A method for early warning of route deviation of ships passing through a lock in a key river section

Abstract

The invention discloses a route deviation warning method for ships passing through a lock in a hub river section, comprising: collecting and preprocessing inland waterway information data and historical ship data of the waterway; determining the waterway range through buffer zone analysis; determining the waterway intersection nodes and areas; constructing a ship route prediction model to predict the ship's operating route and the ship's stay time at different anchor points; planning a route that meets the lock passing plan on the basis of generating the shortest route, and obtaining a final predicted operating route; deviation judgment: comparing the ship's position information with the predicted operating route to determine whether the ship has deviated from the predicted route, and if so, giving an alarm; when the ship has an abnormal situation such as deviation, starting a warning mechanism to remind the ship's driver; optimizing the navigation path, improving navigation efficiency, reducing time and fuel waste and labor costs; reducing the possibility of accidents and improving navigation safety.

Description

Technical Field

[0001] The invention belongs to the field of inland river ship traffic, and in particular relates to a route deviation early warning method for ships passing through a lock in a hub river section. Background Art

[0002] Inland waterway shipping is an important part of my country's water transportation. However, due to the complex and changeable inland waters, turbulent water flow and numerous tributaries, it is very easy for ships that are not familiar with the environment to enter the fork and deviate from the original route, increasing the risk of accidents. The existing early warning method mainly monitors ships through manual combined with radar, cameras and other equipment, but due to the influence of environmental factors, it can generally only monitor the deviation of ships for a short time, and there is a lack of effective monitoring methods for deviations over a longer period of time. In order to ensure the shipping safety of ships passing through the locks in the hub river section, it is of great significance to use artificial intelligence, big data, edge computing and other technologies to realize automatic early warning of ships that deviate from the route.

[0003] With the development of Internet of Things technology, more and more sensors are being used in the field of water transportation, which can collect a large amount of data and process and analyze it in real time, providing more accurate services for inland waterway transportation. At the same time, machine learning methods also provide a more scientific prediction method for the field of inland waterway transportation, which can analyze and predict future ship route deviations based on historical data, and improve the accuracy and generalization of predictions. The application of these technologies will help improve the safety and efficiency of inland waterway transportation, which is of great significance to ensuring water transportation safety and economic development. Summary of the invention

[0004] In view of the problems or defects in the prior art, the present invention provides a method for warning of route deviation of ships passing through locks in a hub river section, so as to analyze and predict future ship route deviations, improve the accuracy and generalization ability of the prediction, and improve the safety and efficiency of inland waterway shipping.

[0005] In order to achieve the above-mentioned purpose, a method for early warning of route deviation of ships passing through a lock in a hub river section comprises the following steps:

[0006] Step 1: Collect inland waterway information data and historical ship data of the waterway;

[0007] Step 2: Preprocess the data collected in step 1;

[0008] Step 3: Determine the scope of the waterway through buffer zone analysis;

[0009] Step 4: Perform intersection analysis using density-based clustering algorithms to determine waterway intersection nodes and areas;

[0010] Step 5: Based on the current ship positioning information and the data obtained in step 2, a ship route prediction model is constructed to predict the ship's route and the ship's stay time at different anchor points;

[0011] Step 6: Based on the shortest route generated, combined with the actual lock-passing plan, a route that complies with the lock-passing plan is planned according to the locations of the departure port and the destination port, and the final predicted operation route is obtained;

[0012] Step 7: Deviation judgment: Compare the ship's position information with the predicted route to determine whether the ship has deviated from the predicted route. If so, an alarm is issued;

[0013] Step 8: When the ship has abnormal conditions such as deviation, the early warning mechanism is activated to promptly remind the ship driver through the message push function of the single Beidou shipborne intelligent terminal.

[0014] Furthermore, the collection of inland waterway information data includes: measuring and recording the river width, water depth, anchor points, etc. using measuring instruments; these data can be collected through existing GIS maps or nautical charts; the anchor points are usually some iconic buildings or important landmarks.

[0015] Furthermore, the collection of historical ship data of the waterway includes: obtaining information about ships that have passed through the waterway over the years through existing electronic nautical charts or satellite images, including information such as ship type, track, and cargo capacity.

[0016] Furthermore, the method of preprocessing the data collected in step 1 in step 2 includes removing abnormal and duplicate data.

[0017] Further, the method of buffer zone analysis in step 3 is:

[0018] ① Determine the scope of the buffer zone: According to the results of the determination of the waterway scope, artificially set an appropriate buffer zone radius;

[0019] ② Generate buffer area: Use the spatial buffer function to generate a buffer area with a specified radius centered on the waterway range of the Three Gorges River section, and use Euclidean buffering to calculate the area;

[0020] ③ Calculate the intersection: Using the spatial intersection function, the intersection between the ship trajectory and the buffer zone can be calculated and used as a reference standard for evaluating the ship's actions;

[0021] ④ Verify spatial relationships: Use the spatial topological relationship function (ST_Relate) to verify spatial relationships;

[0022] Furthermore, the method of using Euclidean buffer to calculate the area is:

[0023] First, the center point of the buffer zone, i.e. the target geometric object, needs to be determined; then the buffer radius r is determined according to the set buffer zone distance;

[0024] For a point object P(x,y) in a plane coordinate system, its Euclidean buffer can be expressed as: B={(x,y)| (x-x_0)²+(y-y_0)²≤r²}, where (x_0,y_0) is the coordinate of the center point;

[0025] For line and surface objects, the Euclidean buffer needs to be calculated for each vertex and the final buffer is obtained by taking the union of all buffers. For polygons, the topological correctness of the buffer needs to be checked to determine whether there is overlap or disconnection in the buffer.

[0026] Furthermore, the method for performing intersection analysis using a density-based clustering algorithm in step 4 is:

[0027] Step 1: Obtain ship trajectory data and remove abnormal and duplicate data;

[0028] Step 2: Convert the trajectory data into spatial coordinate data and add timestamp and other information;

[0029] Step 3: Calculate the distance between each point according to the formula dist(p,q) = sqrt((x_p - x_q)^2 + (y_p - y_q)^2);

[0030] Step 4: Divide all points into three categories: core points, boundary points and noise points according to the size of the neighborhood radius;

[0031] Step 5: According to the minimum number of points MinPts, the clusters formed by all core points are identified as different intersection areas;

[0032] Step 6: Further analysis and judgment are performed based on the cluster density and cluster size to determine the final intersection nodes and areas.

[0033] Furthermore, the specific method of step 5 is:

[0034] (1) Initialization: Set the distance between the starting point and all other nodes to infinity, and the distance from the starting point to itself to 0;

[0035] (2) Determine the starting point: Select the node with the smallest distance from all undetermined nodes as the current starting point and mark the node as determined;

[0036] (3) Relaxation operation: Scan all neighbor nodes of the current starting point. For each neighbor node v, calculate whether the distance from the starting point to the node is less than the known distance. If so, update the distance.

[0037] (4) Repeat steps (2) and (3) until all nodes are determined;

[0038] The specific formulas involved in the Dijkstra algorithm are as follows:

[0039] 1) Distance: used to store the distance from the starting point to each node. During initialization, the distance from the starting point to itself is set to 0, and the distances of other nodes are set to infinity; during relaxation operations, the distance value is updated by comparing the known distance with the newly calculated distance;

[0040] 2) Neighbor nodes: used to describe the relationship between nodes; for each node u, we need to traverse all its neighbor nodes v and calculate the distance from the starting point to v;

[0041] 3) Predecessor node: used to store the predecessor node of the shortest path. In the relaxation operation, if the distance from u to v is found to be shorter, the predecessor node of v is updated to u;

[0042] In the algorithm implementation, a priority queue (such as a heap) is used to maintain the node with the smallest current distance, and steps 2) and 3) are iterated until all nodes are determined; in each relaxation operation, the distance and predecessor node are recalculated and updated using the above formula; the final predecessor node can be used to reversely trace the shortest path.

[0043] Furthermore, the specific method of step 7 is:

[0044] (1) Distance calculation: The Fréchet distance is used to calculate the distance from the current position of the ship to the predicted route to assess whether the ship has deviated from the predicted route.

[0045] The Fréchet distance is a generalization of the Euclidean distance in general. It considers the difference in each dimension between two vectors, rather than just the square root of the sum of the squares of the differences in each dimension. For two vectors u and v, the Fréchet distance d(u,v) between them can be calculated according to the following formula:

[0046]

[0047] in, n is the dimension of the vector, ui and vi is a vector u and v In the idimensional coordinates; during navigation, we can regard the current position of the ship as vector u and the point on the predicted route as vector v, and evaluate whether the ship has deviated from the predicted route by calculating the Fréchet distance between them; if the distance exceeds the preset threshold, it means that the ship has deviated from the predicted route and measures need to be taken to adjust it;

[0048] (2) Deviation judgment: Based on the distance calculation results, judge whether the ship has deviated from the predicted route. The specific steps include: setting a threshold value. When the distance between the ship and the predicted route exceeds the threshold value, it is judged as deviation; according to the specific situation, the threshold value can be dynamically adjusted to improve the judgment accuracy.

[0049] Compared with the prior art, the present invention has the following effects:

[0050] (1) The hub section ship route deviation warning method of the present invention can timely detect the navigation deviation of the ship through the lock through real-time monitoring and warning, reduce the possibility of accidents, and improve navigation safety;

[0051] (2) The hub section ship route deviation warning method of the present invention, with the help of the intelligent decision support system, the ship driver can more accurately select the route, control the speed and heading, optimize the navigation path, improve the navigation efficiency, and reduce the waste of time and fuel;

[0052] (3) The route deviation warning method for ships passing through the lock in the hub river section of the present invention reduces labor costs: Compared with the traditional manual patrol and observation method, the route deviation warning of this patent can reduce the dependence on human resources and reduce labor costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 A schematic diagram of a flow chart of a preferred embodiment of the present invention;

[0054] Figure 2 It is a schematic diagram of the flow chart of yaw judgment in a preferred embodiment of the present invention;

[0055] Figure 3 A schematic diagram of the early warning process of a preferred embodiment of the present invention; DETAILED DESCRIPTION

[0056] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0057] Please refer to Figure 1The present invention relates to a method for early warning of route deviation of ships passing through a lock in a hub river section, comprising the following steps:

[0058] Step 1: Collect inland waterway information data and historical ship data of the waterway;

[0059] The collection of inland waterway information data includes: measuring and recording the river width, water depth, anchor points, etc. using measuring instruments; these data can be collected through existing GIS maps or nautical charts; the anchor points are usually some iconic buildings or important landmarks.

[0060] The collection of historical ship data of the waterway includes: obtaining information about ships that have passed through the waterway over the years through existing electronic nautical charts, satellite images or other real-time monitoring systems. These data include information such as ship type, track, cargo capacity, etc.

[0061] Step 2: Preprocess the data collected in step 1; the method for preprocessing the data collected in step 1 includes removing abnormal and duplicate data.

[0062] Step 3: Determine the scope of the waterway through buffer zone analysis; the purpose of analyzing the buffer zone is to give the scope of the waterway, that is, taking the waterway as the center, set a buffer zone of a certain distance, and define the area it contains as the scope of the entire waterway.

[0063] The specific method is:

[0064] ① Determine the scope of the buffer zone: According to the results of the determination of the waterway scope, artificially set an appropriate buffer zone radius;

[0065] ② Generate buffer area: Use the spatial buffer function to generate a buffer area with a specified radius centered on the waterway range of the Three Gorges River section, and use Euclidean buffering to calculate the area;

[0066] ③ Calculate the intersection: Using the spatial intersection function, the intersection between the ship's trajectory and the buffer zone can be calculated and used as a reference for evaluating the ship's actions; if there is an intersection, TRUE is returned, otherwise FALSE is returned. In PostGIS, the ST_Intersects function can be used to compare data types such as Point, LineString, and Polygon. Its syntax format is: ST_Intersects(geometry A, geometry B), where geometry A and B are the two geometric objects to be compared;

[0067] First, determine the target geometric object and the research area to be analyzed, and project the target object and the research area so that they are in the same coordinate system;

[0068] Use the ST_Intersects function to compare the spatial relationship between the target object and the study area and obtain the intersection result;

[0069] Based on the intersection results, determine whether the target object is inside or on the boundary of the study area, or intersects with the study area, etc.;

[0070] ④ Verify spatial relationships: Use the spatial topological relationship function (ST_Relate) to verify spatial relationships;

[0071] Determine the spatial topological relationship between two geometric objects, that is, determine whether they conform to the predetermined spatial topological rules. In PostGIS, the ST_Relate function can be used to compare data types such as Point, LineString, and Polygon. Its syntax format is: ST_Relate (geometry A, geometry B, matrix), where geometry A and B are the two geometric objects to be compared, and matrix is ​​the spatial topological relationship matrix;

[0072] First, determine the target geometric object and the research area to be analyzed, and project the target object and the research area so that they are in the same coordinate system;

[0073] Determine the spatial topological relationship matrix, which describes the spatial topological relationship between two geometric objects. For example, "TFF" means that the relationship between the target object and the study area is intersecting but not inside, where the character "*" indicates that the two geometric objects may intersect, contain or be contained, and the characters "F" and "T" indicate that the two geometric objects are unrelated or related;

[0074] Using the ST_Relate function, the spatial topological relationship between the target object and the study area is compared to obtain the corresponding spatial topological relationship matrix;

[0075] According to the obtained spatial topological relationship matrix, determine whether the spatial topological relationship between the two geometric objects conforms to the predetermined topological rules; for example, "TFF" means that the target object intersects with the study area but is not inside it, which conforms to the spatial topological rule "TFF";

[0076] ⑤ Overlay buffer: In special cases, multiple buffer areas need to be overlaid. Using the buffer overlay function, multiple buffers can be overlaid on each other to generate more complex buffer areas, and used to analyze various situations that may exist when the ship moves; commonly used buffer overlay functions include ST_Union function and ST_Multi function.

[0077] For each geometric object to be analyzed, construct their buffers. The size of the buffer is usually specified by the user, such as setting a distance threshold or buffer radius.

[0078] Use spatial overlay functions, such as ST_Intersection, ST_Union, etc., to intersect or union the buffers, so that you can get the buffers that make up the polygons and the parts that are not overlaid.

[0079] According to the results of the intersection or union operation, the spatial relationship between two or more geometric objects is analyzed. For example, in the part where two buffers intersect, the common area of ​​the two geometric objects can be analyzed; in the part where the two buffers do not intersect, the empty area between the two geometric objects can be analyzed.

[0080] Group multiple geometric objects for analysis, use the ST_Collect function to merge all geometric objects into a MULTIGEOMETRY object, and perform corresponding overlay buffer analysis.

[0081] The method of calculating the area using Euclidean buffer is:

[0082] First, the center point of the buffer zone, i.e. the target geometric object, needs to be determined; then the buffer radius r is determined according to the set buffer zone distance;

[0083] For a point object P(x,y) in a plane coordinate system, its Euclidean buffer can be expressed as: B(P,r)={(x,y)|(x-x_0)²+(y-y_0)²≤r²}, where (x_0,y_0) are the coordinates of the center point;

[0084] For line and surface objects, the Euclidean buffer needs to be calculated for each vertex and the final buffer is obtained by taking the union of all buffers. For polygons, the topological correctness of the buffer needs to be checked to determine whether there is overlap or disconnection in the buffer.

[0085] In actual operation, spatial analysis software or library functions of programming languages ​​can be used to implement the calculation of Euclidean buffers. In PostGIS, the ST_Buffer function can be used to generate Euclidean buffers. Its syntax format is: ST_Buffer(geometry, radius), where geometry is the target geometry object and radius is the buffer radius.

[0086] Step 4: Perform intersection analysis through density-based clustering algorithm to determine the channel intersection nodes and areas; in the hub river section, there are multiple tributaries that merge into the main river channel, forming a complex intersection structure; in order to accurately evaluate the driving status of ships in these areas, it is necessary to clarify the location of the intersection area on the channel.

[0087] The specific method is:

[0088] Step 1: Obtain ship trajectory data and remove abnormal and duplicate data;

[0089] Step 2: Convert the trajectory data into spatial coordinate data and add timestamp and other information;

[0090] Step 3: Calculate the distance between each point according to the formula dist(p,q) = sqrt((x_p - x_q)^2 + (y_p - y_q)^2);

[0091] Step 4: Divide all points into three categories: core points, boundary points and noise points according to the size of the neighborhood radius Eps;

[0092] Step 5: According to the minimum number of points MinPts, the clusters formed by all core points are identified as different intersection areas;

[0093] Step 6: Further analysis and judgment are performed based on the cluster density and cluster size to determine the final intersection nodes and areas.

[0094] Step 5: Based on the current ship positioning information and the data obtained in step 2, a ship route prediction model is constructed to predict the ship's route and the ship's stay time at different anchor points;

[0095] The specific method is:

[0096] (1) Initialization: Set the distance between the starting point and all other nodes to infinity, and the distance from the starting point to itself to 0;

[0097] (2) Determine the starting point: Select the node with the smallest distance from all undetermined nodes as the current starting point and mark the node as determined;

[0098] (3) Relaxation operation: Scan all neighbor nodes of the current starting point. For each neighbor node v, calculate whether the distance from the starting point to the node is less than the known distance. If so, update the distance.

[0099] (4) Repeat steps (2) and (3) until all nodes are determined;

[0100] The specific formulas involved in the Dijkstra algorithm are as follows:

[0101] 1) Distance: used to store the distance from the starting point to each node. During initialization, the distance from the starting point to itself is set to 0, and the distances of other nodes are set to infinity; during relaxation operations, the distance value is updated by comparing the known distance with the newly calculated distance;

[0102] 2) Neighbor nodes: used to describe the relationship between nodes; for each node u, we need to traverse all its neighbor nodes v and calculate the distance from the starting point to v;

[0103] 3) Predecessor node: used to store the predecessor node of the shortest path. In the relaxation operation, if the distance from u to v is found to be shorter, the predecessor node of v is updated to u;

[0104] In the algorithm implementation, a priority queue (such as a heap) is used to maintain the node with the smallest current distance, and steps 2) and 3) are iterated until all nodes are determined; in each relaxation operation, the distance and predecessor node are recalculated and updated using the above formula; the final predecessor node can be used to reversely trace the shortest path.

[0105] Step 6: Based on the shortest route generated, combined with the actual lock-passing plan, a route that complies with the lock-passing plan is planned according to the locations of the departure port and the destination port, and the final predicted operation route is obtained;

[0106] Step 7: Deviation judgment: Compare the ship's position information with the predicted route to determine whether the ship has deviated from the predicted route. If so, an alarm is issued;

[0107] The specific method is:

[0108] (1) Distance calculation: The Fréchet distance is used to calculate the distance from the current position of the ship to the predicted route to assess whether the ship has deviated from the predicted route.

[0109] The Fréchet distance is a generalization of the Euclidean distance in general. It considers the difference in each dimension between two vectors, rather than just the square root of the sum of the squares of the differences in each dimension. For two vectors u and v, the Fréchet distance d(u,v) between them can be calculated according to the following formula:

[0110]

[0111] in, n is the dimension of the vector, ui and vi is a vector u and v In the idimensional coordinates; during navigation, we can regard the current position of the ship as vector u and the point on the predicted route as vector v, and evaluate whether the ship has deviated from the predicted route by calculating the Fréchet distance between them; if the distance exceeds the preset threshold, it means that the ship has deviated from the predicted route and measures need to be taken to adjust it;

[0112] Deviation judgment: Based on the distance calculation results, judge whether the ship has deviated from the predicted route. The specific steps include: setting a threshold, when the distance between the ship and the predicted route exceeds the threshold, it is judged as deviation; according to the specific situation, the threshold can be dynamically adjusted to improve the judgment accuracy.

[0113] Step 8: When the ship has abnormal conditions such as deviation, the early warning mechanism is activated to promptly remind the ship driver through the message push function of the single Beidou shipborne intelligent terminal.

[0114] The specific early warning strategy is: according to the results of deviation detection, formulate corresponding early warning strategies, that is, remind the ship driver or trigger an alarm. According to the actual situation, early warning can be carried out in various ways such as sound, vibration, and light prompts to remind the driver to adjust the ship's operating status in time.

[0115] Message push: After the early warning mechanism is activated, the early warning information needs to be pushed to the ship driver in time so that he can respond in time. The specific steps include: sending early warning information to the driver through the message push function of the single Beidou shipborne intelligent terminal; adjusting the push method, format and content according to factors such as message type and importance.

[0116] Warning records: At the same time, warning events need to be recorded and saved for subsequent analysis and processing. Specific steps include: recording the time, location, cause and other information of the warning event; uploading the recorded data to the cloud platform or central server for subsequent analysis and processing; and performing data mining and analysis based on the recorded information to improve the accuracy and effectiveness of the warning mechanism.

[0117] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

Claims

1. A method for early warning of route deviation of ships passing through a lock in a key river section, characterized in that: The steps include: Step 1: Collect inland waterway information data and historical ship data of the waterway; Step 2: Preprocess the data collected in step 1; Step 3: Determine the scope of the waterway through buffer zone analysis; the purpose of analyzing the buffer zone is to give the scope of the waterway, that is, taking the waterway as the center, a buffer zone of a certain distance is set, and the area contained in it is defined as the scope of the entire waterway; the method of buffer zone analysis in step 3 is: ① Determine the buffer zone range: set the buffer zone radius based on the results of the channel range determination; ② Generate buffer area: Use the spatial buffer function to generate a buffer area with a specified radius centered on the waterway range of the Three Gorges River section, and use Euclidean buffering to calculate the area; ③ Calculate the intersection: Use the spatial intersection function to calculate the intersection between the ship trajectory and the buffer zone, and use it as a reference standard for evaluating the ship's actions; ④ Verify spatial relationships: Use spatial topological relationship functions to verify spatial relationships; Step 4: Perform intersection analysis using density-based clustering algorithms to determine waterway intersection nodes and areas; Step 5: Based on the current ship positioning information and the data obtained in step 2, a ship route prediction model is constructed through the Dijkstra algorithm to predict the ship's route and the ship's stay time at different anchor points; Step 6: Based on the shortest route generated, combined with the actual lock-passing plan, a route that complies with the lock-passing plan is planned according to the locations of the departure port and the destination port, and the final predicted operation route is obtained; Step 7: Deviation judgment: Compare the ship's position information with the predicted route to determine whether the ship has deviated from the predicted route. If so, an alarm is issued; Step 8: When the ship has an abnormal deviation, the early warning mechanism is activated and the ship driver is reminded in time through the message push function of the single Beidou shipborne intelligent terminal.

2. The method for early warning of route deviation of ships passing through a lock in a key river section according to claim 1 is characterized in that :The collection of inland waterway information data includes: the river width, water depth, and anchor points need to be measured and recorded using measuring instruments; these data are collected through existing GIS maps or nautical charts; the anchor points are iconic buildings or important landmarks.

3. The method for early warning of route deviation of ships passing through a lock in a hub river section according to claim 1 is characterized in that: The collection of historical ship data of the waterway includes: obtaining information on ships that have passed through the waterway over the years through existing electronic nautical charts or satellite images; these data include ship type, track, and cargo capacity information.

4. The method for early warning of route deviation of ships passing through a lock in a key river section according to claim 1 is characterized in that: The method of preprocessing the data collected in step 1 in step 2 includes removing abnormal and duplicate data.

5. The method for early warning of route deviation of ships passing through a lock in a key river section according to claim 4 is characterized in that: The method of calculating the area using Euclidean buffer is: First, the center point of the buffer zone, i.e. the target geometric object, needs to be determined; then the buffer radius r is determined according to the set buffer zone distance; For a point object P(x,y) in a plane coordinate system, its Euclidean buffer is expressed as: B={(x,y)| (x-x_0)²+(y-y_0)²≤r²}, where (x_0,y_0) are the coordinates of the center point; For line and surface objects, the Euclidean buffer needs to be calculated for each vertex separately, and all buffers are taken as the union to get the final buffer; for polygons, the topological correctness of the buffer needs to be checked to determine whether there is overlap or disconnection in the buffer.

6. The method for early warning of route deviation of ships passing through a lock in a key river section according to claim 1 is characterized in that: The method for performing intersection analysis by density-based clustering algorithm in step 4 is: Step 1: Obtain ship trajectory data and remove abnormal and duplicate data; Step 2: Convert the trajectory data into spatial coordinate data and add timestamp information; Step 3: Calculate the distance between each point according to the formula dist(p,q) = sqrt((x_p - x_q)^2 + (y_p - y_q)^2); where the coordinates of point P are (x_p, y_p) and the coordinates of point q are (x_q, y_q); Step 4: Divide all points into three categories: core points, boundary points and noise points according to the size of the neighborhood radius; Step 5: According to the minimum number of points MinPts, the clusters formed by all core points are identified as different intersection areas; Step 6: Further analysis and judgment are performed based on the cluster density and cluster size to determine the final intersection nodes and areas.

7. The method for early warning of route deviation of ships passing through a lock in a key river section according to claim 1 is characterized in that: The specific method of step 5 is: (1) Initialization: Set the distance between the starting point and all other nodes to infinity, and the distance from the starting point to itself to 0; (2) Determine the starting point: Select the node with the smallest distance from all undetermined nodes as the current starting point and mark the node as determined; (3) Relaxation operation: Scan all neighbor nodes of the current starting point. For each neighbor node v, calculate whether the distance from the starting point to the node is less than the known distance. If so, update the distance. (4) Repeat steps (2) and (3) until all nodes are determined; The specific formulas involved in the Dijkstra algorithm are as follows: 1) Distance: used to store the distance from the starting point to each node; during initialization, the distance from the starting point to itself is set to 0, and the distances of other nodes are set to infinity; during relaxation operations, the distance value is updated by comparing the known distance with the newly calculated distance; 2) Neighbor nodes: used to describe the relationship between nodes; for each node u, we need to traverse all its neighbor nodes v and calculate the distance from the starting point to v; 3) Predecessor node: used to store the predecessor node of the shortest path; in the relaxation operation, if it is found that the distance from u to v is shorter, the predecessor node of v is updated to u; In the algorithm implementation, a priority queue is used to maintain the node with the smallest current distance, and steps 2) and 3) are iterated until all nodes are determined; In each relaxation operation, the distance and predecessor nodes are recalculated and updated using the above formula; the resulting predecessor nodes are used to trace the shortest path in reverse.

8. The method for early warning of route deviation of ships passing through a lock in a key river section according to claim 1 is characterized in that: The specific method of step 7 is as follows: (1) distance calculation: using the Fréchet distance to calculate the distance between the current position of the ship and the predicted route, so as to evaluate whether the ship has deviated from the predicted route; The Fréchet distance is a generalization of the Euclidean distance in general. It considers the difference in each dimension between two vectors, rather than just the square root of the sum of the squares of the differences in each dimension. For two vectors u and v, the Fréchet distance d(u,v) between them is calculated according to the following formula: in, n is the dimension of the vector, ui and vi is a vector u and v In the i dimensional coordinates; during navigation, the current position of the ship is regarded as vector u, and the point on the predicted route is regarded as vector v. The Flechet distance between them is calculated to evaluate whether the ship has deviated from the predicted route; if the distance exceeds the preset threshold, it means that the ship has deviated from the predicted route and measures need to be taken to adjust it; (2) Deviation judgment: Based on the distance calculation results, determine whether the ship has deviated from the predicted route. The specific steps include: setting a threshold value. When the distance between the ship and the predicted route exceeds the threshold value, it is judged as deviation; dynamically adjusting the threshold value according to the specific situation to improve the judgment accuracy.

Images