude of each second environment image to obtain initial edge feature data of each second environment image;
[0114] Threshold segmentation and data extraction are performed on the initial edge feature data of each second environment image to obtain a plurality of edge features.
[0115] In this embodiment, constructing the three-dimensional model of the sea area according to the first feature point set and the first label set corresponding to each of the first environment images specifically includes:
[0116] Matching the first feature point set and the first tag set corresponding to each of the first environment images to obtain some obstacle information; wherein the obstacle information includes: obstacle position information and obstacle geometry information;
[0117] Constructing a sea area obstacle location grid map based on the obstacle location information;
[0118] Constructing a sea area obstacle geometric view according to the obstacle geometric information;
[0119] A three-dimensional model of the sea area is constructed based on the sea area obstacle location grid map and the sea area obstacle geometric view.
[0120] In an optional embodiment, constructing a sea area obstacle location grid map according to the obstacle location information specifically includes:
[0121] The obstacle location information includes: the longitude and latitude of the obstacle;
[0122] Get the center point of each obstacle;
[0123] According to the center point of each obstacle, combined with the longitude of each obstacle and the latitude of each obstacle, a grid map of the sea area obstacle locations is constructed.
[0124] In an optional embodiment, constructing a sea area obstacle geometric view according to the obstacle geometric information specifically includes:
[0125] The obstacle geometric information includes: the volume of the obstacle and the height of the obstacle;
[0126] A sea area obstacle geometric view is constructed based on the submission of the obstacle and the height of the obstacle.
[0127] Step S102: obtaining the current position and target position of the ship, combining the three-dimensional model of the sea area, and obtaining an initial path according to a preset path solving algorithm.
[0128] In this embodiment, the current position and target position of the ship are obtained, combined with the three-dimensional model of the sea area, and the initial path is obtained according to a preset path solving algorithm, which specifically includes:
[0129] Get the current position and target position of the ship;
[0130] According to the current position and target position of the ship, combined with the sea obstacle position grid map of the three-dimensional sea model, an incentive value function is constructed:
[0131]
[0132] Among them, τ(i,j) is the excitation value between the i-th grid and the j-th grid, d se is the Euclidean distance between the current position and the target position of the ship, d si is the Euclidean distance between the current position of the ship and the i-th grid, d ij is the Euclidean distance between the i-th grid and the j-th grid, d je is the Euclidean distance between the i-th grid and the target position of the ship;
[0133] According to the sea area obstacle position grid map of the three-dimensional sea area model, a steering function is constructed:
[0134]
[0135] Among them, μ i Indicates the steering value starting from the i-th grid, d ij is the Euclidean distance between the i-th grid and the j-th grid, The steering angle between the i-th grid and the j-th grid;
[0136] According to the geometric view of the sea obstacles in the three-dimensional model of the sea area, a collision function is constructed:
[0137]
[0138] Among them, D i,j is the collision risk from the i-th grid to the j-th grid, K is the number of obstacles between the i-th grid and the j-th grid, V k is the volume of the kth obstacle, H k is the highest height of the volume of the kth obstacle, and a is the weight value;
[0139] According to the incentive value function, the steering function and the collision function, a path search function is constructed:
[0140]
[0141] Where P is the path search function, α, β and γ are the weights of the incentive value, steering value and collision risk, τ(i,j) is the incentive value between the i-th grid and the j-th grid, μ i represents the steering value starting from the i-th grid, D i,j is the collision risk from the i-th grid to the j-th grid, n represents the number of grids in the historical path, j∈allowed i It means that the current i-th grid and j-th grid are accessible. Indicates that the i-th grid and the j-th grid are inaccessible;
[0142] According to the search function, an initial path is obtained in combination with the current position and the target position of the ship.
[0143] Step S103: Obtain historical ship navigation trajectories, and train a preset initial route prediction model in combination with the historical ship navigation trajectories according to a preset model training method to obtain an optimal route prediction model; input the current position and the target position into the optimal route prediction model to obtain a predicted path.
[0144] In this embodiment, the historical ship navigation track is obtained, and a preset initial route prediction model is trained in combination with the historical ship navigation track according to a preset model training method to obtain an optimal route prediction model; the current position and the target position are input into the optimal route prediction model to obtain a predicted path, specifically including:
[0145] Obtain the preset initial route prediction model and historical ship route trajectory;
[0146] According to the current position and the target position, a reward function and a loss function of an initial route prediction model are set;
[0147] Inputting the historical ship route trajectory into the initial route prediction model, combining the reward function and the loss function, iteratively training the initial route prediction model to obtain an optimal route prediction model;
[0148] The current position and the target position are input into the optimal route prediction model to obtain a predicted path.
[0149] Step S104: Based on a preset deviation algorithm, the initial path and the predicted path are combined to obtain a path deviation score. According to the path deviation score and combined with the three-dimensional model of the sea area, the initial path is corrected to obtain an optimal path.
[0150] In this embodiment, the preset deviation algorithm is combined with the initial path and the predicted path to obtain a path deviation score. According to the path deviation score, combined with the three-dimensional model of the sea area, the initial path is corrected to obtain the optimal path, which specifically includes:
[0151] According to the sea area obstacle position grid map of the three-dimensional sea area model, the initial path and the predicted path are divided respectively to obtain a first grid set corresponding to the initial path and a second grid set corresponding to the predicted path;
[0152] According to a preset first calculation method, the first grid set and the second grid set are combined to obtain a first path score of the initial path and a second path score of the predicted path respectively; wherein the preset first calculation method includes:
[0153]
[0154] Among them, S is the path score, x i and i are the horizontal and vertical coordinate values of the center point of the i-th grid, respectively, and n is the number of grids in the grid set;
[0155] According to the first path score and the second path score, combined with the current position and the target position of the ship, a path deviation score is calculated:
[0156]
[0157] Where Deviate is the path deviation score, S1 is the first path score, S2 is the second path score, x s and s are the horizontal and vertical coordinate values of the current position of the ship, respectively. e and e are the horizontal and vertical coordinate values of the ship’s target position respectively;
[0158] A preset deviation standard value is obtained, and the initial path is corrected in combination with the path deviation score and the sea area obstacle position grid map of the sea area three-dimensional model to obtain an optimal path.
[0159] Step S105: Acquire the environmental information and working information of the ship, and combine the optimal path to obtain the optimal sailing speed of the ship.
[0160] In this embodiment, the acquiring of the environmental information and the working information of the ship and combining the optimal path to obtain the optimal sailing speed of the ship specifically include:
[0161] Acquiring environmental information and operating information of the ship, wherein the environmental information includes: water flow direction and water flow velocity; the operating information includes: total amount of ship fuel and ship weight;
[0162] Constructing an environmental function of the ship according to the flow direction of the water body and the flow velocity of the water body;
[0163] According to the total amount of fuel of the ship and the weight of the ship, combined with the environmental function, construct a speed control function of the ship;
[0164] According to the optimal path, combined with the three-dimensional model of the sea area and the speed control function, the optimal sailing speed of the ship is obtained.
[0165] In an optional embodiment, constructing an environmental function of the ship according to the water flow direction and the water flow velocity; constructing a speed control function of the ship according to the total amount of fuel of the ship and the weight of the ship in combination with the environmental function, specifically includes:
[0166] According to the fluctuation information of water body, construct the water flow direction function:
[0167]
[0168] B(t)=B0+A×cos(σt+θ),
[0169] Among them, φ(x,y,t) is the flow direction of the water body, x is the value of the wave function on the x-axis, y is the value of the wave function on the y-axis, t is the current time, c is the phase velocity of the wave function; B0 is the pre-value of the wave amplitude, B(t) is the amplitude function, A is the amplitude, θ is the phase angle, and σ is the wave number;
[0170] According to the water flow direction function, the water flow velocity function is obtained:
[0171]
[0172] Among them, X(x,y,t) is the velocity component of the water body on the X axis at time t, and Y(x,y,t) is the velocity component of the water body on the Y axis at time t;
[0173] According to the water flow direction function and the water flow velocity function, the ship environment function is constructed:
[0174]
[0175] Where E is the environmental function of the ship, X(x,y,t) is the velocity component of the water body on the X-axis at time t, Y(x,y,t) is the velocity component of the water body on the Y-axis at time t, and φ(x,y,t) is the flow direction of the water body;
[0176] According to the total amount of fuel of the ship and the weight of the ship, combined with the environmental function, a speed control function of the ship is constructed:
[0177]
[0178] Where V is the speed control function of the ship, C d is the air resistance coefficient, A is the windward area of the ship, E is the environmental function of the ship, G is the acceleration of gravity, m is the weight of the ship, δ is the turning coefficient of the ship, M is the total amount of ship fuel, and F is the fuel efficiency of the ship fuel.
[0179] Step S106: Control the ship to sail according to the optimal path and the optimal sailing speed.
[0180] This embodiment recognizes the environmental image of the ship through an image recognition method, thereby constructing a three-dimensional model of the sea area, and obtaining an initial path through the three-dimensional model of the sea area; constructs an optimal route prediction through the historical ship navigation trajectory to obtain a predicted path; corrects the initial path through the predicted path to obtain an optimal path, so that the optimal path not only takes into account the historical ship navigation information, but also takes into account the current sea area information, avoids obstacles in the path and other factors that affect the navigation safety of the ship, thereby ensuring the safety of the ship's navigation in maritime transportation; obtains the optimal navigation speed of the ship through the environmental information and working information of the ship, combined with the optimal path, avoids the occurrence of situations such as stranding or fuel shortage caused by environmental or self factors during navigation, ensures that the ship can safely reach the target location, and improves the safety of ship navigation.
[0181] Embodiment 2
[0182] Please refer to Figure 2 , which is a structural schematic diagram of a maritime transport route planning system provided by an embodiment of the present invention, including: a sea area three-dimensional model construction module 201, an initial path acquisition module 202, a predicted path acquisition module 203, an optimal path acquisition module 204, an optimal navigation speed acquisition module 205 and a ship navigation module 206.
[0183] Among them, the sea area three-dimensional model construction module 201 is used to obtain several environmental images of the ship, obtain several first label sets and several first feature point sets according to a preset image recognition method, and construct a three-dimensional model of the sea area based on the first label sets and the first feature point sets.
[0184] In this embodiment, the sea area three-dimensional model construction module 201 includes: a sea area three-dimensional model construction unit;
[0185] The sea area three-dimensional model building unit is used to collect environmental images of a preset range of the ship according to a preset remote sensing image technology to obtain a plurality of environmental images of the ship;
[0186] Performing image cleaning on each of the environment images to obtain a plurality of first environment images;
[0187] Inputting each of the first environment images into a preset labeler, processing an initial label set corresponding to each of the first environment images output by the labeler, and obtaining a first label set corresponding to each of the first environment images;
[0188] Extracting feature points from each of the first environment images to obtain an initial feature point set corresponding to each of the first environment images;
[0189] According to a preset edge feature extraction method, combining each of the first environment images, a plurality of edge features are obtained;
[0190] Merging the initial feature point set and the edge feature of each of the first environment images to obtain a first feature point set corresponding to each of the first environment images;
[0191] A three-dimensional model of the sea area is constructed according to the first feature point set and the first label set corresponding to each of the first environment images.
[0192] In this embodiment, the sea area three-dimensional model construction unit includes: an edge feature extraction subunit;
[0193] The edge feature extraction subunit is used to construct a filter factor according to a two-dimensional Gaussian function, and in combination with the filter factor, use Gaussian filtering to filter each of the first environment images to obtain a plurality of second environment images;
[0194] Calculating the gradient magnitude and direction of each second environment image, and iterating the gradient magnitude of each second environment image to obtain initial edge feature data of each second environment image;
[0195] Threshold segmentation and data extraction are performed on the initial edge feature data of each second environment image to obtain a plurality of edge features.
[0196] In this embodiment, the sea area three-dimensional model construction unit includes: a sea area three-dimensional model construction sub-unit;
[0197] The sea area three-dimensional model construction subunit is used to match the first feature point set and the first label set corresponding to each of the first environment images to obtain a number of obstacle information; wherein the obstacle information includes: obstacle position information and obstacle geometry information;
[0198] Constructing a sea area obstacle location grid map based on the obstacle location information;
[0199] Constructing a sea area obstacle geometric view according to the obstacle geometric information;
[0200] A three-dimensional model of the sea area is constructed based on the sea area obstacle location grid map and the sea area obstacle geometric view.
[0201] The initial path acquisition module 202 is used to acquire the current position and target position of the ship, and obtain the initial path according to a preset path solving algorithm in combination with the three-dimensional model of the sea area.
[0202] In this embodiment, the initial path acquisition module 202 includes: an initial path acquisition unit;
[0203] The initial path acquisition unit is used to acquire the current position and target position of the ship;
[0204] According to the current position and target position of the ship, combined with the sea obstacle position grid map of the three-dimensional sea model, an incentive value function is constructed:
[0205]
[0206] Among them, τ(i,j) is the excitation value between the i-th grid and the j-th grid, d se is the Euclidean distance between the current position and the target position of the ship, d si is the Euclidean distance between the current position of the ship and the i-th grid, d ij is the Euclidean distance between the i-th grid and the j-th grid, d je is the Euclidean distance between the i-th grid and the target position of the ship;
[0207] According to the sea area obstacle position grid map of the three-dimensional sea area model, a steering function is constructed:
[0208]
[0209] Among them, μ i Indicates the steering value starting from the i-th grid, d ijis the Euclidean distance between the i-th grid and the j-th grid, The steering angle between the i-th grid and the j-th grid;
[0210] According to the geometric view of the sea obstacles in the three-dimensional model of the sea area, a collision function is constructed:
[0211]
[0212] Among them, D i,j is the collision risk from the i-th grid to the j-th grid, K is the number of obstacles between the i-th grid and the j-th grid, V k is the volume of the kth obstacle, H k is the highest height of the volume of the kth obstacle, and a is the weight value;
[0213] According to the incentive value function, the steering function and the collision function, a path search function is constructed:
[0214]
[0215] Where P is the path search function, α, β and γ are the weights of the incentive value, steering value and collision risk, τ(i,j) is the incentive value between the i-th grid and the j-th grid, μ i represents the steering value starting from the i-th grid, D i,j is the collision risk from the i-th grid to the j-th grid, n represents the number of grids in the historical path, j∈allowed i It means that the current i-th grid and j-th grid are accessible. Indicates that the i-th grid and the j-th grid are inaccessible;
[0216] According to the search function, an initial path is obtained in combination with the current position and the target position of the ship.
[0217] The predicted path acquisition module 203 is used to obtain the historical ship navigation trajectory, and train the preset initial route prediction model in combination with the historical ship navigation trajectory according to the preset model training method to obtain the optimal route prediction model; the current position and the target position are input into the optimal route prediction model to obtain the predicted path.
[0218] In this embodiment, the predicted path acquisition module 203 includes: a predicted path acquisition unit;
[0219] The predicted path acquisition unit is used to acquire a preset initial route prediction model and a historical ship route trajectory;
[0220] According to the current position and the target position, a reward function and a loss function of an initial route prediction model are set;
[0221] Inputting the historical ship route trajectory into the initial route prediction model, combining the reward function and the loss function, iteratively training the initial route prediction model to obtain an optimal route prediction model;
[0222] The current position and the target position are input into the optimal route prediction model to obtain a predicted path.
[0223] The optimal path acquisition module 204 is used to obtain a path deviation score based on a preset deviation algorithm in combination with the initial path and the predicted path, and to correct the initial path according to the path deviation score in combination with the three-dimensional sea area model to obtain an optimal path.
[0224] In this embodiment, the optimal path acquisition module 204 includes: an optimal path acquisition unit;
[0225] The optimal path acquisition unit is used to divide the initial path and the predicted path according to the sea area obstacle position grid map of the three-dimensional sea area model, and obtain a first grid set corresponding to the initial path and a second grid set corresponding to the predicted path;
[0226] According to a preset first calculation method, the first grid set and the second grid set are combined to obtain a first path score of the initial path and a second path score of the predicted path respectively; wherein the preset first calculation method includes:
[0227]
[0228] Among them, S is the path score, x i and i are the horizontal and vertical coordinate values of the center point of the i-th grid, respectively, and n is the number of grids in the grid set;
[0229] According to the first path score and the second path score, combined with the current position and the target position of the ship, a path deviation score is calculated:
[0230]
[0231] Among them, Deviate is the path deviation score, S1 is the first path score, S2 is the second path score, x s and s are the horizontal and vertical coordinate values of the current position of the ship, respectively. e and e are the horizontal and vertical coordinate values of the ship’s target position respectively;
[0232] A preset deviation standard value is obtained, and the initial path is corrected in combination with the path deviation score and the sea area obstacle position grid map of the sea area three-dimensional model to obtain an optimal path.
[0233] The optimal sailing speed acquisition module 205 is used to acquire the environmental information and working information of the ship, and obtain the optimal sailing speed of the ship in combination with the optimal path.
[0234] In this embodiment, the optimal sailing speed acquisition module 205 includes: an optimal sailing speed acquisition unit;
[0235] The optimal navigation speed acquisition unit is used to acquire environmental information and working information of the ship, wherein the environmental information includes: water flow direction and water flow velocity; the working information includes: total amount of ship fuel and ship weight;
[0236] Constructing an environmental function of the ship according to the flow direction of the water body and the flow velocity of the water body;
[0237] According to the total amount of fuel of the ship and the weight of the ship, combined with the environmental function, construct a speed control function of the ship;
[0238] According to the optimal path, combined with the three-dimensional model of the sea area and the speed control function, the optimal sailing speed of the ship is obtained.
[0239] The ship navigation module 206 is used to control the ship to navigate according to the optimal path and the optimal navigation speed.
[0240] This system recognizes the environmental image of the ship through an image recognition method, thereby constructing a three-dimensional model of the sea area, and obtains an initial path through the three-dimensional model of the sea area; constructs an optimal route prediction through the historical ship navigation trajectory to obtain a predicted path; corrects the initial path through the predicted path to obtain the optimal path, so that the optimal path not only takes into account the historical ship navigation information, but also takes into account the current sea area information, avoids obstacles in the path and other factors that affect the navigation safety of the ship, thereby ensuring the safety of the ship's navigation in maritime transportation; obtains the optimal navigation speed of the ship through the environmental information and working information of the ship, combined with the optimal path, avoids the occurrence of ship grounding or fuel shortage due to environmental or self-factors during navigation, ensures that the ship can safely reach the target location, and improves the safety of ship navigation.
[0241] In summary, the embodiment of the present invention recognizes the environmental image of the ship through an image recognition method, thereby constructing a three-dimensional model of the sea area, and obtaining an initial path through the three-dimensional model of the sea area; constructing an optimal route prediction through historical ship navigation trajectories to obtain a predicted path; correcting the initial path through the predicted path to obtain an optimal path, so that the optimal path not only takes into account historical ship navigation information, but also takes into account current sea area information, avoiding factors that affect the navigation safety of the ship such as obstacles in the path, thereby ensuring the safety of the navigation of the ship in maritime transportation; through the environmental information and working information of the ship, combined with the optimal path, the optimal navigation speed of the ship is obtained, avoiding the occurrence of situations such as stranding or fuel shortage caused by environmental or self factors during navigation, ensuring that the ship can safely reach the target location, and improving the safety of the ship's navigation.
[0242] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.
Claims
1. A method for planning a maritime transport route, characterized in that: include: Acquire a number of environmental images of the ship, obtain a number of first label sets and a number of first feature point sets according to a preset image recognition method, and construct a three-dimensional model of the sea area according to the first label sets and the first feature point sets; Obtain the current position and target position of the ship, and obtain an initial path based on the three-dimensional model of the sea area and a preset path solving algorithm; Obtaining historical ship navigation trajectories, and training a preset initial route prediction model in combination with the historical ship navigation trajectories according to a preset model training method to obtain an optimal route prediction model; inputting the current position and the target position into the optimal route prediction model to obtain a predicted path; Based on a preset deviation algorithm, the initial path and the predicted path are combined to obtain a path deviation score, and according to the path deviation score and the three-dimensional model of the sea area, the initial path is corrected to obtain an optimal path; Acquire environmental information and working information of the ship, and obtain the optimal sailing speed of the ship in combination with the optimal path; The ship is controlled to navigate according to the optimal path and the optimal navigation speed.
2. A method for planning a maritime transport route according to claim 1, characterized in that: The method of acquiring a plurality of environmental images of the ship, obtaining a plurality of first label sets and a plurality of first feature point sets according to a preset image recognition method, and constructing a three-dimensional model of the sea area according to the first label sets and the first feature point sets, specifically includes: According to the preset remote sensing image technology, environmental images of a preset range of the ship are collected to obtain several environmental images of the ship; Performing image cleaning on each of the environment images to obtain a plurality of first environment images; Inputting each of the first environment images into a preset labeler, processing an initial label set corresponding to each of the first environment images output by the labeler, and obtaining a first label set corresponding to each of the first environment images; Extracting feature points from each of the first environment images to obtain an initial feature point set corresponding to each of the first environment images; According to a preset edge feature extraction method, combining each of the first environment images, a plurality of edge features are obtained; Merging the initial feature point set and the edge feature of each of the first environment images to obtain a first feature point set corresponding to each of the first environment images; A three-dimensional model of the sea area is constructed according to the first feature point set and the first label set corresponding to each of the first environment images.
3. A method for planning a maritime transport route according to claim 2, characterized in that: The method of extracting edge features according to a preset method is combined with each of the first environment images to obtain a plurality of edge features, specifically including: Constructing a filter factor according to a two-dimensional Gaussian function, and filtering each of the first environment images using Gaussian filtering in combination with the filter factor to obtain a plurality of second environment images; Calculating the gradient magnitude and direction of each second environment image, and iterating the gradient magnitude of each second environment image to obtain initial edge feature data of each second environment image; Threshold segmentation and data extraction are performed on the initial edge feature data of each second environment image to obtain a plurality of edge features.
4. A method for planning a maritime transport route according to claim 2, characterized in that: The constructing the three-dimensional model of the sea area according to the first feature point set and the first label set corresponding to each of the first environment images specifically includes: Matching the first feature point set and the first tag set corresponding to each of the first environment images to obtain some obstacle information; wherein the obstacle information includes: obstacle position information and obstacle geometry information; Constructing a sea area obstacle location grid map based on the obstacle location information; Constructing a sea area obstacle geometric view according to the obstacle geometric information; A three-dimensional model of the sea area is constructed based on the sea area obstacle location grid map and the sea area obstacle geometric view.
5. A method for planning a marine transport route according to claim 4, characterized in that: The current position and target position of the ship are obtained, and an initial path is obtained according to a preset path solving algorithm in combination with the three-dimensional model of the sea area, specifically including: Get the current position and target position of the ship; According to the current position and target position of the ship, combined with the sea obstacle position grid map of the three-dimensional sea model, an incentive value function is constructed: Among them, τ(i,j) is the excitation value between the i-th grid and the j-th grid, d se is the Euclidean distance between the current position and the target position of the ship, d si is the Euclidean distance between the current position of the ship and the i-th grid, d ij is the Euclidean distance between the i-th grid and the j-th grid, d je is the Euclidean distance between the i-th grid and the target position of the ship; According to the sea area obstacle position grid map of the three-dimensional sea area model, a steering function is constructed: Among them, μ i Indicates the steering value starting from the i-th grid, d ij is the Euclidean distance between the i-th grid and the j-th grid, The steering angle between the i-th grid and the j-th grid; According to the geometric view of the sea area obstacles of the three-dimensional model of the sea area, a collision function is constructed: Among them, D i,j is the collision risk from the i-th grid to the j-th grid, K is the number of obstacles between the i-th grid and the j-th grid, V k is the volume of the kth obstacle, H k is the highest height of the volume of the kth obstacle, and a is the weight value; According to the incentive value function, the steering function and the collision function, a path search function is constructed: Where P is the path search function, α, β and γ are the weights of the incentive value, steering value and collision risk, τ(i,j) is the incentive value between the i-th grid and the j-th grid, μ i represents the steering value starting from the i-th grid, D i,j is the collision risk from the i-th grid to the j-th grid, n represents the number of grids in the historical path, j∈allowed i It means that the current i-th grid and j-th grid are accessible. Indicates that the i-th grid and the j-th grid are inaccessible; According to the search function, an initial path is obtained in combination with the current position and the target position of the ship.
6. A method for planning a marine transport route according to claim 1, characterized in that: The obtaining of historical ship navigation tracks, training a preset initial route prediction model in combination with the historical ship navigation tracks according to a preset model training method to obtain an optimal route prediction model; inputting the current position and the target position into the optimal route prediction model to obtain a predicted path, specifically includes: Obtain the preset initial route prediction model and historical ship route trajectory; According to the current position and the target position, a reward function and a loss function of an initial route prediction model are set; Inputting the historical ship route trajectory into the initial route prediction model, combining the reward function and the loss function, iteratively training the initial route prediction model to obtain an optimal route prediction model; The current position and the target position are input into the optimal route prediction model to obtain a predicted path.
7. A method for planning a maritime transport route according to any one of claims 5 to 6, characterized in that: The preset deviation algorithm is used to combine the initial path and the predicted path to obtain a path deviation score, and the initial path is corrected according to the path deviation score and combined with the three-dimensional model of the sea area to obtain the optimal path, which specifically includes: According to the sea area obstacle position grid map of the three-dimensional sea area model, the initial path and the predicted path are divided respectively to obtain a first grid set corresponding to the initial path and a second grid set corresponding to the predicted path; According to a preset first calculation method, the first grid set and the second grid set are combined to obtain a first path score of the initial path and a second path score of the predicted path respectively; wherein the preset first calculation method includes: Among them, S is the path score, x i and i are the horizontal and vertical coordinate values of the center point of the i-th grid, respectively, and n is the number of grids in the grid set; According to the first path score and the second path score, combined with the current position and the target position of the ship, a path deviation score is calculated: Where Deviate is the path deviation score, S1 is the first path score, S2 is the second path score, x s and s are the horizontal and vertical coordinate values of the current position of the ship, respectively. e and e are the horizontal and vertical coordinate values of the ship’s target position respectively; A preset deviation standard value is obtained, and the initial path is corrected in combination with the path deviation score and the sea area obstacle position grid map of the sea area three-dimensional model to obtain an optimal path.
8. A method for planning a marine transport route according to claim 1, characterized in that: The acquiring of the environmental information and the working information of the ship and combining the optimal path to obtain the optimal sailing speed of the ship specifically includes: Acquiring environmental information and operating information of the ship, wherein the environmental information includes: water flow direction and water flow velocity; the operating information includes: total amount of ship fuel and ship weight; Constructing an environmental function of the ship according to the flow direction of the water body and the flow velocity of the water body; According to the total amount of fuel of the ship and the weight of the ship, combined with the environmental function, construct a speed control function of the ship; According to the optimal path, combined with the three-dimensional model of the sea area and the speed control function, the optimal sailing speed of the ship is obtained.
9. A marine transport route planning system, characterized in that: include: Sea area three-dimensional model building module, initial path acquisition module, predicted path acquisition module, optimal path acquisition module, optimal navigation speed acquisition module and ship navigation module; The three-dimensional model building module of the sea area is used to obtain a number of environmental images of the ship, obtain a number of first label sets and a number of first feature point sets according to a preset image recognition method, and build a three-dimensional model of the sea area according to the first label sets and the first feature point sets; The initial path acquisition module is used to acquire the current position and target position of the ship, and obtain the initial path according to the preset path solving algorithm in combination with the three-dimensional model of the sea area; The predicted path acquisition module is used to acquire the historical ship navigation track, and train the preset initial route prediction model in combination with the historical ship navigation track according to the preset model training method to obtain the optimal route prediction model; the current position and the target position are input into the optimal route prediction model to obtain the predicted path; The optimal path acquisition module is used to obtain a path deviation score based on a preset deviation algorithm, in combination with the initial path and the predicted path, and to correct the initial path according to the path deviation score and in combination with the three-dimensional model of the sea area to obtain an optimal path; The optimal sailing speed acquisition module is used to acquire the environmental information and working information of the ship, and obtain the optimal sailing speed of the ship in combination with the optimal path; The ship navigation module is used to control the ship to navigate according to the optimal path and the optimal navigation speed.
10. A marine transport route planning system as claimed in claim 9, characterized in that: The initial path acquisition module includes: an initial path acquisition unit; The initial path acquisition unit is used to acquire the current position and target position of the ship; According to the current position and target position of the ship, combined with the sea obstacle position grid map of the three-dimensional sea model, an incentive value function is constructed: Among them, τ(i,j) is the excitation value between the i-th grid and the j-th grid, d se is the Euclidean distance between the current position and the target position of the ship, d si is the Euclidean distance between the current position of the ship and the i-th grid, d ij is the Euclidean distance between the i-th grid and the j-th grid, d je is the Euclidean distance between the i-th grid and the target position of the ship; According to the sea area obstacle position grid map of the three-dimensional sea area model, a steering function is constructed: Among them, μ i Indicates the steering value starting from the i-th grid, d ij is the Euclidean distance between the i-th grid and the j-th grid, The steering angle between the i-th grid and the j-th grid; According to the geometric view of the sea obstacles in the three-dimensional model of the sea area, a collision function is constructed: Among them, D i,j is the collision risk from the i-th grid to the j-th grid, K is the number of obstacles between the i-th grid and the j-th grid, V k is the volume of the kth obstacle, H k is the highest height of the volume of the kth obstacle, and a is the weight value; According to the incentive value function, the steering function and the collision function, a path search function is constructed: Where P is the path search function, α, β and γ are the weights of the incentive value, steering value and collision risk, τ(i,j) is the incentive value between the i-th grid and the j-th grid, μ i represents the steering value starting from the i-th grid, D i,j is the collision risk from the i-th grid to the j-th grid, n represents the number of grids in the historical path, j∈allowed i It means that the current i-th grid and j-th grid are accessible. Indicates that the i-th grid and the j-th grid are inaccessible; According to the search function, an initial path is obtained in combination with the current position and the target position of the ship.
