Ship route and speed joint planning method
By constructing a grid map of the sea area and optimizing routes using real-time meteorological data, and combining the Theta* algorithm with route and speed optimization, the problem of the initial route being difficult to guarantee navigation safety in traditional methods has been solved, achieving safe and efficient route planning.
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- CHINA SHIP SCIENTIFIC RESEARCH CENTER
- Filing Date
- 2025-11-13
- Publication Date
- 2026-07-09
AI Technical Summary
Traditional methods for joint route and speed planning rely on initial routes recommended for ocean routes, which makes it difficult to ensure navigation safety in complex and ever-changing marine environments, resulting in the optimization of the target route and speed without guaranteeing safety.
By constructing a grid map of the sea area, using path planning algorithms and a joint optimization method for route speed, combined with real-time meteorological data and the Theta* algorithm, the initial route is optimized to avoid obstacles and severe weather, redundant turning points are eliminated, and the positions of turning points are adjusted to ensure the safety and accuracy of the route.
It improves navigation safety and efficiency, reduces computational complexity, ensures the accuracy and safety of route planning, reduces computational load, and improves computational efficiency.
Smart Images
Figure CN2025134551_09072026_PF_FP_ABST
Abstract
Description
A method for joint planning of ship routes and speeds Technical Field
[0001] This application relates to the field of intelligent ships, and in particular to a method for joint planning of ship routes and speeds. Background Technology
[0002] With the continuous development of economic globalization, the shipping industry, as the core of supply chains and international trade transportation, has had a crucial impact on the world economy. Due to the long hours spent at sea, the marine environment is complex and variable, and the energy consumption required for ship navigation is substantial. Ensuring safety throughout the voyage and reducing ship energy consumption have always been key considerations in voyage planning. The joint planning of ship routes and speeds can effectively ensure navigation safety, reduce ship energy consumption, shorten voyage time, and improve navigation efficiency, which is of great significance for promoting the development of intelligent ships and improving ship energy efficiency. Technical issues
[0003] Traditional route and speed joint planning methods use recommended ocean routes as the initial route, and then employ route and speed joint optimization methods to jointly optimize the initial route and speed to obtain the ship's target route and speed, aiming for better navigation efficiency. Currently, common route and speed joint optimization methods mainly involve establishing a multi-objective optimization model and iteratively solving the optimization results using existing technologies such as genetic algorithms. Subsequent optimization processes primarily involve further fine-tuning the initial route. Therefore, the initial route largely determines the final target route. However, the initial route determined by this method requires sufficient prior knowledge. Due to the complex and variable marine environment, the initial routes recommended by ocean routes rarely guarantee navigation safety. This leads to situations where, even after route and speed joint optimization, the final target route and speed may still not guarantee navigation safety. Technical solutions
[0004] This application addresses the aforementioned problems and technical requirements by proposing a joint planning method for ship routes and speeds. The technical solution of this application is as follows:
[0005] A method for joint planning of ship route and speed includes the following steps:
[0006] Obtain geographical information of the sea areas traversed by the ship from the starting point to the destination, and construct a grid map of the sea areas based on the geographical information.
[0007] Initialize parameter i=1 and use the grid node where the navigation start point is located as the first node;
[0008] In the i-th iteration, based on the grid map and real-time meteorological data, the path planning algorithm determines the (i+1)-th node according to the cost function of the neighboring nodes of the i-th node, and determines the i-th iteration candidate path from the starting point of navigation to the (i+1)-th node. The i-th iteration candidate path does not pass through the obstacle grid in the grid map, and the real-time meteorological data of the estimated arrival time of the ship at each position of the i-th iteration candidate path all meet the ship's passage conditions.
[0009] Enter the (i+1)th iteration until the (i+1)th node is the target point, and obtain the initial route of the ship from the starting point to the target point based on all determined nodes;
[0010] The route and speed are optimized based on the initial route using a combined route and speed optimization method to obtain the optimized route and speed of the ship from the starting point to the target point.
[0011] A further technical solution is that when the real-time wave height data of the estimated arrival time at each position of the candidate path in the i-th iteration is less than the wave height threshold, the real-time meteorological data of the estimated arrival time at the current position of the ship is determined to meet the ship's passage conditions.
[0012] Further technical solutions include: the method for joint planning of ship routes and speeds also includes:
[0013] Determine the starting time T0 when the ship departs from the starting point of its voyage;
[0014] Interpolation determines the length D of the navigation route between any position on the i-th iteration candidate path and the starting point of the navigation, and determines the estimated time ΔT=D / V for the ship to travel from the starting point to the current position along the i-th iteration candidate path, where V is the ship's speed.
[0015] The estimated arrival time of the ship at its current position on the candidate path of the i-th iteration is determined as T0+△T.
[0016] A further technical solution involves using the Theta* algorithm for path planning. This algorithm determines the (i+1)th node based on the cost function of the neighboring nodes of the i-th node, and identifies the i-th iteration candidate path from the starting point to the (i+1)-th node, including:
[0017] The p-th neighbor of the i-th node, whose cost function is ranked from smallest to largest, is selected as the p-th candidate node. The initial value of the integer parameter p is 1.
[0018] Detect whether the parent node of the p-th candidate node and the i-th node are visible; when the straight-line connection route between any two grid nodes does not pass through the obstacle grid, and the real-time meteorological data of the estimated arrival time of the ship at each position of the straight-line connection route between the two grid nodes all meet the ship's passage conditions, the two grid nodes are determined to be visible; otherwise, the two grid nodes are determined to be invisible; where the parent node of the first node is the first node itself.
[0019] When the p-th candidate node is not visible to the parent node of the i-th node, check whether the p-th candidate node and the i-th node are visible. When the p-th candidate node and the i-th node are visible, use the p-th candidate node as the (i+1)-th node and the i-th node as the parent node of the (i+1)-th node, and determine the i-th iteration candidate path, which includes the path from the starting point to the i-th node and the path from the i-th node to the (i+1)-th node. When the p-th candidate node is not visible to the i-th node, let p = p + 1 and execute the step of using the p-th neighbor node with the largest cost function among the neighbor nodes of the i-th node as the p-th candidate node again.
[0020] When the p-th candidate node is visible to the parent node of the i-th node, the p-th candidate node is taken as the (i+1)-th node, and the parent node of the i-th node is taken as the parent node of the (i+1)-th node. The i-th iteration candidate path is determined to include the path from the starting point of the voyage to the parent node of the i-th node and the path from the parent node of the i-th node to the (i+1)-th node.
[0021] The further technical solution is to construct a grid map of the sea area, which includes: dividing the sea area into grids and marking the grid status of the grid containing physical obstacles in the sea area as obstacle grids; determining the water depth data of the sea area and marking the grid status of the grids whose water depth does not meet the conditions for ship passage as obstacle grids, thereby constructing a grid map of the sea area.
[0022] A further technical solution is that constructing a grid map of the sea area also includes: performing sparsification processing on the constructed grid map, merging grids within a specified row and column range into one grid; and detecting the grid state of the grids within a specified row and column range of the grid map, wherein at least one grid in the grid state of the grid map is an obstacle grid, and the merged grid is marked as an obstacle grid.
[0023] The further technical solution is to obtain the initial route of the ship from the starting point to the target point based on all the determined nodes, including: taking the node where the target point is located as the first turning point, retrieving the parent node of each turning point in reverse as each turning point until the starting point is reached, and obtaining the initial route formed by connecting several turning points in sequence.
[0024] The route and speed joint optimization method is used to optimize the initial route, including:
[0025] Redundant turning points in the initial route are removed according to a predetermined elimination method to preprocess the initial route and obtain the route to be optimized. The route and speed joint optimization method is then used to optimize the route to be optimized.
[0026] A further technical solution is to remove redundant turning points in the initial route according to a predetermined elimination method, including: for two adjacent turning points, the kth turning point and the (k+1)th turning point, whose straight-line distance does not meet the distance requirement on the initial route, if the kth turning point and the (k+2)th turning point are visible to each other, the (k+1)th turning point is eliminated and the kth turning point and the (k+2)th turning point are directly connected; or, if the (k-1)th turning point and the (k+1)th turning point are visible to each other, the kth turning point is eliminated and the (k-1)th turning point and the (k+1)th turning point are directly connected.
[0027] The further technical solution is to optimize the route based on the route to be optimized by using the route and speed joint optimization method. This includes connecting the m turning points of the route to be optimized on the same great circle route. The great circle route is the great circle arc navigation curve of the ship from the starting point of navigation to the target point. The great circle route between two adjacent turning points is divided into n reference points at equal distances, and n turning points to be optimized are obtained based on the position of the reference points.
[0028] When using the route and speed joint optimization method to optimize the route to be optimized, m turning points and n turning points to be optimized are adjusted to obtain the optimized route.
[0029] The further technical solution is to obtain n turning points to be optimized based on the position of the reference point, including: according to the angle bisector principle, with the reference point as the foot of the perpendicular, the reference point is moved a predetermined distance along the direction perpendicular to the great circle route to obtain n turning points to be optimized. Beneficial effects
[0030] This application proposes a method for joint planning of ship routes and speeds. During the initial route planning process, real-time weather data for both obstacles and the ship's journey to each location along the path are simultaneously monitored. This ensures that the shortest initial route from the starting point to the target point accurately avoids obstacles and severe weather such as typhoons, significantly improving navigation safety and laying a solid foundation for subsequent joint optimization of routes and speeds.
[0031] A detailed grid map was constructed based on ship data, and the actual water depth data was modeled as an obstacle grid, which accurately screened out shallow water areas and ensured that dangerous areas with shallow water could be effectively avoided when planning navigation routes.
[0032] By discretizing the continuous space of the real-world environment map into a grid map, the computational complexity is simplified. Furthermore, by sparsifying the grid map and reducing the grid density, the computational load of the algorithm can be effectively reduced, enabling efficient computation for joint optimization of flight routes and speeds.
[0033] Redundant turning points in the initial route are removed using a predetermined elimination method to obtain the route to be optimized, reducing the complexity of the route. Simultaneously, it ensures that the distance between adjacent turning points in the route is not too close, which is crucial for the safety and accuracy of navigation. Based on this, a route-speed joint optimization algorithm is used to adjust the turning points of the route to be optimized, accurately calculating the direction and range of movement for each turning point, ensuring the accuracy, smoothness, and safety of the route's turning. Attached Figure Description
[0034] Figure 1 is a flowchart of the joint optimization method for ship route and speed.
[0035] Figure 2 shows two path examples of the Theta* algorithm.
[0036] Figure 3 is an example diagram of the Theta* algorithm searching for the initial route.
[0037] Figure 4 shows the initial flight path with redundant nodes.
[0038] Figure 5 is a schematic diagram of adjusting the turning point of the route to be optimized. Embodiments of the present invention
[0039] The specific embodiments of this application will be further described below with reference to the accompanying drawings.
[0040] This application discloses a method for joint planning of ship route and speed. Please refer to the flowchart shown in Figure 1. This method for joint optimization of ship route and speed includes:
[0041] Step 1: Obtain geographical information of the sea areas traversed by the ship from the starting point to the destination through world ocean route data, including water depth data and physical obstacles, and construct a grid map of the sea areas based on the geographical information.
[0042] Geographical information includes at least the coverage area of physical obstacles within the sea area. Physical obstacles include islands, reefs, and military restricted areas within the sea area, which are impassable. The sea area traversed by the ship is divided into grids using a grid mapping method. The grids containing physical obstacles such as islands, reefs, and military restricted areas are marked as obstacle grids, while the grids of other grids are marked as navigable grids.
[0043] In another embodiment, the geographic condition information also includes water depth data at different locations within the sea area. Based on the water depth data, grids whose water depth does not meet the vessel passage requirements are marked as obstacle grids. When the water depth of a grid is less than the vessel's draft requirement, it is determined that the water depth of that grid does not meet the vessel passage requirements. By consulting the vessel's navigation report, relevant vessel information can be determined, including vessel type, cargo load, displacement, and draft, thereby determining the vessel's draft requirement. This step accurately filters out physical obstacles and shallow water areas in the navigation area and models them as obstacles, ensuring that these dangerous areas can be effectively avoided when planning routes.
[0044] Considering the impact of data density on computational efficiency in environmental modeling, and given that the distance between adjacent grid nodes in the existing grid map is only 1.4221 nautical miles, the grid density in the existing grid map is too high. This causes the path planning algorithm to consume a significant amount of computation time when calculating the initial route. To improve the computational efficiency of the path planning algorithm while ensuring the safety of route planning, in one embodiment, the constructed grid map is further sparsified by merging grids within a specified row and column range into one grid, and the grid status within the specified row and column range of the grid map is detected. When at least one grid within the specified row and column range of the grid map is an obstacle grid, the merged grid is marked as an obstacle grid; otherwise, the merged grid is marked as a navigable grid. For example, obstacle grids in the grid map are marked as 1, navigable grids are marked as 0, and 9 grids in every 3 rows and 3 columns are merged into one grid. The marking values of the 9 grids in every 3 rows and 3 columns are detected; if any of these 9 grids is marked as 1, the merged grid is marked as 1. This method ensures that no obstacle area is overlooked, reducing computational load while maintaining accurate obstacle identification, thus improving the efficiency and safety of flight path planning. The number of grid rows and columns included in the specified row and column range can be adjusted according to actual needs.
[0045] Step 2: Initialize parameter i=1 and use the grid node where the starting point of the voyage is located as the first node.
[0046] Step 3: In the i-th iteration, based on the grid map and real-time meteorological data, a path planning algorithm is used to determine the (i+1)-th node according to the cost function of the neighboring nodes of the i-th node, and to determine the i-th iteration candidate path from the starting point to the (i+1)-th node. The path planning algorithm can be any existing path planning algorithm. When searching for the (i+1)-th node, the selected i-th iteration candidate path does not pass through the obstacle grid in the grid map, and the real-time meteorological data for the estimated arrival time of the ship at each location on the i-th iteration candidate path all meet the ship's passage conditions.
[0047] When searching for the candidate path in the i-th iteration, it is necessary to estimate the arrival time of the ship at each position on the candidate path in the i-th iteration. In one embodiment, the method for estimating the arrival time is as follows:
[0048] Determine the starting time T0 when the ship departs from the navigation starting point. Interpolate the candidate path for the i-th iteration, using linear interpolation. Determine the length D of the navigation route between any position on the i-th iteration candidate path and the navigation starting point, and determine the estimated time ΔT = D / V for the ship to travel from the navigation starting point to the current position along the i-th iteration candidate path, where V is the ship's speed. Determine the estimated arrival time of the ship at the current position on the i-th iteration candidate path as T0 + ΔT.
[0049] In one embodiment, real-time wave height data of the estimated arrival time at the current location is read. When the wave height data is less than the wave height threshold, it is determined that the real-time meteorological data of the estimated arrival time at the current location meets the passage conditions of the ship. Here, the wave height threshold is set according to the actual safety requirements of the ship's navigation.
[0050] Step 4: Enter the (i+1)th iteration and repeat the iterative steps of searching for candidate paths in Step 3 until the (i+1)th node is the target point. Based on all the nodes determined by the iteration process, the initial route of the ship from the starting point to the target point is obtained.
[0051] Step 5: Optimize the initial route using the route and speed joint optimization method to obtain the optimized route and speed of the ship from the starting point to the target point.
[0052] The route and speed joint optimization method can construct a multi-objective optimization model that includes the total route length, total sailing time, total sailing fuel consumption, and total sailing cost. Then, the nodes of the initial route and the ship speed of each segment in the initial route are input into the multi-objective optimization model, and optimization algorithms such as particle swarm optimization and genetic optimization are used to solve the problem. The optimized route and optimized speed can be obtained. This part can use existing route and speed joint optimization solution methods, which will not be elaborated on in this embodiment.
[0053] In another embodiment, the path planning algorithm used in step 3 is the Theta* algorithm. Based on the standard Theta* algorithm, this application obtains an improved dynamic Theta* algorithm by real-time detection of meteorological data. The specific steps of using the improved dynamic Theta* algorithm to determine the (i+1)th node based on the cost function of the neighboring nodes of the i-th node, and to determine the i-th iteration candidate path from the starting point to the (i+1)th node, include:
[0054] The p-th neighbor of the i-th node, whose cost function is ranked from smallest to largest, is selected as the p-th candidate node. The initial value of the integer parameter p is 1.
[0055] Detect whether the parent node of the p-th candidate node and the i-th node are visible; when the straight connection route between any two grid nodes does not pass through the obstacle grid, and the real-time meteorological data of the estimated arrival time of the ship at each position of the straight connection route between the two grid nodes all meet the ship's passage conditions, the two grid nodes are determined to be visible; otherwise, the two grid nodes are determined to be invisible; where the parent node of the first node is the first node itself.
[0056] When the p-th candidate node is not visible to the parent node of the i-th node, check whether the p-th candidate node and the i-th node are visible. When the p-th candidate node and the i-th node are visible, take the p-th candidate node as the (i+1)-th node and the i-th node as the parent node of the (i+1)-th node, and determine the i-th iteration candidate path, which includes the path from the starting point of the flight to the i-th node and the path from the i-th node to the (i+1)-th node. When the p-th candidate node is not visible to the i-th node, let p = p+1 and execute the step of taking the p-th neighbor node with the largest cost function among the neighbor nodes of the i-th node as the p-th candidate node again.
[0057] When the p-th candidate node is visible to the parent node of the i-th node, the p-th candidate node is taken as the (i+1)-th node, and the parent node of the i-th node is taken as the parent node of the (i+1)-th node. The i-th iteration candidate path is determined to include the path from the starting point of the voyage to the parent node of the i-th node and the path from the parent node of the i-th node to the (i+1)-th node.
[0058] In this application, when detecting whether the straight-line connection route from the ship's starting point to the (i+1)th node s' is a candidate path for the i-th iteration, since the visible connection route from the ship's starting point to the parent node of the i-th node has already been obtained when detecting the (i-1)th iteration candidate path, it is only necessary to accumulate the visible connection route from the parent node of the i-th node to the (i+1)th node into the (i-1)th iteration candidate path to obtain the i-th iteration candidate path. When detecting the visible connection route from the parent node of the i-th node to the (i+1)th node, the Theta* algorithm compares the cost functions of two different paths from the parent node of the i-th node to the (i+1)th node. As shown in Figure 2, the cost function of the path from the parent node of the i-th node s' is calculated. The distance cost g(s) to the i-th node s and the straight-line lengths between the i-th node s and the (i+1)-th node s' Thus, the cost function of Path1 is obtained. This corresponds to the path Path1 shown by the dashed lines in Figures 2(a) and 2(b). Starting from the parent node of the i-th node s... Length of the straight line to the (i+1)th node s' Thus, the cost function of Path2 is obtained. This corresponds to the path Path2 shown by the solid line in Figures 2(a) and 2(b).
[0059] According to the triangle inequality, the length of Path2 is less than or equal to the length of Path1, but Path1 is a visible path, while Path2 needs to detect the parent node of the i-th node s. Is the (i+1)th node s' visible? As shown in Figure 2(a), if the parent node of the i-th node s is visible... If Path2 is visible to the (i+1)th node s', it means that Path2 is a visible path, and the Theta* algorithm will preferentially select Path2 as the i-th iteration candidate path for the ship from the starting point of the voyage to the (i+1)th node s'. As shown in Figure 2(b), if the parent node of the i-th node s... If the i+1th node s' is not visible, then the Theta* algorithm will select Path1 as the candidate path for the i-th iteration.
[0060] Figure 3 illustrates an example of the Theta* algorithm searching for an initial route. The g(s) value of each node and its parent node are marked at the grid nodes in the figure. In the example, the horizontal coordinates (1,2,3,4,5) can be understood as equally spaced longitudes, and the vertical coordinates (A,B,C) represent different dimensions. Hollow circles represent the current node, and dashed arrows represent nodes generated during the expansion process. The Theta* algorithm first expands outward from the starting point A4 as the current node and sets the parent node of its visible neighbors to A4 itself, as shown in Figure 3(a). The requirement for visibility is that the line connecting the neighbor node and its parent node A4 does not pass through obstacles or wave height areas exceeding a set threshold. If the f(s) value of a visible neighbor node is smaller than its original value, it is added to the open list (A3, A5, B3, B4, B5), and the current node A4 is added to the closed list. In the next iteration, the node B3 with the smallest f(s) value is selected from the open list as the current node to expand its neighbors, as shown in Figure 3(b). Node B2 is a visible neighbor of the current node B3. It is not visible to B3's parent node A4, so Theta* updates it according to Path 1 in Figure 2(b) and sets its parent node to B3. On the other hand, nodes C2, C3, and C4 are also visible neighbors of the current node B3, but they are visible to B3's parent node A4. Therefore, Theta* updates them according to Path 2 in Figure 2(a) and sets their parent node to A4. Because A4 is in the closed list, and the f(s) values of A3 and B4 are greater than or equal to the values corresponding to the previous iteration, the other visible neighbors of the current node B3, A4, A3, and B4, are not updated. Similar to the previous step, the updated B2, C2, C3, and C4 are added to the open list, and the current node B3 is added to the closed list. Afterwards, B2 expands its visible neighbors as the current node, as shown in Figure 3(c). A1 and A2 are visible neighbors of B2, but not visible to B2's parent node B3. Therefore, Theta* updates them according to Path 1 in Figure 2(b). Since B1 and C1 are visible to B3, Theta* updates them according to Path 2 in Figure 2(a). Finally, C1, as the node with the smallest f(s) value in the open list, is selected as the current node. Thus, the Theta* algorithm has found a feasible path, as shown in Figure 3(d).
[0061] In one embodiment, each node and its parent node on the initial path are obtained based on the improved dynamic Theta* algorithm. Based on all the determined nodes, the initial route of the ship from the starting point to the target point is obtained, including: taking the node where the target point is located as the first turning point, retrieving the parent node of each turning point in reverse as each turning point until the starting point is reached, and obtaining an initial route formed by connecting several turning points in sequence.
[0062] The initial flight path obtained based on the improved dynamic Theta* algorithm often contains redundant turning points that are too close together, as shown in Figure 4. and These redundant turning points not only increase the complexity of the navigation route but also reduce computational efficiency. Therefore, in one embodiment, redundant turning points in the initial route are first removed according to a predetermined removal method to preprocess the initial route and obtain the route to be optimized. Then, the route and speed joint optimization method is used to optimize the route based on the route to be optimized, which can improve the efficiency of route and speed optimization. In one embodiment, removing redundant turning points in the initial route according to the predetermined removal method includes: for the k-th turning point among two adjacent turning points whose straight-line distance does not meet the distance requirement on the initial route. and the (k+1)th turning point The distance requirement here is set according to actual navigation needs. When the k-th turning point... With the (k+2)th turning point When points are visible to each other, remove the (k+1)th turning point P. k +1 And directly connect the kth turning point and the (k+2)th turning point Or, at the (k-1)th turning point With the (k+1)th turning point When points are visible to each other, remove the k-th turning point. And directly connect to the (k-1)th turning point and the (k+1)th turning point Among them, the turning point It is along the direction from the starting point of the voyage to the target point and the turning point. The next turning point directly connected, turning point It is along the direction from the starting point of the voyage to the target point and the turning point. The previous turning point that is directly connected.
[0063] In the initial route, besides some overly dense turning points, there are also some overly sparse turning points. Although the preprocessing method removes redundant points from the route to be optimized, the remaining nodes may still be too sparse. This can lead to excessively large turning angles when the ship turns at each node, making turning difficult. Moreover, overly sparse turning points will reduce the accuracy of subsequent route speed optimization results. Therefore, before using route speed joint optimization to optimize the route to be optimized, further processing of the turning points of the route to be optimized is required.
[0064] In one embodiment, as shown in Figure 5, the m turning points of the route to be optimized are connected on the same great circle route, where the m turning points are P1, P2, ..., P...m A great circle course is a great circle arc navigation curve for a ship from its starting point to its destination. The great circle course between two adjacent turning points is divided into n reference points B1, B2, ..., B... n Based on the angle bisector principle, taking the reference point as the foot of the perpendicular, the reference point is moved a predetermined distance along a direction perpendicular to the great circle route, resulting in n turning points x1, x2, ..., xn to be optimized. n When using the route and speed joint optimization method to optimize a route to be optimized, m turning points and n turning points to be optimized are adjusted to obtain the optimized route. The speeds of m+n-1 segments are adjusted to obtain the optimized speed.
[0065] The above descriptions are merely preferred embodiments of this application, and this application is not limited to the above embodiments. It is understood that other improvements and variations that can be directly derived or conceived by those skilled in the art without departing from the spirit and concept of this application should be considered to be included within the protection scope of this application.
Claims
1. A method for joint planning of ship route and speed, characterized in that, The joint planning method for ship routes and speeds includes: Obtain geographical condition information of the sea area traversed by the ship from the starting point to the destination, and construct a grid map of the sea area based on the geographical condition information; Initialize parameter i=1 and use the grid node where the navigation start point is located as the first node; In the i-th iteration, based on the grid map and real-time meteorological data, a path planning algorithm is used to determine the (i+1)-th node according to the cost function of the neighboring nodes of the i-th node, and to determine the i-th iteration candidate path from the starting point to the (i+1)-th node. The i-th iteration candidate path does not pass through the obstacle grid in the grid map, and the real-time meteorological data of the estimated arrival time of the ship at each position of the i-th iteration candidate path all meet the passage conditions of the ship. Enter the (i+1)th iteration until the (i+1)th node is the target point, and obtain the initial route of the ship from the starting point to the target point based on all the determined nodes; The route and speed are optimized based on the initial route using a combined route and speed optimization method to obtain the optimized route and speed of the ship from the starting point to the target point.
2. The method for joint planning of ship routes and speeds according to claim 1, characterized in that, When the real-time wave height data of the estimated arrival time of the vessel at each position of the candidate path in the i-th iteration is less than the wave height threshold, it is determined that the real-time meteorological data of the estimated arrival time of the vessel at the current position meets the passage conditions of the vessel.
3. The method for joint planning of ship routes and speeds according to claim 2, characterized in that, The joint planning method for ship routes and speeds also includes: Determine the starting time of the ship's departure from the navigation point. ； Interpolation determines the length D of the navigation route between any position on the i-th iteration candidate path and the starting point of the navigation, and determines the estimated time ΔT=D / V for the ship to travel from the starting point to the current position along the i-th iteration candidate path, where V is the ship's speed. The estimated arrival time of the ship at its current position on the candidate path for the i-th iteration is determined as follows: +△T。 4. The method for joint planning of ship routes and speeds according to claim 1, characterized in that, The path planning algorithm used is the Theta* algorithm. The process of using this algorithm to determine the (i+1)th node based on the cost function of the neighboring nodes of the i-th node, and determining the i-th iteration candidate path from the starting point to the (i+1)-th node, includes: The p-th neighbor of the i-th node, whose cost function is ranked from smallest to largest, is selected as the p-th candidate node. The initial value of the integer parameter p is 1. Detect whether the parent node of the p-th candidate node and the i-th node are visible; when the straight-line connection route between any two grid nodes does not pass through the obstacle grid, and the real-time meteorological data of the estimated arrival time of the ship at each position of the straight-line connection route between the two grid nodes all meet the passage conditions of the ship, the two grid nodes are determined to be visible; otherwise, the two grid nodes are determined to be invisible; wherein, the parent node of the first node is the first node itself. When the p-th candidate node is not visible to the parent node of the i-th node, check whether the p-th candidate node and the i-th node are visible; when the p-th candidate node and the i-th node are visible, use the p-th candidate node as the (i+1)-th node and the i-th node as the parent node of the (i+1)-th node, and determine that the i-th iteration candidate path includes the path from the starting point to the i-th node and the path from the i-th node to the (i+1)-th node; when the p-th candidate node is not visible to the i-th node, let p = p + 1 and execute the step of using the p-th neighbor node with the largest cost function among the neighbor nodes of the i-th node as the p-th candidate node again. When the p-th candidate node is visible to the parent node of the i-th node, the p-th candidate node is taken as the (i+1)-th node, and the parent node of the i-th node is taken as the parent node of the (i+1)-th node. The i-th iteration candidate path is determined to include the path from the starting point of the voyage to the parent node of the i-th node and the path from the parent node of the i-th node to the (i+1)-th node.
5. The method for joint optimization of ship route and speed according to claim 1, characterized in that, The construction of the grid map of the sea area includes: The sea area is divided into grids, and the grid state of the grid containing the physical obstacles in the sea area is marked as an obstacle grid; Determine the water depth data of the sea area, mark the grid state of grids whose water depth does not meet the conditions for ship passage as obstacle grids, and construct a grid map of the sea area.
6. The method for joint planning of ship routes and speeds according to claim 5, characterized in that, The construction of the grid map of the sea area also includes: The constructed grid map is subjected to sparsification processing, merging grids within a specified row and column range into one grid; and the grid state of the grids within the specified row and column range of the grid map is detected, and at least one grid in the specified row and column range of the grid map is an obstacle grid, and the merged grid is marked as an obstacle grid.
7. The method for joint planning of ship routes and speeds according to claim 4, characterized in that, Based on all determined nodes, the initial route of the ship from the starting point to the destination is obtained, including: Using the node where the target point is located as the first turning point, the parent node of each turning point is retrieved in reverse order as each turning point until the starting point of the voyage is reached, thus obtaining the initial route formed by connecting several turning points in sequence. The method of optimizing the route and speed together based on the initial route includes: Redundant turning points in the initial route are removed according to a predetermined elimination method to preprocess the initial route and obtain the route to be optimized. The route and speed joint optimization method is then used to optimize the route to be optimized.
8. The method for joint planning of ship routes and speeds according to claim 7, characterized in that, Redundant turning points in the initial route are removed according to a predetermined removal method, including: For the kth turning point and the (k+1)th turning point, which are adjacent turning points whose straight-line distance on the initial route does not meet the distance requirement, if the kth turning point and the (k+2)th turning point are visible to each other, the (k+1)th turning point is discarded and the kth turning point and the (k+2)th turning point are directly connected; or, if the (k-1)th turning point and the (k+1)th turning point are visible to each other, the kth turning point is discarded and the (k-1)th turning point and the (k+1)th turning point are directly connected.
9. The method for joint planning of ship routes and speeds according to claim 7, characterized in that, The method of optimizing routes and speeds together optimizes the route to be optimized, including: Connect the m turning points of the route to be optimized on the same great circle route. The great circle route is the great circle arc navigation curve of the ship from the starting point of the voyage to the target point. Divide the great circle route between two adjacent turning points into n reference points at equal distances, and obtain n turning points to be optimized based on the positions of the reference points. When optimizing the route based on the route to be optimized using the route and speed joint optimization method, the m turning points and n turning points to be optimized are adjusted to obtain the optimized route.
10. The method for joint planning of ship routes and speeds according to claim 9, characterized in that, The n turning points to be optimized based on the position of the reference point include: Based on the angle bisector principle, with the reference point as the foot of the perpendicular, the reference point is moved a predetermined distance along a direction perpendicular to the great circle route to obtain n turning points to be optimized.