Intelligent navigation system for ship

Abstract

The invention discloses an intelligent navigation system for a ship, and the system comprises an obtaining module which is used for obtaining the navigation data of the ship within a preset time period in the navigation process of the current route; the preprocessing module is used for preprocessing the navigation data to obtain preprocessed navigation data; the evaluation module is used for inputting the preprocessed navigation data into a pre-trained route evaluation model to obtain a first evaluation risk value; the optimization module is used for comparing the first assessment risk value with a preset assessment threshold value and judging whether to adjust and optimize the current route according to a comparison result; by acquiring data of the ship in the sailing process in real time and preprocessing the data, the system can find potential sailing risks in time, such as channel congestion and meteorological condition deterioration; based on evaluation of the data, the system can intelligently adjust the ship route and avoid a dangerous area, so that the safety of ship navigation is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent navigation of ships, and in particular to an intelligent navigation system for ships. Background Art

[0002] During the navigation process of a ship, in order to improve the safety of the ship's route, reduce the increase in the ship's navigation time caused by marine environmental problems and the hidden dangers to personal and property safety during navigation, the navigation route will be planned based on experience before sailing; however, the ocean route is long and the routes along the way are complex and changeable. When the existing ships are sailing, they use satellites to locate the ship's route. Due to the low efficiency of analyzing changes in the marine environment, it is easy for the ship to encounter safety hazards during the navigation process due to changes in the marine environment. During the navigation process of the ship, the ship cannot make timely assessments of changes in the marine environment, and cannot make timely adjustments to the route before facing changes in the marine environment, resulting in damage to the ship itself, affecting its normal passage, and causing losses to personal and property safety during the navigation of the ship. Therefore, an intelligent navigation system for ships is urgently needed to solve such problems. Summary of the invention

[0003] The present invention aims to solve at least one of the technical problems in the above-mentioned technologies to a certain extent. To this end, the purpose of the present invention is to propose an intelligent navigation system for ships, which can timely detect potential navigation risks, such as channel congestion, deterioration of weather conditions, etc., by acquiring data of ships in real time during navigation and preprocessing them; based on the evaluation of these data, the system can intelligently adjust the route to avoid dangerous areas, thereby significantly improving the safety of ship navigation.

[0004] To achieve the above object, an embodiment of the present invention provides an intelligent navigation system for a ship, comprising:

[0005] An acquisition module is used to acquire the navigation data of the ship during the current route navigation within a preset time period;

[0006] A preprocessing module, used for preprocessing the navigation data to obtain preprocessed navigation data;

[0007] An evaluation module, used for inputting the preprocessed navigation data into a pre-trained route evaluation model to obtain a first evaluation risk value;

[0008] The optimization module is used to compare the first evaluation risk value with a preset evaluation threshold value, and determine whether to adjust and optimize the current route according to the comparison result.

[0009] Preferably, the preprocessing module comprises:

[0010] An outlier processing submodule is used to identify outliers in the navigation data and remove the outliers to obtain initial data;

[0011] The data cleaning submodule is used to clean the initial data based on the preset data cleaning rules to obtain the final data;

[0012] A submodule is determined for taking the final data as the preprocessed navigation data.

[0013] Preferably, the outlier processing submodule includes:

[0014] Grouping unit for:

[0015] Grouping the navigation data to obtain a plurality of navigation data groups;

[0016] Take any navigation data set as the target data set;

[0017] Sorting the values ​​in the target data group based on the time series to obtain a sorted target data group; and evenly dividing the sorted target data group into a plurality of target sub-data groups;

[0018] Any target sub-data group is selected as the first data group, and any group of data groups adjacent to the first data group is selected as the second data group;

[0019] A first computing unit is configured to:

[0020] Calculate the absolute value of the difference between any two adjacent data in the first data group to obtain a plurality of first difference values;

[0021] Calculate the absolute value of the difference between any two adjacent data in the second data group to obtain a plurality of second difference values;

[0022] Calculating the Pearson correlation coefficients between the first difference values ​​and the second difference values ​​at the corresponding positions;

[0023] Compare the Pearson correlation coefficient with a preset correlation coefficient threshold, and take the first difference and the second difference corresponding to when the Pearson correlation coefficient is greater than or equal to the preset correlation coefficient threshold as abnormal differences; and mark the value corresponding to the abnormal difference as abnormal;

[0024] Traverse each target sub-data group and obtain the number of abnormal markings corresponding to each value;

[0025] Compare the number of abnormal markings corresponding to each value with the preset abnormal number threshold, and take the value when the number of abnormal markings corresponding to the value is greater than or equal to the preset abnormal number threshold as the first abnormal value;

[0026] Calculate the median of all values ​​in the target data set except the first outlier;

[0027] Calculating the ratio of each first outlier value to the median, and when it is determined that the ratio is less than a preset ratio threshold, taking the first outlier value corresponding to the ratio as the second outlier value;

[0028] Calculate the mean of all values ​​in the target data set except the first outlier;

[0029] Calculate the absolute difference between each second outlier and the mean, and use the ratio of the absolute difference to the mean as the deviation coefficient of the second outlier;

[0030] The second outlier value with a deviation coefficient greater than a preset deviation threshold is used as the final outlier value of the target data group;

[0031] The outlier elimination unit is used to eliminate the final outlier value.

[0032] Preferably, the method for obtaining the route evaluation model includes:

[0033] Obtain historical navigation data of the ship during navigation;

[0034] Standardizing the historical navigation data and using the standardized historical navigation data as sample data;

[0035] Extracting features from the sample data to obtain a sample training feature set;

[0036] Inputting the sample training feature set into a preset neural network model for iterative training, and when the model converges, determining the model algorithm to obtain an initial route evaluation model;

[0037] The cross-validation method is used to optimize the initial route evaluation model to obtain the optimized route evaluation model.

[0038] Preferably, the model algorithm comprises:

[0039]

[0040] Wherein, T represents the risk assessment value of the ship sailing on the route; α represents the average risk value of other historical ships in the preset time period of the route; n represents the total number of categories of risk factors in the navigation data of the ship's route navigation process; b represents the influence coefficient of the risk factors in the navigation data of the ship's route navigation process on the navigation of the route; when the risk factors in the navigation data of the navigation process have an impact on the navigation of the route, b takes a value of 1; when the risk factors in the navigation data of the navigation process have no impact on the navigation of the route, b takes a value of 0; Q represents the risk probability of the risk factors in the corresponding navigation data of the navigation process bringing to the navigation of the route when the risk factors in the navigation data of the navigation process have an impact on the navigation of the route; M i represents the risk value of the i-th risk factor; f(x)dx represents the probability density function; t represents the starting time within the preset time period.

[0041] Preferably, the optimization module includes:

[0042] A first judgment submodule, used for comparing the first assessment risk value with a preset assessment threshold;

[0043] A first optimization submodule, configured to, if the first assessed risk value is greater than or equal to a preset assessment threshold, allow the ship to continue sailing based on the current route;

[0044] The second optimization submodule is used to:

[0045] If the first assessed risk value is less than a preset assessment threshold, obtaining information of a plurality of backup routes and obtaining second assessed risk values ​​corresponding to the plurality of backup routes, and when it is determined that the maximum value of the second assessed risk value is greater than or equal to the preset assessment threshold, taking the backup route corresponding to the maximum value of the second assessed risk value as the target route;

[0046] The current route of the ship is adjusted and optimized based on the target route.

[0047] Preferably, it also includes: a navigation collision avoidance module, which is used to determine the collision avoidance strategy of the current ship according to the navigation data, and control the current ship to execute the collision avoidance strategy;

[0048] The navigation collision avoidance module comprises:

[0049] The second judgment submodule is used to judge based on the navigation data and determine the temporary obstacles and their types on the current route;

[0050] A first strategy submodule, used for generating a first collision avoidance strategy when determining that the category of the temporary obstacle is a static target, and controlling the current ship to execute the first collision avoidance strategy;

[0051] The second strategy submodule is used to generate a second collision avoidance strategy when it is determined that the category of the temporary obstacle is a dynamic target, and control the current ship to execute the second collision avoidance strategy.

[0052] Preferably, the first strategy submodule includes:

[0053] The first strategy generating unit is used for obtaining, when it is determined that the category of the temporary obstacle is a static target, a line connecting the current position of the current ship and the tangent points on the left and right sides of the static obstacle to be avoided, and recording the line as a collision avoidance route;

[0054] The first strategy control unit is used to control the current ship to make a strategy adjustment to the current route based on the collision avoidance route.

[0055] Preferably, the second strategy submodule includes:

[0056] An acquisition unit, used to acquire navigation data of all dynamic obstacles of the current ship in a preset navigation area; the dynamic obstacles are other target ships;

[0057] The second computing unit is configured to:

[0058] Calculate the collision risk between the current ship and other target ships to obtain several first collision risks;

[0059] The plurality of first collision risk degrees are compared with a preset collision risk threshold value respectively; if the first collision risk degree is less than the preset collision risk threshold value, a target ship corresponding to the first collision risk degree being less than the preset collision risk threshold value is taken as a first target ship, to obtain a plurality of first target ships;

[0060] If the first collision risk is greater than or equal to a preset collision risk threshold, taking a target ship corresponding to the case where the first collision risk is greater than or equal to the preset collision risk threshold as a second target ship, and obtaining a plurality of second target ships;

[0061] The third computing unit is configured to:

[0062] respectively calculating the second collision risk between any two of the plurality of second target ships to obtain a plurality of second collision risk;

[0063] Calculating the difference between the plurality of second collision risk levels, taking the second target ship corresponding to the difference being less than or equal to a preset difference threshold as the third target ship, and taking the preset area where the third target ship is located as the first prohibited navigation area for the current ship;

[0064] The fourth computing unit is configured to:

[0065] respectively calculating the third collision risk between other second target ships except the third target ship and a plurality of first target ships;

[0066] The second target ship and a plurality of first target ships corresponding to the third collision risk level being greater than or equal to a preset collision risk level threshold are used as fourth target ships, and the preset area where the fourth target ship is located is used as the second prohibited navigation area for the current ship;

[0067] The second strategy generating unit is used to take the area other than the first prohibited navigation area and the second prohibited navigation area in the preset navigation area as a safe area;

[0068] The second strategy control unit is used to control the current ship to adjust the strategy of the current route based on the safety area.

[0069] Preferably, it also includes: a verification unit, which is used to:

[0070] After the second strategy control unit controls the current ship to sail for a preset time after the strategy adjustment of the current route based on the safety area, calculates a fourth collision risk between the current ship and the first target ship in the preset navigation area of ​​the route after the strategy adjustment;

[0071] Determine whether the maximum value of the fourth collision risk is less than or equal to the preset collision risk threshold. If so, the collision avoidance verification is passed, and the current ship continues to sail based on the route after the strategy adjustment; if not, the route strategy is re-adjusted until the collision avoidance verification is passed.

[0072] The present invention discloses an intelligent navigation system for ships. By acquiring and preprocessing navigation data and then inputting these data into a pre-trained route evaluation model, the system can accurately evaluate the potential risks of the ship on the current route. When the first evaluation risk value evaluated exceeds the preset evaluation threshold, the system can issue a warning in time and adjust the route, thereby effectively avoiding the risks that may be encountered during navigation and significantly improving the safety of ship navigation; the system processes and evaluates the navigation data in an automated and intelligent manner, reducing the need for human intervention. This not only improves the speed of decision-making, but also enhances the scientificity and accuracy of decision-making. The captain and crew can make more reasonable navigation decisions based on the evaluation results and optimization suggestions provided by the system; the optimization module intelligently determines whether the current route needs to be adjusted based on the evaluation results and the preset evaluation threshold. This dynamic adjustment mechanism enables the ship to select the optimal navigation path according to the real-time navigation environment and conditions, thereby saving navigation time, reducing fuel consumption, and improving navigation efficiency; by optimizing route selection and reducing unnecessary navigation distance and waiting time, the system can significantly reduce the fuel consumption and operating costs of the ship. This has important economic significance for shipping companies.

[0073] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0074] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0076] Figure 1 is a block diagram of an intelligent navigation system for a ship according to an embodiment of the present invention;

[0077] Figure 2 is a block diagram of a preprocessing module according to one embodiment of the present invention;

[0078] Figure 3 4 is a block diagram of an outlier processing submodule according to an embodiment of the present invention. DETAILED DESCRIPTION

[0079] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0080] Example 1

[0081] like Figure 1 As shown, an intelligent navigation system for a ship comprises:

[0082] An acquisition module is used to acquire the navigation data of the ship during the current route navigation within a preset time period;

[0083] A preprocessing module, used for preprocessing the navigation data to obtain preprocessed navigation data;

[0084] An evaluation module, used for inputting the preprocessed navigation data into a pre-trained route evaluation model to obtain a first evaluation risk value;

[0085] The optimization module is used to compare the first evaluation risk value with a preset evaluation threshold value, and determine whether to adjust and optimize the current route according to the comparison result.

[0086] In this embodiment, the navigation data is information obtained by acquiring, analyzing and integrating multi-source data during the ship's navigation process through ship-side sensing equipment using advanced sensing technology and sensor information fusion technology.

[0087] In this embodiment, the navigation data includes: environmental information, ship motion information, and temporary obstruction information; wherein the environmental information includes: wind speed, wind direction, wave height, and water depth; the ship motion information includes: the position of the ship, the speed of the ship, and the heading of the ship; the temporary obstruction information includes: the direction and distance of sea and seabed obstacles relative to the ship, and the static and dynamic information of other ships; the static information of other ships includes: length, width, ship type, and target port; the dynamic information of other ships includes: position, speed, and heading; and the other ships are all ships other than the ship in navigation.

[0088] In this embodiment, the preprocessing includes identifying and eliminating outliers and data cleaning.

[0089] In this embodiment, the pre-trained route evaluation model is a route evaluation model that has been trained in advance.

[0090] The beneficial effect of the above technical solution is that by acquiring and preprocessing navigation data in real time, and then inputting these data into the pre-trained route evaluation model, the system can accurately evaluate the potential risks of the ship on the current route. When the first evaluation risk value evaluated exceeds the preset evaluation threshold, the system can issue a warning in time and adjust the route, thereby effectively avoiding the risks that may be encountered during navigation and significantly improving the safety of ship navigation; the system processes and evaluates navigation data in an automated and intelligent manner, reducing the need for human intervention. This not only improves the speed of decision-making, but also enhances the scientificity and accuracy of decision-making. The captain and crew can make more reasonable navigation decisions based on the evaluation results and optimization suggestions provided by the system; the optimization module intelligently determines whether the current route needs to be adjusted based on the evaluation results and the preset evaluation threshold. This dynamic adjustment mechanism enables the ship to select the optimal navigation path according to the real-time navigation environment and conditions, thereby saving navigation time, reducing fuel consumption, and improving navigation efficiency; by optimizing route selection and reducing unnecessary navigation distance and waiting time, the system can significantly reduce the fuel consumption and operating costs of the ship. This is of great economic significance to shipping companies.

[0091] Example 2

[0092] like Figure 2 As shown, the preprocessing module includes:

[0093] An outlier processing submodule is used to identify outliers in the navigation data and remove the outliers to obtain initial data;

[0094] The data cleaning submodule is used to clean the initial data based on the preset data cleaning rules to obtain the final data;

[0095] A submodule is determined for taking the final data as the preprocessed navigation data.

[0096] In this embodiment, the preset data cleaning rules include identifying and processing duplicate data, data formatting, data verification and data consistency checking, and data screening and conversion.

[0097] The beneficial effects of the above technical solution are: through outlier processing and data cleaning, inaccurate or invalid data are effectively removed, ensuring that subsequent analysis and evaluation are based on reliable information, thereby improving the correctness of the entire system decision; clean and accurate data can make the system run more stable and reduce errors and instability caused by abnormal data; high-quality data is input into the route evaluation model to obtain more accurate evaluation results, which is conducive to making more reasonable route optimization decisions.

[0098] Example 3

[0099] like Figure 3 As shown, the outlier processing submodule includes:

[0100] Grouping unit for:

[0101] Grouping the navigation data to obtain a plurality of navigation data groups;

[0102] Take any navigation data set as the target data set;

[0103] Sorting the values ​​in the target data group based on the time series to obtain a sorted target data group; and evenly dividing the sorted target data group into a plurality of target sub-data groups;

[0104] Any target sub-data group is selected as the first data group, and any group of data groups adjacent to the first data group is selected as the second data group;

[0105] A first computing unit is configured to:

[0106] Calculate the absolute value of the difference between any two adjacent data in the first data group to obtain a plurality of first difference values;

[0107] Calculate the absolute value of the difference between any two adjacent data in the second data group to obtain a plurality of second difference values;

[0108] Calculating the Pearson correlation coefficients between the first difference values ​​and the second difference values ​​at the corresponding positions;

[0109] Compare the Pearson correlation coefficient with a preset correlation coefficient threshold, and take the first difference and the second difference corresponding to when the Pearson correlation coefficient is greater than or equal to the preset correlation coefficient threshold as abnormal differences; and mark the value corresponding to the abnormal difference as abnormal;

[0110] Traverse each target sub-data group and obtain the number of abnormal markings corresponding to each value;

[0111] Compare the number of abnormal markings corresponding to each value with the preset abnormal number threshold, and take the value when the number of abnormal markings corresponding to the value is greater than or equal to the preset abnormal number threshold as the first abnormal value;

[0112] Calculate the median of all values ​​in the target data set except the first outlier;

[0113] Calculating the ratio of each first outlier value to the median, and when it is determined that the ratio is less than a preset ratio threshold, taking the first outlier value corresponding to the ratio as the second outlier value;

[0114] Calculate the mean of all values ​​in the target data set except the first outlier;

[0115] Calculate the absolute difference between each second outlier and the mean, and use the ratio of the absolute difference to the mean as the deviation coefficient of the second outlier;

[0116] The second outlier value with a deviation coefficient greater than a preset deviation threshold is used as the final outlier value of the target data group;

[0117] The outlier elimination unit is used to eliminate the final outlier value.

[0118] In this embodiment, the navigation data may be grouped according to the data types in the navigation data.

[0119] The working principle of the above calculation scheme is as follows: Assume that the target data group after sorting based on the time series is ABCDEFGHI JKL; Assume that the sorted target data group is evenly divided into three target sub-data groups, and three target sub-data groups ABCD, EFGH, and IJKL are obtained; any target sub-data group is taken as the first data group, and here it is assumed that the EFGH target sub-data group is taken as the first data group; the second data group is "ABCD" "I JKL", assuming that "ABCD" is taken as the second data group; the first difference is |EF|, |FG|, |GH|; similarly, the second difference is |AB|, |BC|, |CD|; calculate the Pearson correlation coefficient between a number of first differences and the three second differences at the corresponding positions, that is, calculate the Pearson correlation coefficient between |FG| and |BC|, the Pearson correlation coefficient between |EF| and |AB|, and the Pearson correlation coefficient between |GH| and |CD|; compare the Pearson correlation coefficient with the preset correlation coefficient threshold, when the Pearson correlation coefficient is greater than or equal to the preset correlation coefficient threshold, it means that the value corresponding to the abnormal difference is likely to be abnormal; assuming that the Pearson correlation coefficient between |FG| and |BC| is greater than the preset correlation coefficient threshold, F, G, B, and C are marked as abnormal;

[0120] Traverse each target data group, here we need to calculate the first difference corresponding to the first data group "EFGH" and the second difference corresponding to the second data group "I JKL"; at this time, calculate the Pearson correlation coefficient between several first differences and the three second differences at the corresponding positions; that is, calculate the Pearson correlation coefficient between |FG| and |JK|, the Pearson correlation coefficient between |EF| and |IJ|, and the Pearson correlation coefficient between |GH| and |KL|; compare the Pearson correlation coefficient with the preset correlation coefficient threshold, when the Pearson correlation coefficient is greater than or equal to the preset correlation coefficient threshold, it means that the value corresponding to the abnormal difference is likely to be abnormal; assuming that the Pearson correlation coefficient between |FG| and |JK| is greater than the preset correlation coefficient threshold, F, G, J, K are marked as abnormal;

[0121] The number of abnormal markings corresponding to each value is F marked 2 times, G marked 2 times, B, C, J, K marked once each; the preset abnormal number threshold is 2, then F and G are used as the first abnormal data; the median of other values ​​in the target data group except FG is obtained, and the ratio of each first abnormal value to the median is calculated. When it is determined that the ratio is less than the preset ratio threshold, the first abnormal value corresponding to the ratio is used as the second abnormal value; the mean of all values ​​in the target data group except the first abnormal value is calculated; the absolute difference between each second abnormal value and the mean is calculated, and the ratio of the absolute difference to the mean is used as the deviation coefficient of the second abnormal value; the second abnormal value whose deviation coefficient is greater than the preset deviation threshold is used as the final abnormal value of the target data group; the abnormal value elimination unit is used to eliminate the final abnormal value.

[0122] The beneficial effects of the above technical solution are: through detailed grouping, calculation and comparison steps, outliers in navigation data can be accurately found, thereby improving the accuracy of outlier identification; comprehensive consideration of multiple factors and statistical indicators, such as the Pearson correlation coefficient, the number of abnormal markings, etc., can effectively reduce the misjudgment of normal fluctuations and ensure that only truly abnormal data are eliminated; it can adapt to different data distributions and characteristics and can work effectively in various complex data situations; after eliminating outliers, data that can better reflect the actual situation can be obtained, providing a high-quality foundation for subsequent preprocessing and evaluation.

[0123] Example 4

[0124] The method for obtaining the route evaluation model includes:

[0125] Obtain historical navigation data of the ship during navigation;

[0126] Standardizing the historical navigation data and using the standardized historical navigation data as sample data;

[0127] Extracting features from the sample data to obtain a sample training feature set;

[0128] Inputting the sample training feature set into a preset neural network model for iterative training, and when the model converges, determining the model algorithm to obtain an initial route evaluation model;

[0129] The cross-validation method is used to optimize the initial route evaluation model to obtain the optimized route evaluation model.

[0130] The beneficial effects of the above technical solution are: ensuring that the data has a uniform scale and specification, making subsequent model training more effective and improving model performance and generalization ability; helping to highlight important data features so that the model can better understand and process data; being able to continuously optimize model parameters through iterative training to better fit the data and improve the accuracy of route evaluation; finding a suitable algorithm to improve the operational efficiency and accuracy of the model; enhancing the reliability and stability of the model, reducing the risk of overfitting, and making the model more accurate in predicting new data; providing scientific and accurate data support and evaluation methods for ship route planning and decision-making.

[0131] Example 5

[0132] The model algorithm comprises:

[0133]

[0134] Wherein, T represents the risk assessment value of the ship sailing on the route; α represents the average risk value of other historical ships in the preset time period of the route; n represents the total number of categories of risk factors in the navigation data of the ship's route navigation process; b represents the influence coefficient of the risk factors in the navigation data of the ship's route navigation process on the navigation of the route; when the risk factors in the navigation data of the navigation process have an impact on the navigation of the route, b takes a value of 1; when the risk factors in the navigation data of the navigation process have no impact on the navigation of the route, b takes a value of 0; Q represents the risk probability of the risk factors in the corresponding navigation data of the navigation process bringing to the navigation of the route when the risk factors in the navigation data of the navigation process have an impact on the navigation of the route; M i represents the risk value of the i-th risk factor; f(x)dx represents the probability density function; t represents the starting time within the preset time period.

[0135] The beneficial effect of the above technical solution is that the risk assessment value of a ship sailing on a specific route can be calculated more accurately through a clear formula, providing a quantitative standard for risk assessment; it takes into account a variety of factors, such as the average risk value, the number of risk factor categories, the impact coefficient, the risk probability and the risk value of each risk factor, so as to make the assessment more comprehensive; because it involves variables such as preset time periods, the risk can be dynamically assessed according to different times and situations to adapt to changes in actual conditions; based on rigorous mathematical models and calculation formulas, the risk assessment has a certain degree of scientificity and rationality.

[0136] Example 6

[0137] Optimization modules, including:

[0138] A first judgment submodule, used for comparing the first assessment risk value with a preset assessment threshold;

[0139] A first optimization submodule, configured to, if the first assessed risk value is greater than or equal to a preset assessment threshold, allow the ship to continue sailing based on the current route;

[0140] The second optimization submodule is used to:

[0141] If the first assessed risk value is less than a preset assessment threshold, obtaining information of a plurality of backup routes and obtaining second assessed risk values ​​corresponding to the plurality of backup routes, and when it is determined that the maximum value of the second assessed risk value is greater than or equal to the preset assessment threshold, taking the backup route corresponding to the maximum value of the second assessed risk value as the target route;

[0142] The current route of the ship is adjusted and optimized based on the target route.

[0143] The beneficial effects of the above technical solution are: continuously monitoring the risk status through the first judgment submodule to detect changes in risk situations in a timely manner; ensuring that the ship continues to sail along the current route when the first assessed risk value is large to avoid confusion caused by unnecessary route changes, and evaluating alternative routes when the risk value is small to find safer routes, thereby improving overall navigation safety; reasonably selecting alternative routes and optimizing them to avoid waste of resources and meaningless route adjustments; being able to flexibly adjust routes according to different risk conditions to adapt to various complex situations; obtaining alternative route information and evaluations in advance to improve decision-making efficiency and timeliness; and optimizing route selection to help reduce costs and improve efficiency.

[0144] Example 7

[0145] It also includes: a navigation collision avoidance module, which is used to determine the collision avoidance strategy of the current ship according to the navigation data, and control the current ship to execute the collision avoidance strategy;

[0146] The navigation collision avoidance module comprises:

[0147] The second judgment submodule is used to judge based on the navigation data and determine the temporary obstacles and their types on the current route;

[0148] A first strategy submodule, used for generating a first collision avoidance strategy when determining that the category of the temporary obstacle is a static target, and controlling the current ship to execute the first collision avoidance strategy;

[0149] The second strategy submodule is used to generate a second collision avoidance strategy when it is determined that the category of the temporary obstacle is a dynamic target, and control the current ship to execute the second collision avoidance strategy.

[0150] The beneficial effects of the above technical solution are: through a special navigation collision avoidance module, various possible collision hazards are handled in a timely and effective manner, greatly improving the safety of ship navigation; static targets and dynamic targets are distinguished and corresponding strategies are formulated to make response measures more accurate and adapt to the characteristics of different types of obstacles; different sub-modules quickly make judgments and generate strategies, which improves decision-making efficiency and timeliness and reduces reaction time; targeted collision avoidance strategies effectively reduce the probability of collision with temporary obstacles and reduce accidents; can flexibly generate appropriate strategies according to different situations to better adapt to complex and changeable navigation environments; timely and effective avoidance of obstacles ensures smooth navigation of ships and reduces unnecessary pauses and interference.

[0151] Example 8

[0152] The first strategy submodule includes:

[0153] The first strategy generating unit is used for obtaining, when it is determined that the category of the temporary obstacle is a static target, a line connecting the current position of the current ship and the tangent points on the left and right sides of the static obstacle to be avoided, and recording the line as a collision avoidance route;

[0154] The first strategy control unit is used to control the current ship to make a strategy adjustment to the current route based on the collision avoidance route.

[0155] The beneficial effects of the above technical solution are: it can accurately generate collision avoidance routes for static targets and accurately avoid static obstacles; quickly adjust the strategy of the current route to reduce the possibility of collision with static targets; formulate collision avoidance strategies based on accurate positions and connections to make decisions more scientific; effectively reduce the risk of collision with static obstacles and ensure the safe navigation of ships; improve the processing ability and reaction speed of ships when encountering static obstacles; avoid potential losses and hazards caused by collisions with static objects; help maintain good navigation order and reduce chaos caused by static obstacles.

[0156] Example 9

[0157] The second strategy submodule includes:

[0158] An acquisition unit, used to acquire navigation data of all dynamic obstacles of the current ship in a preset navigation area; the dynamic obstacles are other target ships;

[0159] The second computing unit is configured to:

[0160] Calculate the collision risk between the current ship and other target ships to obtain several first collision risks;

[0161] The plurality of first collision risk degrees are compared with a preset collision risk threshold value respectively; if the first collision risk degree is less than the preset collision risk threshold value, a target ship corresponding to the first collision risk degree being less than the preset collision risk threshold value is taken as a first target ship, to obtain a plurality of first target ships;

[0162] If the first collision risk is greater than or equal to a preset collision risk threshold, taking a target ship corresponding to the case where the first collision risk is greater than or equal to the preset collision risk threshold as a second target ship, and obtaining a plurality of second target ships;

[0163] The third computing unit is configured to:

[0164] respectively calculating the second collision risk between any two of the plurality of second target ships to obtain a plurality of second collision risk;

[0165] Calculating the difference between the plurality of second collision risk levels, taking the second target ship corresponding to the difference being less than or equal to a preset difference threshold as the third target ship, and taking the preset area where the third target ship is located as the first prohibited navigation area for the current ship;

[0166] The fourth computing unit is configured to:

[0167] respectively calculating the third collision risk between other second target ships except the third target ship and a plurality of first target ships;

[0168] The second target ship and a plurality of first target ships corresponding to the third collision risk level being greater than or equal to a preset collision risk level threshold are used as fourth target ships, and the preset area where the fourth target ship is located is used as the second prohibited navigation area for the current ship;

[0169] The second strategy generating unit is used to take the area other than the first prohibited navigation area and the second prohibited navigation area in the preset navigation area as a safe area;

[0170] The second strategy control unit is used to control the current ship to adjust the strategy of the current route based on the safety area.

[0171] In this embodiment, the specific calculation method of the collision risk is not limited here, which belongs to the common knowledge in the field and will not be elaborated here.

[0172] The working principle of the above technical solution is: assuming that the current ship is A, and the other target ships are BCDEFG; assuming that the collision risk between A and BCDE is greater than or equal to the preset collision risk threshold, BCDE is used as the second target ship; at this time, the danger of these four ships to the current ship A is relatively large, and the collision risks among the four ships, namely, the collision risks between BC, CD, BD, BE, CE and DE, are calculated to obtain 6 second collision risks; the difference between the 6 second collision risks is calculated, and the second target ship corresponding to the difference when the difference is less than or equal to the preset difference threshold is used as the third target ship; assuming that the difference between the collision risk between BC and the collision risk between CD is less than or equal to the preset difference threshold, the BCD ships are classified into one category, and here the main purpose is to classify dangerous ships; the BCD ships are used as the third target ship. target ship; the preset area where the third target ship is located is used as the first prohibited navigation area of ​​the current ship; the second target ships include ship E in addition to the third target ship, that is, the third collision risk between the second target ship E and the first target ship FG is calculated; it is assumed that the third collision risk between the second target ship E and the first target ship F is greater than or equal to the preset collision risk threshold, because the second target ship E itself is dangerous to the current ship A, and there is a danger between E and the first target ship F, in order to prevent A from participating in the danger between E and F, the preset area where E is located is set as the second prohibited navigation area; in this way, the danger to ship A can be minimized, and the area other than the first prohibited navigation area and the second prohibited navigation area in the preset navigation area is used as a safe area; the second strategy control unit controls the current ship to adjust the strategy of the current route based on the safe area.

[0173] The beneficial effects of the above technical solution are: it carefully considers the situations of various dynamic target ships and their mutual relationships, and comprehensively evaluates the collision risk; accurately classifies target ships into categories with different degrees of danger, so as to facilitate the formulation of targeted strategies; by setting no-fly zones and determining safe areas, the probability of collision with other dynamic obstacles can be greatly reduced; strategies are generated based on detailed calculations and analysis to make decisions more scientific and reasonable; it can adapt to complex multi-ship navigation environments and flexibly adjust ship routes; reduce unnecessary pauses or confusion caused by avoiding other ships, and ensure smooth navigation.

[0174] Example 10

[0175] Also includes: a verification unit, for:

[0176] After the second strategy control unit controls the current ship to sail for a preset time after the strategy adjustment of the current route based on the safety area, calculates a fourth collision risk between the current ship and the first target ship in the preset navigation area of ​​the route after the strategy adjustment;

[0177] Determine whether the maximum value of the fourth collision risk is less than or equal to the preset collision risk threshold. If so, the collision avoidance verification is passed, and the current ship continues to sail based on the route after the strategy adjustment; if not, the route strategy is re-adjusted until the collision avoidance verification is passed.

[0178] The beneficial effects of the above technical scheme are: it realizes continuous verification and monitoring of the effect of strategy adjustment to ensure the effectiveness of the collision avoidance strategy; it can promptly discover problems and make adjustments according to actual conditions to keep the route in a relatively safe state; it improves the reliability and stability of the entire collision avoidance scheme through repeated verification; it better adapts to new situations and changes that may arise in dynamic obstacles during navigation; it further reduces the risk of collision and ensures the high safety of the ship during the entire navigation process; it can quickly discover and correct possible unreasonable adjustments to improve the accuracy of the strategy.

[0179] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An intelligent navigation system for a ship, characterized in that: include: An acquisition module is used to acquire the navigation data of the ship during the current route navigation within a preset time period; A preprocessing module, used for preprocessing the navigation data to obtain preprocessed navigation data; An evaluation module, used for inputting the preprocessed navigation data into a pre-trained route evaluation model to obtain a first evaluation risk value; The optimization module is used to compare the first evaluation risk value with a preset evaluation threshold value, and determine whether to adjust and optimize the current route according to the comparison result.

2. The intelligent navigation system for ships according to claim 1, characterized in that: The preprocessing module comprises: An outlier processing submodule is used to identify outliers in the navigation data and remove the outliers to obtain initial data; The data cleaning submodule is used to clean the initial data based on the preset data cleaning rules to obtain the final data; A submodule is determined for taking the final data as the preprocessed navigation data.

3. The intelligent navigation system for ships according to claim 2, characterized in that: The outlier processing submodule includes: Grouping unit for: Grouping the navigation data to obtain a plurality of navigation data groups; Take any navigation data set as the target data set; Sorting the values ​​in the target data group based on the time series to obtain a sorted target data group; and evenly dividing the sorted target data group into a plurality of target sub-data groups; Any target sub-data group is selected as the first data group, and any group of data groups adjacent to the first data group is selected as the second data group; The first computing unit is configured to: Calculate the absolute value of the difference between any two adjacent data in the first data group to obtain a plurality of first difference values; Calculate the absolute value of the difference between any two adjacent data in the second data group to obtain a plurality of second difference values; Calculating the Pearson correlation coefficients between the first difference values ​​and the second difference values ​​at the corresponding positions; Compare the Pearson correlation coefficient with a preset correlation coefficient threshold, and take the first difference and the second difference corresponding to when the Pearson correlation coefficient is greater than or equal to the preset correlation coefficient threshold as abnormal differences; and mark the value corresponding to the abnormal difference as abnormal; Traverse each target sub-data group and obtain the number of abnormal markings corresponding to each value; Compare the number of abnormal markings corresponding to each value with the preset abnormal number threshold, and take the value when the number of abnormal markings corresponding to the value is greater than or equal to the preset abnormal number threshold as the first abnormal value; Calculate the median of all values ​​in the target data set except the first outlier; Calculating the ratio of each first outlier value to the median, and when it is determined that the ratio is less than a preset ratio threshold, taking the first outlier value corresponding to the ratio as the second outlier value; Calculate the mean of all values ​​in the target data set except the first outlier; Calculate the absolute difference between each second outlier and the mean, and use the ratio of the absolute difference to the mean as the deviation coefficient of the second outlier; The second outlier value with a deviation coefficient greater than a preset deviation threshold is used as the final outlier value of the target data group; The outlier elimination unit is used to eliminate the final outlier value.

4. The intelligent navigation system for ships according to claim 1, characterized in that: The method for obtaining the route evaluation model includes: Obtain historical navigation data of the ship during navigation; Standardizing the historical navigation data and using the standardized historical navigation data as sample data; Extracting features from the sample data to obtain a sample training feature set; Inputting the sample training feature set into a preset neural network model for iterative training, and when the model converges, determining the model algorithm to obtain an initial route evaluation model; The cross-validation method is used to optimize the initial route evaluation model to obtain the optimized route evaluation model.

5. The intelligent navigation system for ships according to claim 4, characterized in that: The model algorithm comprises: Wherein, T represents the risk assessment value of the ship sailing on the route; α represents the average risk value of other historical ships in the preset time period of the route; m represents the total number of categories of risk factors in the navigation data of the ship's route navigation process; b represents the influence coefficient of the risk factors in the navigation data of the ship's route navigation process on the navigation of the route; when the risk factors in the navigation data of the navigation process have an impact on the navigation of the route, b takes a value of 1; when the risk factors in the navigation data of the navigation process have no impact on the navigation of the route, b takes a value of 0; Q represents the risk probability of the risk factors in the corresponding navigation data of the navigation process bringing to the navigation of the route when the risk factors in the navigation data of the navigation process have an impact on the navigation of the route; M i represents the risk value of the i-th risk factor; f(x)dx represents the probability density function; t represents the starting time within the preset time period.

6. The intelligent navigation system for ships according to claim 1, characterized in that: Optimization modules, including: A first judgment submodule, used for comparing the first assessment risk value with a preset assessment threshold; A first optimization submodule, configured to, if the first assessed risk value is greater than or equal to a preset assessment threshold, allow the ship to continue sailing based on the current route; The second optimization submodule is used to: If the first assessed risk value is less than a preset assessment threshold, obtaining information of a plurality of backup routes and obtaining second assessed risk values ​​corresponding to the plurality of backup routes, and when it is determined that the maximum value of the second assessed risk value is greater than or equal to the preset assessment threshold, taking the backup route corresponding to the maximum value of the second assessed risk value as the target route; The current route of the ship is adjusted and optimized based on the target route.

7. The intelligent navigation system for ships according to claim 1, characterized in that: Also includes: The navigation and collision avoidance module is used to determine the collision avoidance strategy of the current ship according to the navigation data and control the current ship to execute the collision avoidance strategy; The navigation collision avoidance module comprises: The second judgment submodule is used to judge based on the navigation data and determine the temporary obstacles and their types on the current route; A first strategy submodule, used for generating a first collision avoidance strategy when determining that the category of the temporary obstacle is a static target, and controlling the current ship to execute the first collision avoidance strategy; The second strategy submodule is used to generate a second collision avoidance strategy when it is determined that the category of the temporary obstacle is a dynamic target, and control the current ship to execute the second collision avoidance strategy.

8. The intelligent navigation system for ships according to claim 7, characterized in that: The first strategy submodule includes: The first strategy generating unit is used for obtaining, when it is determined that the category of the temporary obstacle is a static target, a line connecting the current position of the current ship and the tangent points on the left and right sides of the static obstacle to be avoided, and recording the line as a collision avoidance route; The first strategy control unit is used to control the current ship to make a strategy adjustment to the current route based on the collision avoidance route.

9. The intelligent navigation system for ships according to claim 7, characterized in that: The second strategy submodule includes: An acquisition unit, used to acquire navigation data of all dynamic obstacles of the current ship in a preset navigation area; the dynamic obstacles are other target ships; The second computing unit is configured to: Calculate the collision risk between the current ship and other target ships to obtain several first collision risks; The plurality of first collision risk degrees are compared with a preset collision risk threshold value respectively; if the first collision risk degree is less than the preset collision risk threshold value, a target ship corresponding to the first collision risk degree being less than the preset collision risk threshold value is taken as a first target ship, to obtain a plurality of first target ships; If the first collision risk is greater than or equal to a preset collision risk threshold, taking a target ship corresponding to the case where the first collision risk is greater than or equal to the preset collision risk threshold as a second target ship, and obtaining a plurality of second target ships; The third computing unit is configured to: respectively calculating the second collision risk between any two of the plurality of second target ships to obtain a plurality of second collision risk; Calculating the difference between the plurality of second collision risk levels, taking the second target ship corresponding to the difference being less than or equal to a preset difference threshold as the third target ship, and taking the preset area where the third target ship is located as the first prohibited navigation area for the current ship; The fourth computing unit is configured to: respectively calculating the third collision risk between other second target ships except the third target ship and a plurality of first target ships; The second target ship and a plurality of first target ships corresponding to the third collision risk level being greater than or equal to a preset collision risk level threshold are used as fourth target ships, and the preset area where the fourth target ship is located is used as the second prohibited navigation area for the current ship; The second strategy generating unit is used to take the area other than the first prohibited navigation area and the second prohibited navigation area in the preset navigation area as a safe area; The second strategy control unit is used to control the current ship to adjust the strategy of the current route based on the safety area.

10. The intelligent navigation system for ships according to claim 9, characterized in that: Also includes: Verification unit for: After the second strategy control unit controls the current ship to sail for a preset time after the strategy adjustment of the current route based on the safety area, calculates a fourth collision risk between the current ship and the first target ship in the preset navigation area of ​​the route after the strategy adjustment; Determine whether the maximum value of the fourth collision risk is less than or equal to the preset collision risk threshold. If so, the collision avoidance verification is passed, and the current ship continues to sail based on the route after the strategy adjustment; if not, the route strategy is re-adjusted until the collision avoidance verification is passed.

Images