Route planning method and system based on neural network
Through the neural network-based route planning method, similar ships are screened and route planning models are trained in real time. Combined with charts and weather data, the problem of insufficient route planning efficiency and safety in the existing technology is solved, and more reliable and safe routes are achieved planning.
Patent Information
- Application Number
- CN202410868714.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-01
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2044-07-01
AI Technical Summary
The existing route planning technology has failed to effectively improve the efficiency and accuracy of route planning, and has not fully considered the screening of data sources, which has affected navigation safety.
Using a neural network-based route planning method, by obtaining the ship information and area data of the target ship, screening out similar ships based on the similarity algorithm, and training the route planning algorithm model in real time, combining navigation area chart data and weather forecast information to optimize the route Plan for increased security.
It improves the efficiency and accuracy of route planning, enhances the safety of ship navigation, and reduces the occurrence of navigation accidents.
Smart Images

Figure CN118408555B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of data processing, and in particular to a route planning method and system based on neural network. Background Art
[0002] With the increase in the demand for maritime navigation and the increasing emphasis on navigation safety, the demand for automatic planning of ship routes is also increasing. A large number of data processing technologies have been introduced to meet these needs. How to accurately predict safe routes has become an important technical issue. In the existing route prediction and planning technology, the route planning algorithm is generally trained simply through historical navigation data, and no further screening of data sources is considered to improve the adaptability and accuracy of the planning results of the planning algorithm model. It can be seen that the existing technology has defects that need to be solved urgently. Summary of the invention
[0003] The technical problem to be solved by the present invention is to provide a route planning method and system based on neural network, which can effectively improve the efficiency and accuracy of route planning, improve the safety of ship navigation, and reduce navigation accidents.
[0004] In order to solve the above technical problems, the first aspect of the present invention discloses a route planning method based on a neural network, the method comprising:
[0005] In response to a route planning request of a target ship, obtaining ship information of the target ship and regional data of the ocean area where the target ship is located;
[0006] According to the ship information, based on a similarity algorithm, multiple similar ships are selected from multiple candidate ships;
[0007] According to the navigation data of the similar ships, a route planning algorithm model of the target ship is obtained by real-time training;
[0008] The ship information and the area data are input into the route planning algorithm model for planning calculation to obtain the planned route of the target ship.
[0009] As an optional embodiment, in the first aspect of the present invention, the ship information includes ship type, ship position, ship historical navigation data, ship crew information and ship sensor parameters; the ship sensor parameters include ship speed, ship acceleration, ship heading, ship temperature, ship humidity and ship image data; the area data includes obstacle area data, land boundary data, floating animal data and weather forecast data.
[0010] As an optional implementation, in the first aspect of the present invention, the selecting a plurality of similar ships from a plurality of candidate ships based on the ship information and a similarity algorithm comprises:
[0011] The satellite communication equipment of the target ship broadcasts information to all other ships within a preset radius, and obtains response information from the other ships; the response information includes the computing equipment status information of the other ships;
[0012] Inputting the computing equipment status information of each of the other ships into a preset equipment status prediction algorithm model to obtain equipment status prediction parameters of each of the other ships;
[0013] Determine all other ships that have sent the response information and whose equipment status prediction parameter is greater than a preset first parameter threshold as candidate ships;
[0014] Calculating information similarity between the ship information of each candidate ship and the ship information of the target ship;
[0015] Sorting all the candidate ships from large to small according to the information similarity to obtain a ship sequence;
[0016] All the candidate ships that are a preset number before the ship sequence and whose information similarity is greater than a preset first similarity threshold are determined as a plurality of similar ships.
[0017] As an optional implementation, in the first aspect of the present invention, the information similarity is calculated according to the following rules:
[0018] For the ship information of the candidate ship and the ship information of the target ship, a first vector distance between vectorized representations of basic ship information in the two ship information is calculated; the basic ship information includes ship type, ship position and ship crew information;
[0019] Calculating a second vector distance between vectorized representations of historical ship navigation data in the two ship information; the historical ship navigation data includes the route shape of the ship in the historical time period, the navigation area and the nautical chart data of the area;
[0020] calculating a third vector distance between the vectorized representations of the ship sensing parameters in the two ship information;
[0021] The inverse of the weighted sum of the first vector distance, the second vector distance and the third vector distance is calculated to obtain the information similarity between the ship information of the candidate ship and the ship information of the target ship; wherein the weights corresponding to the second vector distance, the third vector distance and the first vector distance decrease in sequence.
[0022] As an optional implementation, in the first aspect of the present invention, the route planning algorithm model of the target ship is obtained by real-time training based on the navigation data of the similar ship, including:
[0023] Acquire safe route data of each of the similar ships and corresponding navigation area chart data; the navigation area chart data includes obstacle areas and weather forecast information;
[0024] Inputting each of the navigation area chart data into a preset route planning algorithm model to obtain a planned route result, and adjusting the model parameters of the route planning algorithm model according to the corresponding safe route data and the planned route result;
[0025] Repeat the previous step until the route similarity between the planned route result predicted by the route planning algorithm model and the corresponding safe route data is lower than a preset second similarity threshold, so as to obtain the route planning algorithm model of the target ship.
[0026] As an optional implementation, in the first aspect of the present invention, the step of inputting each of the navigation area chart data into a preset route planning algorithm model to obtain a planned route result, and adjusting the model parameters of the route planning algorithm model according to the corresponding safe route data and the planned route result, comprises:
[0027] The objective function is set to maximize the sum of the distances between the planned route and each obstacle area;
[0028] The restriction conditions include that the distance between the planned route and any obstacle area is greater than a distance threshold and the average value of the weather danger values corresponding to the weather forecast information of all areas through which the planned route passes is less than a danger threshold; the weather danger value is obtained by predicting the weather forecast information using a pre-trained classifier model;
[0029] According to the objective function and the constraint conditions, a route planning algorithm model is established based on a spatial search algorithm;
[0030] Inputting the nautical chart data of each navigation area into the route planning algorithm model for iterative calculation until the planned route result is calculated;
[0031] Calculating the route difference between the corresponding safe route data and the planned route result and the route corresponding to each area part;
[0032] Determine the region where the route difference is greater than a first difference threshold as a high difference portion;
[0033] Determine the area portion where the route difference is greater than a second difference threshold as a low difference portion; the second difference threshold is less than the first difference threshold;
[0034] When the total number of all the high-difference parts is greater than a first number threshold, randomly adjusting the distance threshold to obtain a new distance threshold for next planning;
[0035] When the total number of all the low-difference parts is greater than a second number threshold, the danger threshold is randomly adjusted to obtain a new danger threshold for next planning.
[0036] As an optional implementation, in the first aspect of the present invention, the step of inputting the ship information and the area data into the route planning algorithm model for planning calculation to obtain the planned route of the target ship includes:
[0037] According to the ship information of the target ship at the current time point and the ship information stored at multiple historical time points, it is determined whether there is any information change of the target ship, and a first determination result is obtained;
[0038] If the first judgment result is no, the area data is input into the route planning algorithm model to calculate the planned route of the target ship;
[0039] If the first judgment result is yes, the satellite communication device of the target ship broadcasts a regional data acquisition request to all other ships within a preset radius, and acquires other regional data of the area sent by all other ships;
[0040] Calculating an average value of the differences between all other regional data and the regional data to obtain a difference parameter corresponding to the regional data;
[0041] Determine whether the difference parameter is greater than a preset second parameter threshold, and obtain a second determination result;
[0042] If the second judgment result is no, the area data is input into the route planning algorithm model to calculate the planned route of the target ship;
[0043] If the second judgment result is yes, inputting each of the other area data into the route planning algorithm model to obtain the corresponding first planned route;
[0044] An average value of all the first planned routes is calculated to obtain the planned route of the target ship.
[0045] As an optional implementation, in the first aspect of the present invention, judging whether there is any information change of the target ship according to the ship information of the target ship at the current time point and the ship information stored at multiple historical time points includes:
[0046] Determine whether the information difference between the ship information of the target ship at the current time point and the ship information at the most recent historical time point is greater than a preset third difference threshold, and obtain a third determination result;
[0047] If the third judgment result is no, determining that there is no information change of the target ship;
[0048] If the third judgment result is yes, the ship information sequence corresponding to the target ship is input into the trained information gradient judgment neural network model to obtain the information gradient probability corresponding to the target ship; the ship information sequence is an information sequence obtained by arranging all ship information corresponding to the target ship at the current time point and all historical time points from early to late according to the storage time points; the information gradient judgment neural network model is trained by a training data set including a plurality of training ship information sequences and corresponding information gradient annotations of whether the normal;
[0049] Determine whether the information gradual change probability is less than a preset probability threshold, and obtain a fourth determination result;
[0050] If the fourth judgment result is no, determining that there is no information change of the target ship;
[0051] If the fourth judgment result is yes, it is determined that there is information change of the target ship.
[0052] A second aspect of an embodiment of the present invention discloses a route planning system based on a neural network, the system comprising:
[0053] An acquisition module, used for acquiring ship information of the target ship and regional data of the ocean area where the target ship is located in response to a route planning request of the target ship;
[0054] A screening module, used for screening a plurality of similar ships from a plurality of candidate ships based on the ship information and a similarity algorithm;
[0055] A training module, used for obtaining a route planning algorithm model of the target ship through real-time training based on the navigation data of the similar ship;
[0056] The planning module is used to input the ship information and the area data into the route planning algorithm model for planning calculation to obtain the planned route of the target ship.
[0057] As an optional embodiment, in the second aspect of the present invention, the ship information includes ship type, ship position, ship historical navigation data, ship crew information and ship sensor parameters; the ship sensor parameters include ship speed, ship acceleration, ship heading, ship temperature, ship humidity and ship image data; the area data includes obstacle area data, land boundary data, floating animal data and weather forecast data.
[0058] As an optional implementation, in the second aspect of the present invention, the specific manner in which the screening module screens out a plurality of similar ships from a plurality of candidate ships based on the ship information and a similarity algorithm includes:
[0059] The satellite communication equipment of the target ship broadcasts information to all other ships within a preset radius, and obtains response information from the other ships; the response information includes the computing equipment status information of the other ships;
[0060] Inputting the computing equipment status information of each of the other ships into a preset equipment status prediction algorithm model to obtain equipment status prediction parameters of each of the other ships;
[0061] Determine all other ships that have sent the response information and whose equipment status prediction parameter is greater than a preset first parameter threshold as candidate ships;
[0062] Calculating information similarity between the ship information of each candidate ship and the ship information of the target ship;
[0063] Sorting all the candidate ships from large to small according to the information similarity to obtain a ship sequence;
[0064] All the candidate ships that are a preset number before the ship sequence and whose information similarity is greater than a preset first similarity threshold are determined as a plurality of similar ships.
[0065] As an optional implementation, in the second aspect of the present invention, the information similarity is calculated according to the following rules:
[0066] For the ship information of the candidate ship and the ship information of the target ship, a first vector distance between vectorized representations of basic ship information in the two ship information is calculated; the basic ship information includes ship type, ship position and ship crew information;
[0067] Calculating a second vector distance between vectorized representations of historical ship navigation data in the two ship information; the historical ship navigation data includes the route shape of the ship in the historical time period, the navigation area and the nautical chart data of the area;
[0068] calculating a third vector distance between the vectorized representations of the ship sensing parameters in the two ship information;
[0069] The inverse of the weighted sum of the first vector distance, the second vector distance and the third vector distance is calculated to obtain the information similarity between the ship information of the candidate ship and the ship information of the target ship; wherein the weights corresponding to the second vector distance, the third vector distance and the first vector distance decrease in sequence.
[0070] As an optional implementation, in the second aspect of the present invention, the specific manner in which the training module obtains the route planning algorithm model of the target ship through real-time training based on the navigation data of the similar ship includes:
[0071] Acquire safe route data of each of the similar ships and corresponding navigation area chart data; the navigation area chart data includes obstacle areas and weather forecast information;
[0072] Inputting each of the navigation area chart data into a preset route planning algorithm model to obtain a planned route result, and adjusting the model parameters of the route planning algorithm model according to the corresponding safe route data and the planned route result;
[0073] Repeat the previous step until the route similarity between the planned route result predicted by the route planning algorithm model and the corresponding safe route data is lower than a preset second similarity threshold, so as to obtain the route planning algorithm model of the target ship.
[0074] As an optional implementation, in the second aspect of the present invention, the training module inputs each of the navigation area chart data into a preset route planning algorithm model to obtain a planned route result, and adjusts the model parameters of the route planning algorithm model according to the corresponding safe route data and the planned route result, including:
[0075] The objective function is set to maximize the sum of the distances between the planned route and each obstacle area;
[0076] The restriction conditions include that the distance between the planned route and any obstacle area is greater than a distance threshold and the average value of the weather danger values corresponding to the weather forecast information of all areas through which the planned route passes is less than a danger threshold; the weather danger value is obtained by predicting the weather forecast information using a pre-trained classifier model;
[0077] According to the objective function and the constraint conditions, a route planning algorithm model is established based on a spatial search algorithm;
[0078] Inputting the nautical chart data of each navigation area into the route planning algorithm model for iterative calculation until the planned route result is calculated;
[0079] Calculating the route difference between the corresponding safe route data and the planned route result and the route corresponding to each area part;
[0080] Determine the region where the route difference is greater than a first difference threshold as a high difference portion;
[0081] Determine the area portion where the route difference is greater than a second difference threshold as a low difference portion; the second difference threshold is less than the first difference threshold;
[0082] When the total number of all the high-difference parts is greater than a first number threshold, randomly adjusting the distance threshold to obtain a new distance threshold for next planning;
[0083] When the total number of all the low-difference parts is greater than a second number threshold, the danger threshold is randomly adjusted to obtain a new danger threshold for next planning.
[0084] As an optional implementation, in the second aspect of the present invention, the planning module inputs the ship information and the area data into the route planning algorithm model for planning calculation to obtain a specific method of the planned route of the target ship, including:
[0085] According to the ship information of the target ship at the current time point and the ship information stored at multiple historical time points, it is determined whether there is any information change of the target ship, and a first determination result is obtained;
[0086] If the first judgment result is no, the area data is input into the route planning algorithm model to calculate the planned route of the target ship;
[0087] If the first judgment result is yes, the satellite communication device of the target ship broadcasts a regional data acquisition request to all other ships within a preset radius, and acquires other regional data of the area sent by all other ships;
[0088] Calculating an average value of the differences between all other regional data and the regional data to obtain a difference parameter corresponding to the regional data;
[0089] Determine whether the difference parameter is greater than a preset second parameter threshold, and obtain a second determination result;
[0090] If the second judgment result is no, the area data is input into the route planning algorithm model to calculate the planned route of the target ship;
[0091] If the second judgment result is yes, inputting each of the other area data into the route planning algorithm model to obtain the corresponding first planned route;
[0092] An average value of all the first planned routes is calculated to obtain the planned route of the target ship.
[0093] As an optional implementation, in the second aspect of the present invention, the specific manner in which the planning module determines whether there is any information change of the target ship based on the ship information of the target ship at the current time point and the ship information stored at multiple historical time points includes:
[0094] Determine whether the information difference between the ship information of the target ship at the current time point and the ship information at the most recent historical time point is greater than a preset third difference threshold, and obtain a third determination result;
[0095] If the third judgment result is no, determining that there is no information change of the target ship;
[0096] If the third judgment result is yes, the ship information sequence corresponding to the target ship is input into the trained information gradient judgment neural network model to obtain the information gradient probability corresponding to the target ship; the ship information sequence is an information sequence obtained by arranging all ship information corresponding to the target ship at the current time point and all historical time points from early to late according to the storage time points; the information gradient judgment neural network model is trained by a training data set including a plurality of training ship information sequences and corresponding information gradient annotations of whether the normal;
[0097] Determine whether the information gradual change probability is less than a preset probability threshold, and obtain a fourth determination result;
[0098] If the fourth judgment result is no, determining that there is no information change of the target ship;
[0099] If the fourth judgment result is yes, it is determined that there is information change of the target ship.
[0100] The third aspect of the present invention discloses another route planning system based on a neural network, the system comprising:
[0101] A memory storing executable program code;
[0102] a processor coupled to the memory;
[0103] The processor calls the executable program code stored in the memory to execute part or all of the steps in the neural network-based route planning method disclosed in the first aspect of the present invention.
[0104] The fourth aspect of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute some or all of the steps in the neural network-based route planning method disclosed in the first aspect of the present invention.
[0105] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:
[0106] The present invention can screen out multiple similar ships based on ship information, and then perform real-time model training based on the data of similar ships to obtain a planning algorithm model with more data relevance, so as to obtain a more reliable and safe route through planning calculation, thereby effectively improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents. BRIEF DESCRIPTION OF THE DRAWINGS
[0107] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0108] Figure 1 It is a flowchart of a route planning method based on a neural network disclosed in an embodiment of the present invention.
[0109] Figure 2 It is a structural schematic diagram of a route planning system based on a neural network disclosed in an embodiment of the present invention.
[0110] Figure 3 It is a structural schematic diagram of another route planning system based on neural network disclosed in an embodiment of the present invention. DETAILED DESCRIPTION
[0111] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings 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.
[0112] The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, device, product or equipment that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or equipment.
[0113] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.
[0114] The present invention discloses a route planning method and system based on a neural network, which can screen out multiple similar ships based on ship information, and then perform model training in real time based on the data of similar ships to obtain a planning algorithm model with more data relevance, so as to obtain a more reliable and safe route through planning calculation, thereby effectively improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents. The following are detailed descriptions.
[0115] For example, see Figure 1 , Figure 1 is a flow chart of a route planning method based on a neural network disclosed in an embodiment of the present invention. Figure 1 The described neural network-based route planning method can be applied to a data processing system / data processing device / data processing server (wherein the server includes a local processing server or a cloud processing server). Figure 1 As shown, the route planning method based on neural network may include the following operations:
[0116] 101. In response to a route planning request of a target ship, obtain ship information of the target ship and regional data of the ocean area where the target ship is located.
[0117] 102. According to the ship information and based on a similarity algorithm, multiple similar ships are selected from multiple candidate ships.
[0118] 103. Based on the navigation data of similar ships, the route planning algorithm model of the target ship is trained in real time.
[0119] 104. The ship information and the area data are input into the route planning algorithm model for planning calculation to obtain the planned route of the target ship.
[0120] It can be seen that the above-mentioned embodiment of the invention can screen out multiple similar ships based on ship information, and then perform real-time model training based on the data of similar ships to obtain a planning algorithm model with more data relevance, so as to obtain a more reliable and safe route through planning calculation, thereby effectively improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents.
[0121] As an optional embodiment, in the above steps, the ship information includes ship type, ship position, ship historical navigation data, ship crew information and ship sensor parameters; the ship sensor parameters include ship speed, ship acceleration, ship heading, ship temperature, ship humidity and ship image data; the area data includes obstacle area data, land boundary data, floating animal data and weather forecast data.
[0122] It can be seen that through the above optional embodiments, the content of ship information and regional data is clarified, which can better characterize the characteristics of ships and regions, so that subsequent ship screening and route planning are more accurate, thereby assisting in improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents.
[0123] As an optional embodiment, in the above step, selecting multiple similar ships from multiple candidate ships based on the ship information and a similarity algorithm includes:
[0124] The satellite communication equipment of the target ship broadcasts information to all other ships within a preset radius and obtains response information from other ships; optionally, the response information includes status information of computing equipment of other ships;
[0125] Inputting the computing equipment status information of each other ship into a preset equipment status prediction algorithm model to obtain equipment status prediction parameters of each other ship;
[0126] Determine all other ships that have sent response information and whose equipment status prediction parameter is greater than a preset first parameter threshold as candidate ships;
[0127] Calculate the information similarity between the ship information of each candidate ship and the ship information of the target ship;
[0128] Sort all candidate ships from large to small according to information similarity to obtain a ship sequence;
[0129] All candidate ships that are a preset number in front of the ship sequence and whose information similarity is greater than a preset first similarity threshold are determined as a plurality of similar ships.
[0130] It can be seen that through the above-mentioned optional embodiments, it is possible to communicate with ships within a radius based on information broadcasting, and then screen out similar ships that are more similar to the target ship based on equipment status prediction and similarity calculation, so as to be used for subsequent training of a planning algorithm model with more data relevance, thereby assisting in improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents.
[0131] As an optional embodiment, in the above steps, the information similarity is calculated according to the following rules:
[0132] For the ship information of the candidate ship and the ship information of the target ship, a first vector distance between vectorized representations of basic ship information in the two ship information is calculated; optionally, the basic ship information includes ship type, ship position and ship crew information;
[0133] Calculate a second vector distance between the vectorized representations of the historical navigation data of the ships in the two ship information; optionally, the historical navigation data of the ships include the route shape of the ships in the historical time period, the navigation passing area and the nautical chart data of the passing area;
[0134] calculating a third vector distance between the vectorized representations of the ship sensing parameters in the two ship information;
[0135] The inverse of the weighted sum of the first vector distance, the second vector distance and the third vector distance is calculated to obtain the information similarity between the ship information of the candidate ship and the ship information of the target ship; wherein the weights corresponding to the second vector distance, the third vector distance and the first vector distance decrease in sequence.
[0136] It can be seen that through the above optional embodiments, the detailed calculation rules of information similarity are clarified, and the similarity between ship information parameters can be accurately calculated, so as to be used in the subsequent accurate screening of similar ships that are more similar to the target ship, thereby assisting in improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents.
[0137] As an optional embodiment, in the above steps, the route planning algorithm model of the target ship is obtained by real-time training based on the navigation data of similar ships, including:
[0138] Obtain safe route data for each similar ship and corresponding navigation area chart data; optionally, the navigation area chart data includes obstacle areas and weather forecast information;
[0139] Inputting the nautical chart data of each navigation area into the preset route planning algorithm model to obtain the planned route result, and adjusting the model parameters of the route planning algorithm model according to the corresponding safe route data and the planned route result;
[0140] Repeat the previous step until the route similarity between the planned route result predicted by the route planning algorithm model and the corresponding safe route data is lower than the preset second similarity threshold, so as to obtain the route planning algorithm model of the target ship.
[0141] It can be seen that through the above-mentioned optional embodiments, the model parameters of the route planning algorithm model can be trained and adjusted through the navigation data of each similar ship to obtain a planning algorithm model with more data relevance, thereby assisting in improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents.
[0142] As an optional embodiment, in the above steps, the nautical chart data of each navigation area is input into a preset route planning algorithm model to obtain a planned route result, and the model parameters of the route planning algorithm model are adjusted according to the corresponding safe route data and the planned route result, including:
[0143] The objective function is set to maximize the sum of the distances between the planned route and each obstacle area;
[0144] The setting restriction conditions include that the distance between the planned route and any obstacle area is greater than a distance threshold and the average value of weather danger values corresponding to weather forecast information of all areas passed by the planned route is less than a danger threshold; optionally, the weather danger value is obtained by predicting the weather forecast information using a pre-trained classifier model;
[0145] According to the objective function and constraints, a route planning algorithm model is established based on the spatial search algorithm;
[0146] Input the nautical chart data of each navigation area into the route planning algorithm model for iterative calculation until the planned route result is calculated;
[0147] Calculate the route difference between the corresponding safe route data and the planned route results for each area;
[0148] Determine a region where the route difference is greater than a first difference threshold as a high difference region;
[0149] Determine the area portion where the route difference is greater than a second difference threshold as a low difference portion; optionally, the second difference threshold is less than the first difference threshold;
[0150] When the total number of all high-difference parts is greater than a first number threshold, randomly adjusting the distance threshold to obtain a new distance threshold for next planning;
[0151] When the total number of all low-difference parts is greater than the second number threshold, the danger threshold is randomly adjusted to obtain a new danger threshold for the next planning.
[0152] It can be seen that through the above-mentioned optional embodiments, a route planning algorithm model can be established based on the preset objective function and constraints, and the different planning parameters of the model can be adjusted according to the difference between each planning result and the corresponding safe route data, so as to obtain a planning algorithm model with more data relevance, thereby assisting in improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents.
[0153] As an optional embodiment, in the above steps, the ship information and the area data are input into the route planning algorithm model for planning calculation to obtain the planned route of the target ship, including:
[0154] According to the ship information of the target ship at the current time point and the ship information stored at multiple historical time points, it is determined whether there is any information change of the target ship, and a first determination result is obtained;
[0155] If the first judgment result is no, the regional data is input into the route planning algorithm model to calculate the planned route of the target ship;
[0156] If the first judgment result is yes, the satellite communication device of the target ship broadcasts a regional data acquisition request to all other ships within a preset radius, and acquires other regional data of the area sent by all other ships;
[0157] Calculate the average value of the differences between all other regional data and the regional data to obtain the difference parameter corresponding to the regional data;
[0158] Determine whether the difference parameter is greater than a preset second parameter threshold, and obtain a second determination result;
[0159] If the second judgment result is no, the regional data is input into the route planning algorithm model to calculate the planned route of the target ship;
[0160] If the second determination result is yes, inputting each other area data into the route planning algorithm model to obtain the corresponding first planned route;
[0161] The average of all the first planned routes is calculated to obtain the planned route of the target ship.
[0162] It can be seen that through the above-mentioned optional embodiments, it is possible to determine whether the target ship is in a data hijacking situation based on the judgment of changes in ship information and regional data, and in the event of a hijacking situation, use the nautical chart data sent by other ships to plan the route, so as to prevent the target ship from being hijacked by data to plan an erroneous route and cause navigation accidents, thereby improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents.
[0163] As an optional embodiment, in the above step, judging whether there is any information change of the target ship according to the ship information of the target ship at the current time point and the ship information stored at multiple historical time points includes:
[0164] Determine whether the information difference between the ship information of the target ship at the current time point and the ship information at the most recent historical time point is greater than a preset third difference threshold, and obtain a third determination result;
[0165] If the third judgment result is no, it is determined that there is no information change of the target ship;
[0166] If the third judgment result is yes, the ship information sequence corresponding to the target ship is input into the trained information gradient judgment neural network model to obtain the information gradient probability corresponding to the target ship; optionally, the ship information sequence is an information sequence obtained by arranging all ship information corresponding to the target ship at the current time point and all historical time points from early to late according to the storage time points; the information gradient judgment neural network model is trained by a training data set including a plurality of training ship information sequences and corresponding information gradient annotations of whether the normal;
[0167] Determine whether the information gradual change probability is less than a preset probability threshold, and obtain a fourth determination result;
[0168] If the fourth judgment result is no, it is determined that there is no information change of the target ship;
[0169] If the fourth judgment result is yes, it is determined that there is information change on the target ship.
[0170] Specifically, the information gradient judgment neural network model is a neural network model of a CNN network structure, which includes multiple convolutional layers, a fully connected layer and a SoftMax layer. It is used to input multiple training ship information sequences as training data and extract vector features, and then output the judgment result of whether the gradient is normal from the fully connected layer. The loss function is judged by the SoftMax layer in combination with the corresponding information whether the gradient is normal to optimize the model parameters until the information gradient judgment neural network model is trained.
[0171] It can be seen that through the above optional embodiments, it is possible to fully and comprehensively judge whether there is any information change in the target ship based on the difference between the ship information and the historical ship information and the information gradient judgment neural network model, so as to facilitate the subsequent determination of whether there is data hijacking in the target ship, so as to prevent the target ship from being hijacked by data to plan an erroneous route and cause navigation accidents, thereby improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents.
[0172] For example 2, please refer to Figure 2 , Figure 2 is a schematic diagram of the structure of a route planning system based on a neural network disclosed in an embodiment of the present invention. Figure 2 The described neural network-based route planning system can be applied to a data processing system / data processing device / data processing server (wherein the server includes a local processing server or a cloud processing server). Figure 2 As shown, the neural network-based route planning system may include:
[0173] The acquisition module 201 is used to respond to the route planning request of the target ship and acquire the ship information of the target ship and the regional data of the ocean area where the target ship is located.
[0174] The screening module 202 is used to screen out a plurality of similar ships from a plurality of candidate ships based on the ship information and a similarity algorithm.
[0175] The training module 203 is used to obtain the route planning algorithm model of the target ship through real-time training based on the navigation data of similar ships.
[0176] The planning module 204 is used to input the ship information and the area data into the route planning algorithm model for planning calculation to obtain the planned route of the target ship.
[0177] It can be seen that the above-mentioned embodiment of the invention can screen out multiple similar ships based on ship information, and then perform real-time model training based on the data of similar ships to obtain a planning algorithm model with more data relevance, so as to obtain a more reliable and safe route through planning calculation, thereby effectively improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents.
[0178] As an optional embodiment, the ship information includes ship type, ship position, ship historical navigation data, ship crew information and ship sensor parameters; the ship sensor parameters include ship speed, ship acceleration, ship heading, ship temperature, ship humidity and ship image data; the area data includes obstacle area data, land boundary data, floating animal data and weather forecast data.
[0179] It can be seen that through the above optional embodiments, the content of ship information and regional data is clarified, which can better characterize the characteristics of ships and regions, so that subsequent ship screening and route planning are more accurate, thereby assisting in improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents.
[0180] As an optional embodiment, the specific manner in which the screening module screens out multiple similar ships from multiple candidate ships based on the ship information and the similarity algorithm includes:
[0181] The satellite communication equipment of the target ship broadcasts information to all other ships within a preset radius and obtains response information from other ships; optionally, the response information includes status information of computing equipment of other ships;
[0182] Inputting the computing equipment status information of each other ship into a preset equipment status prediction algorithm model to obtain equipment status prediction parameters of each other ship;
[0183] Determine all other ships that have sent response information and whose equipment status prediction parameter is greater than a preset first parameter threshold as candidate ships;
[0184] Calculate the information similarity between the ship information of each candidate ship and the ship information of the target ship;
[0185] Sort all candidate ships from large to small according to information similarity to obtain a ship sequence;
[0186] All candidate ships that are a preset number in front of the ship sequence and whose information similarity is greater than a preset first similarity threshold are determined as a plurality of similar ships.
[0187] It can be seen that through the above-mentioned optional embodiments, it is possible to communicate with ships within a radius based on information broadcasting, and then screen out similar ships that are more similar to the target ship based on equipment status prediction and similarity calculation, so as to be used for subsequent training of a planning algorithm model with more data relevance, thereby assisting in improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents.
[0188] As an optional embodiment, the information similarity is calculated according to the following rules:
[0189] For the ship information of the candidate ship and the ship information of the target ship, a first vector distance between vectorized representations of basic ship information in the two ship information is calculated; optionally, the basic ship information includes ship type, ship position and ship crew information;
[0190] Calculate a second vector distance between the vectorized representations of the historical navigation data of the ships in the two ship information; optionally, the historical navigation data of the ships include the route shape of the ships in the historical time period, the navigation passing area and the nautical chart data of the passing area;
[0191] calculating a third vector distance between the vectorized representations of the ship sensing parameters in the two ship information;
[0192] The inverse of the weighted sum of the first vector distance, the second vector distance and the third vector distance is calculated to obtain the information similarity between the ship information of the candidate ship and the ship information of the target ship; wherein the weights corresponding to the second vector distance, the third vector distance and the first vector distance decrease in sequence.
[0193] It can be seen that through the above optional embodiments, the detailed calculation rules of information similarity are clarified, and the similarity between ship information parameters can be accurately calculated, so as to be used in the subsequent accurate screening of similar ships that are more similar to the target ship, thereby assisting in improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents.
[0194] As an optional embodiment, the specific method of obtaining the route planning algorithm model of the target ship by real-time training of the training module according to the navigation data of similar ships includes:
[0195] Obtain safe route data for each similar ship and corresponding navigation area chart data; optionally, the navigation area chart data includes obstacle areas and weather forecast information;
[0196] Inputting the nautical chart data of each navigation area into the preset route planning algorithm model to obtain the planned route result, and adjusting the model parameters of the route planning algorithm model according to the corresponding safe route data and the planned route result;
[0197] Repeat the previous step until the route similarity between the planned route result predicted by the route planning algorithm model and the corresponding safe route data is lower than the preset second similarity threshold, so as to obtain the route planning algorithm model of the target ship.
[0198] It can be seen that through the above-mentioned optional embodiments, the model parameters of the route planning algorithm model can be trained and adjusted through the navigation data of each similar ship to obtain a planning algorithm model with more data relevance, thereby assisting in improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents.
[0199] As an optional embodiment, the training module inputs the nautical chart data of each navigation area into a preset route planning algorithm model to obtain a planned route result, and adjusts the model parameters of the route planning algorithm model according to the corresponding safe route data and the planned route result, including:
[0200] The objective function is set to maximize the sum of the distances between the planned route and each obstacle area;
[0201] The setting restriction conditions include that the distance between the planned route and any obstacle area is greater than a distance threshold and the average value of weather danger values corresponding to weather forecast information of all areas passed by the planned route is less than a danger threshold; optionally, the weather danger value is obtained by predicting the weather forecast information using a pre-trained classifier model;
[0202] According to the objective function and constraints, a route planning algorithm model is established based on the spatial search algorithm;
[0203] Input the nautical chart data of each navigation area into the route planning algorithm model for iterative calculation until the planned route result is calculated;
[0204] Calculate the route difference between the corresponding safe route data and the planned route results for each area;
[0205] Determine a region where the route difference is greater than a first difference threshold as a high difference region;
[0206] Determine the area portion where the route difference is greater than a second difference threshold as a low difference portion; optionally, the second difference threshold is less than the first difference threshold;
[0207] When the total number of all high-difference parts is greater than a first number threshold, randomly adjusting the distance threshold to obtain a new distance threshold for next planning;
[0208] When the total number of all low-difference parts is greater than the second number threshold, the danger threshold is randomly adjusted to obtain a new danger threshold for the next planning.
[0209] It can be seen that through the above-mentioned optional embodiments, a route planning algorithm model can be established based on the preset objective function and constraints, and the different planning parameters of the model can be adjusted according to the difference between each planning result and the corresponding safe route data, so as to obtain a planning algorithm model with more data relevance, thereby assisting in improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents.
[0210] As an optional embodiment, the planning module inputs the ship information and the area data into the route planning algorithm model for planning calculation to obtain the specific method of planning the route of the target ship, including:
[0211] According to the ship information of the target ship at the current time point and the ship information stored at multiple historical time points, it is determined whether there is any information change of the target ship, and a first determination result is obtained;
[0212] If the first judgment result is no, the regional data is input into the route planning algorithm model to calculate the planned route of the target ship;
[0213] If the first judgment result is yes, the satellite communication device of the target ship broadcasts a regional data acquisition request to all other ships within a preset radius, and acquires other regional data of the area sent by all other ships;
[0214] Calculate the average value of the differences between all other regional data and the regional data to obtain the difference parameter corresponding to the regional data;
[0215] Determine whether the difference parameter is greater than a preset second parameter threshold, and obtain a second determination result;
[0216] If the second judgment result is no, the regional data is input into the route planning algorithm model to calculate the planned route of the target ship;
[0217] If the second determination result is yes, inputting each other area data into the route planning algorithm model to obtain the corresponding first planned route;
[0218] The average of all the first planned routes is calculated to obtain the planned route of the target ship.
[0219] It can be seen that through the above-mentioned optional embodiments, it is possible to determine whether the target ship is in a data hijacking situation based on the judgment of changes in ship information and regional data, and in the event of a hijacking situation, use the nautical chart data sent by other ships to plan the route, so as to prevent the target ship from being hijacked by data to plan an erroneous route and cause navigation accidents, thereby improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents.
[0220] As an optional embodiment, the planning module determines whether there is any information change of the target ship according to the ship information of the target ship at the current time point and the ship information stored at multiple historical time points in a specific manner, including:
[0221] Determine whether the information difference between the ship information of the target ship at the current time point and the ship information at the most recent historical time point is greater than a preset third difference threshold, and obtain a third determination result;
[0222] If the third judgment result is no, it is determined that there is no information change of the target ship;
[0223] If the third judgment result is yes, the ship information sequence corresponding to the target ship is input into the trained information gradient judgment neural network model to obtain the information gradient probability corresponding to the target ship; optionally, the ship information sequence is an information sequence obtained by arranging all ship information corresponding to the target ship at the current time point and all historical time points from early to late according to the storage time points; the information gradient judgment neural network model is trained by a training data set including a plurality of training ship information sequences and corresponding information gradient annotations of whether the normal;
[0224] Determine whether the information gradual change probability is less than a preset probability threshold, and obtain a fourth determination result;
[0225] If the fourth judgment result is no, it is determined that there is no information change of the target ship;
[0226] If the fourth judgment result is yes, it is determined that there is information change on the target ship.
[0227] It can be seen that through the above optional embodiments, it is possible to fully and comprehensively judge whether there is any information change in the target ship based on the difference between the ship information and the historical ship information and the information gradient judgment neural network model, so as to facilitate the subsequent determination of whether there is data hijacking in the target ship, so as to prevent the target ship from being hijacked by data to plan an erroneous route and cause navigation accidents, thereby improving the efficiency and accuracy of route planning, improving the safety of ship navigation, and reducing navigation accidents.
[0228] For example 3, please refer to Figure 3, Figure 3 It is another route planning system based on neural network disclosed in the embodiment of the present invention. Figure 3 The described neural network-based route planning system is applied to a data processing system / data processing device / data processing server (wherein the server includes a local processing server or a cloud processing server). Figure 3 As shown, the neural network-based route planning system may include:
[0229] A memory 301 storing executable program codes;
[0230] a processor 302 coupled to the memory 301;
[0231] The processor 302 calls the executable program code stored in the memory 301 to execute the steps of the neural network-based route planning method described in the first embodiment.
[0232] Embodiment 4, the embodiment of the present invention discloses a computer-readable storage medium, which stores a computer program for electronic data exchange, wherein the computer program enables a computer to execute the steps of the neural network-based route planning method described in embodiment 1.
[0233] Embodiment 5, the embodiment of the present invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to enable a computer to execute the steps of the neural network-based route planning method described in Embodiment 1.
[0234] The above describes specific embodiments of the present specification, and other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily have to be performed in the specific order or sequential order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0235] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.
[0236] For the convenience of description, the above device is described in various units according to their functions. Of course, when implementing this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.
[0237] Those skilled in the art will appreciate that the embodiments of this specification may be provided as methods, systems, or computer program products. Therefore, the embodiments of this specification may be in the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the embodiments of this specification may be in the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.
[0238] This specification is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of this specification. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0239] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0240] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0241] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.
[0242] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.
[0243] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.
[0244] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0245] This specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. This specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.
[0246] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.
[0247] Finally, it should be noted that the neural network-based route planning method and system disclosed in the embodiments of the present invention only disclose the preferred embodiments of the present invention, which are only used to illustrate the technical solution of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, it should be understood by those skilled in the art that the technical solutions described in the aforementioned embodiments can still be modified, or some of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A route planning method based on neural network, It is characterized in that The method comprises: In response to a route planning request of a target ship, obtaining ship information of the target ship and regional data of the ocean area where the target ship is located; the ship information includes ship type, ship position, ship historical navigation data, ship crew information and ship sensor parameters; the ship sensor parameters include ship speed, ship acceleration, ship heading, ship temperature, ship humidity and ship image data; the regional data includes obstacle area data, land boundary data, floating animal data and weather forecast data; According to the ship information, based on a similarity algorithm, multiple similar ships are screened out from multiple candidate ships, including: The satellite communication equipment of the target ship broadcasts information to all other ships within a preset radius, and obtains response information from the other ships; the response information includes the computing equipment status information of the other ships; Inputting the computing equipment status information of each of the other ships into a preset equipment status prediction algorithm model to obtain equipment status prediction parameters of each of the other ships; Determine all other ships that have sent the response information and whose equipment status prediction parameter is greater than a preset first parameter threshold as candidate ships; Calculating information similarity between the ship information of each candidate ship and the ship information of the target ship; Sorting all the candidate ships from large to small according to the information similarity to obtain a ship sequence; All the candidate ships that are a preset number of ships in the ship sequence and whose information similarity is greater than a preset first similarity threshold are determined as a plurality of similar ships; the information similarity is calculated according to the following rule: For the ship information of the candidate ship and the ship information of the target ship, a first vector distance between vectorized representations of basic ship information in the two ship information is calculated; the basic ship information includes ship type, ship position and ship crew information; Calculating a second vector distance between vectorized representations of historical navigation data of the ships in the two ship information; the historical navigation data of the ships includes the route shape of the ships in the historical time period, the navigation passing area and the nautical chart data of the passing area; calculating a third vector distance between the vectorized representations of the ship sensing parameters in the two ship information; Calculating the inverse of the weighted sum of the first vector distance, the second vector distance and the third vector distance to obtain information similarity between the ship information of the candidate ship and the ship information of the target ship; wherein the weights corresponding to the second vector distance, the third vector distance and the first vector distance decrease in sequence; According to the navigation data of the similar ships, a route planning algorithm model of the target ship is obtained by real-time training; The ship information and the area data are input into the route planning algorithm model for planning calculation to obtain the planned route of the target ship.
2. The route planning method based on neural network according to claim 1, It is characterized in that The method of obtaining the route planning algorithm model of the target ship through real-time training based on the navigation data of the similar ship includes: Acquire safe route data of each of the similar ships and corresponding navigation area chart data; the navigation area chart data includes obstacle areas and weather forecast information; Inputting each of the navigation area chart data into a preset route planning algorithm model to obtain a planned route result, and adjusting the model parameters of the route planning algorithm model according to the corresponding safe route data and the planned route result; Repeat the previous step until the route similarity between the planned route result predicted by the route planning algorithm model and the corresponding safe route data is lower than a preset second similarity threshold, so as to obtain the route planning algorithm model of the target ship.
3. The route planning method based on neural network according to claim 2, It is characterized in that The step of inputting each of the navigation area chart data into a preset route planning algorithm model to obtain a planned route result, and adjusting the model parameters of the route planning algorithm model according to the corresponding safe route data and the planned route result, comprises: The objective function is set to maximize the sum of the distances between the planned route and each obstacle area; The restriction conditions include that the distance between the planned route and any obstacle area is greater than a distance threshold and the average value of the weather danger values corresponding to the weather forecast information of all areas through which the planned route passes is less than a danger threshold; the weather danger value is obtained by predicting the weather forecast information using a pre-trained classifier model; According to the objective function and the constraint conditions, a route planning algorithm model is established based on a spatial search algorithm; Inputting the nautical chart data of each navigation area into the route planning algorithm model for iterative calculation until the planned route result is calculated; Calculating the route difference between the corresponding safe route data and the planned route result and the route corresponding to each area part; Determine the region where the route difference is greater than a first difference threshold as a high difference portion; Determine the area portion where the route difference is greater than a second difference threshold as a low difference portion; the second difference threshold is less than the first difference threshold; When the total number of all the high-difference parts is greater than a first number threshold, randomly adjusting the distance threshold to obtain a new distance threshold for next planning; When the total number of all the low-difference parts is greater than a second number threshold, the danger threshold is randomly adjusted to obtain a new danger threshold for next planning.
4. The route planning method based on neural network according to claim 1, It is characterized in that The step of inputting the ship information and the area data into the route planning algorithm model for planning calculation to obtain the planned route of the target ship includes: According to the ship information of the target ship at the current time point and the ship information stored at multiple historical time points, it is determined whether there is any information change of the target ship, and a first determination result is obtained; If the first judgment result is no, the area data is input into the route planning algorithm model to calculate the planned route of the target ship; If the first judgment result is yes, the satellite communication device of the target ship broadcasts a regional data acquisition request to all other ships within a preset radius, and acquires other regional data of the ocean area sent by all other ships; Calculating an average value of the differences between all other regional data and the regional data to obtain a difference parameter corresponding to the regional data; Determine whether the difference parameter is greater than a preset second parameter threshold, and obtain a second determination result; If the second judgment result is no, the area data is input into the route planning algorithm model to calculate the planned route of the target ship; If the second judgment result is yes, inputting each of the other area data into the route planning algorithm model to obtain the corresponding first planned route; An average value of all the first planned routes is calculated to obtain the planned route of the target ship.
5. The route planning method based on neural network according to claim 4, It is characterized in that The determining whether there is any information change of the target ship according to the ship information of the target ship at the current time point and the ship information stored at multiple historical time points includes: Determine whether the information difference between the ship information of the target ship at the current time point and the ship information at the most recent historical time point is greater than a preset third difference threshold, and obtain a third determination result; If the third judgment result is no, determining that there is no information change of the target ship; If the third judgment result is yes, the ship information sequence corresponding to the target ship is input into the trained information gradient judgment neural network model to obtain the information gradient probability corresponding to the target ship; the ship information sequence is an information sequence obtained by arranging all ship information corresponding to the target ship at the current time point and all historical time points from early to late according to the storage time points; the information gradient judgment neural network model is trained by a training data set including a plurality of training ship information sequences and corresponding information gradient annotations of whether the normal; Determine whether the information gradual change probability is less than a preset probability threshold, and obtain a fourth determination result; If the fourth judgment result is no, determining that there is no information change of the target ship; If the fourth judgment result is yes, it is determined that there is information change of the target ship.
6. A route planning system based on neural network, It is characterized in that The system comprises: an acquisition module, for acquiring, in response to a route planning request of a target ship, ship information of the target ship and regional data of the ocean area where the target ship is located; the ship information includes ship type, ship position, ship historical navigation data, ship crew information and ship sensor parameters; the ship sensor parameters include ship speed, ship acceleration, ship heading, ship temperature, ship humidity and ship image data; the regional data includes obstacle area data, land boundary data, floating animal data and weather forecast data; A screening module is used to screen out multiple similar ships from multiple candidate ships based on the ship information and a similarity algorithm, including: The satellite communication equipment of the target ship broadcasts information to all other ships within a preset radius, and obtains response information from the other ships; the response information includes the computing equipment status information of the other ships; Inputting the computing equipment status information of each of the other ships into a preset equipment status prediction algorithm model to obtain equipment status prediction parameters of each of the other ships; Determine all other ships that have sent the response information and whose equipment status prediction parameter is greater than a preset first parameter threshold as candidate ships; Calculating information similarity between the ship information of each candidate ship and the ship information of the target ship; Sorting all the candidate ships from large to small according to the information similarity to obtain a ship sequence; All the candidate ships that are a preset number of ships in the ship sequence and whose information similarity is greater than a preset first similarity threshold are determined as a plurality of similar ships; the information similarity is calculated according to the following rule: For the ship information of the candidate ship and the ship information of the target ship, a first vector distance between vectorized representations of basic ship information in the two ship information is calculated; the basic ship information includes ship type, ship position and ship crew information; Calculating a second vector distance between vectorized representations of historical navigation data of the ships in the two ship information; the historical navigation data of the ships includes the route shape of the ships in the historical time period, the navigation passing area and the nautical chart data of the passing area; calculating a third vector distance between the vectorized representations of the ship sensing parameters in the two ship information; Calculating the inverse of the weighted sum of the first vector distance, the second vector distance and the third vector distance to obtain information similarity between the ship information of the candidate ship and the ship information of the target ship; wherein the weights corresponding to the second vector distance, the third vector distance and the first vector distance decrease in sequence; A training module, used for obtaining a route planning algorithm model of the target ship through real-time training based on the navigation data of the similar ship; The planning module is used to input the ship information and the area data into the route planning algorithm model for planning calculation to obtain the planned route of the target ship.
7. A route planning system based on neural network, It is characterized in that The system comprises: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the neural network-based route planning method as described in any one of claims 1-5.
