Dynamic Optimization Method for Meteorological Routes Based on A* Algorithm

Abstract

The present invention discloses a method for dynamic optimization of weather routes based on the A-star algorithm, comprising: obtaining a recommended weather route for the current voyage, and translating the recommended weather route to both sides by a first distance threshold to obtain a recommended route belt; obtaining a weather data grid node; generating a navigation grid node with obstacles and wind, wave, and current with obstacle attributes in the recommended route belt; obtaining a weather route from the departure port position to the destination port position based on the A-star algorithm; determining the optimal weather route for the current voyage; calculating the ship position when the weather update time threshold is passed starting from the departure port; obtaining all ship positions in the optimal weather route, thereby obtaining the entire optimal weather route. The present invention utilizes forecasted weather data to continuously update the navigation environment, continuously optimize the remaining segments, and ultimately obtain the optimal route, thereby maximizing the use of weather conditions to minimize the total cost of ship operation and promote the healthy development of smart shipping.

Description

Technical Field

[0001] The present invention relates to the field of intelligent ship ocean route optimization, and in particular to a weather route dynamic optimization method based on an A-star algorithm. Background Art

[0002] Autonomous navigation, a key feature of smart ships, is gaining increasing attention from both industrial and maritime companies. Furthermore, fuel costs and carbon emission policies pose numerous challenges to the development of smart ships. As we all know, fuel consumption is both a major cost and a primary source of pollutant emissions for ships. Research and application of route optimization is an effective way to reduce fuel consumption.

[0003] Route optimization has different definitions depending on the optimization objective. Simonsen et al. define route optimization as the process of finding the best route for a voyage based on parameters such as the expected time of arrival (ETA), turning points, and speed for each leg, based on meteorological data and the ship's own performance. In a sense, route optimization is the selection of the best route for a given voyage (with a known departure and destination port) taking into account expected weather and sea conditions. The optimality of a route depends on the optimization objective, such as minimizing sailing distance, sailing time, or fuel consumption.

[0004] A large number of theoretical achievements have been accumulated at home and abroad for the route optimization problem of intelligent ships. The main methods used to solve route optimization problems include isochrone method, dynamic programming method, graph search algorithm, intelligent algorithm, artificial intelligence and machine learning methods. These research methods have their unique advantages in dealing with optimization problems, but there are still four problems that hinder the process of converting theoretical achievements into practical applications. Insufficient consideration of meteorological conditions. Some studies do not consider the influence of ocean currents or waves. Insufficient consideration of meteorological conditions will inevitably lead to cumulative errors in the predicted speed, ship position and estimated arrival time, reducing the practicality of the optimized route; dynamic optimization performance is insufficient. The duration of a transoceanic voyage often exceeds the time range of the weather forecast. Therefore, it is necessary to continuously update the meteorological data during the voyage to perform dynamic optimization of the route. Summary of the Invention

[0005] The present invention provides a weather route dynamic optimization method based on the A-star algorithm to overcome the above technical problems.

[0006] In order to achieve the above object, the technical solution of the present invention is:

[0007] The weather route dynamic optimization method based on the A-star algorithm includes the following steps:

[0008] S1: Obtain a recommended climate route for the current voyage, and shift the recommended climate route to both sides by a first distance threshold to obtain a recommended route belt;

[0009] S2: Based on the predicted meteorological data and the longitude and latitude grid, a meteorological data grid including the entire voyage within a meteorological update time threshold is obtained to obtain a meteorological data grid node; the predicted meteorological data includes wind direction, wind speed, wave height, swell direction, swell height, ocean current direction, and ocean current speed;

[0010] S3: generating navigation grid nodes with obstacles and wind, wave and current with obstacle attributes in the recommended route belt;

[0011] S4: Based on the navigation grid nodes and the meteorological data grid nodes, based on the A-star algorithm, obtain the ship position from the departure port P0 to the destination port P in the recommended route belt. N The weather routes between the two countries include the lowest fuel consumption routes, the shortest time routes and the shortest distance routes;

[0012] S5: Determine the probability of selecting the lowest fuel consumption route, the probability of selecting the shortest time route, and the probability of selecting the shortest distance route based on the fuel cost and the voyage time, so as to determine the optimal weather route for the current voyage;

[0013] S6: Calculate the ship position P1 starting from the departure port and passing the weather update time threshold according to the optimal weather route;

[0014] S7: Repeat S2 to S3, and according to the optimal weather route of the current voyage, based on the A-star algorithm, obtain the next weather update time threshold time from the ship position P1 to the destination port position P in the recommended route band. N The updated optimal weather route between the two intervals is obtained; the ship position P2 is obtained when the next weather update time threshold is passed; and so on, the ship position P in the optimal weather route is obtained. i , i∈1,……,N-1, thereby obtaining the entire optimal weather route.

[0015] Furthermore, the shortest distance route is the one with the shortest total distance from the departure port to the destination port, that is,

[0016]

[0017] Where, L represents the total distance from the departure port to the destination port; l i represents the distance between the i-1th meteorological data grid node and the ith meteorological data grid node; n represents the number of meteorological data grid nodes; i represents the meteorological data grid node number;

[0018] The cost function of the A-star algorithm is improved to a function representing the spherical distance, namely:

[0019]

[0020] Where g i is the spherical distance from the current meteorological data grid node to the next meteorological data grid node; is the latitude value of the current meteorological data grid node; is the latitude value of the next meteorological data grid node; i is the longitude value of the current meteorological data grid node; i+1 is the longitude value of the next meteorological data grid node;

[0021]

[0022] Where h i is the distance from the next weather data grid node to the destination node; is the latitude value of the destination node; is the latitude value of the next meteorological data grid node; goal is the longitude value of the destination node; i+1 The longitude value of the next meteorological data grid node.

[0023] Furthermore, the lowest fuel consumption route has the lowest fuel consumption from the departure port to the destination port, that is,

[0024]

[0025] Where: Fuel total The total fuel consumption from the departure port to the destination port; fuel i is the fuel consumption between the i-1th meteorological data grid node and the i-th meteorological data grid node;

[0026] The cost function with minimum fuel consumption as the optimization goal is:

[0027] g fuel =v i ·S(v) (6)

[0028] Where g fuel is the fuel consumption between the ith meteorological data grid node and the i+1th meteorological data grid node; S(v) is the function of fuel consumption versus ship speed v, with ship speed v being the independent variable; v i is the ship speed between the i-th meteorological data grid node and the i+1-th meteorological data grid node;

[0029] The heuristic function with minimum fuel consumption as the optimization objective is:

[0030] h fuel =vmean ·S(v)·t rest (7)

[0031] Where: h fuel is the heuristic function with the minimum fuel consumption as the optimization objective; v mean is the average speed from the i+1th node to the destination node; t rest is the estimated flight time from the i+1th node to the destination node; where:

[0032]

[0033] Where s rest is the distance from the i+1th meteorological data grid node to the destination point;

[0034]

[0035] Where, t total is the estimated total time of the entire voyage; t i is the required flight time for the i-th leg; t i+1 It is the navigation time required from the current meteorological data grid node to the next meteorological data grid node.

[0036] Furthermore, the shortest time route is the one that takes the least time from the departure port to the destination port, i.e.

[0037]

[0038] Where: T is the total sailing time from the departure port to the destination port; s i is the distance of the i-th flight segment, v i is the speed of the i-th segment;

[0039] Use the stall function as the cost function of the shortest time route, that is:

[0040] V lost =V w +V current (11)

[0041] Where V lost V is the speed loss of the ship sailing from the current meteorological data grid node to the next meteorological data grid node; w V is the speed loss caused by wind and waves during the voyage between the current meteorological data grid node and the next meteorological data grid node; current is the speed loss caused by ocean currents during the voyage between the current meteorological data grid node and the next meteorological data grid node; where:

[0042]

[0043] Where b0 and bj are the coefficients of the ship stall equation; m is the number of stall factors; x j is the stall factor; j is the number of the stall factor;

[0044]

[0045] Where V oc is the ocean current velocity, The sheer angle is the angle between the current direction and the bow direction of the ship.

[0046] The heuristic function for the shortest time route is as follows:

[0047]

[0048] Where g t is the heuristic value in time units; t total is the estimated total time of the entire voyage; t i is the required flight time for the i-th leg; t i+1 It is the navigation time required from the current meteorological data grid node to the next meteorological data grid node.

[0049] Furthermore, in S5, the method for determining the optimal weather route for this voyage is as follows:

[0050]

[0051]

[0052]

[0053] Where: P(y=0) represents the probability of the shortest distance route being selected; P(y=1) represents the probability of the shortest time route being selected; P(y=2) represents the probability of the lowest fuel consumption route being selected; y represents the event; X k Indicates the selection support characteristic variables, namely X1 and X2; among them, X1 represents the voyage time; β1 is the parameter representing the voyage time; X2 represents the fuel cost; β2 is the parameter representing the fuel cost; β k Indicates that the maximum likelihood estimation method is used to solve the optimal estimate; k is the number of the influencing factors of route selection;

[0054] So we get the optimal weather route for this voyage:

[0055]

[0056] Beneficial effects: The present invention provides a dynamic optimization method for weather routes based on the A-star algorithm, which constructs a navigation environment model according to meteorological data of wind, waves, surges and ocean currents; obtains a recommended route belt by translating the first distance threshold to both sides of the recommended climate route, thereby limiting the optimization range of the route and improving the operation efficiency of the algorithm; obtains weather routes including the lowest fuel consumption route, the shortest time route and the shortest distance route based on the A-star algorithm, and determines the probability of the lowest fuel consumption route being selected, the probability of the shortest time route being selected and the probability of the shortest distance route being selected according to the fuel cost and the navigation time; determines the optimal weather route for this voyage; and continuously updates the navigation environment using the forecasted weather data, continuously optimizes the remaining segments, and finally obtains the optimal route, so as to maximize the use of meteorological conditions to minimize the total cost of ship operation and promote the healthy development of smart shipping. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0058] Figure 1 Schematic diagram of the restriction effect on the recommended route belt in an embodiment of the present invention;

[0059] Figure 2 Schematic diagram of meteorological data and longitude and latitude grids for a certain area of the Pacific Ocean in an embodiment of the present invention;

[0060] Figure 3 Schematic diagram of an optimized route with the shortest sailing distance as the optimization target in an embodiment of the present invention;

[0061] Figure 4 This is a schematic diagram of an optimized route with minimum fuel consumption as the optimization goal in an embodiment of the present invention;

[0062] Figure 5 Schematic diagram of meteorological conditions in the local sea area east of Tokyo Port on May 1, 2022, in an embodiment of the present invention;

[0063] Figure 6 This is a map of meteorological conditions in the central Pacific on May 10, 2022, in an embodiment of the present invention;

[0064] Figure 7 This is a flow chart of the weather route dynamic optimization method of the present invention. DETAILED DESCRIPTION

[0065] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0066] In order to solve the above problems existing in the prior art, the present invention designs a dynamic optimization method for weather routes that takes into account wind, waves, ocean currents, has strong real-time performance, and can add constraints.

[0067] In order to achieve the above object, the technical solution of the present invention is as follows: A weather route dynamic optimization method based on A star algorithm comprises the following steps, as shown in the attached figure: Figure 7 As shown:

[0068] S1: Obtain the recommended climate route for this voyage from the recommended navigation route database as the initial route, and shift the recommended climate route to both sides by a first distance threshold to obtain a recommended route belt, limit the search area, and improve algorithm efficiency;

[0069] Specifically, the first distance threshold in this embodiment is specified by the sailors based on experience so that the ship can avoid severe sea conditions; in actual operation, the sailors can continuously input different first distance thresholds and generate multiple routes for comparison.

[0070] S2: Based on the predicted meteorological data and the longitude and latitude grid, obtain a meteorological data grid containing the entire voyage within the meteorological update time threshold; the predicted meteorological data includes wind direction, wind speed, wave height, swell direction, swell height, flow direction, and flow speed;

[0071] Specifically, the meteorological data grid in this embodiment is a latitude and longitude grid combined with meteorological data. Setting the meteorological data grid is a commonly used method in the field and will not be described in detail here.

[0072] The forecasted weather data can be obtained from the National Oceanic and Atmospheric Administration (NOAA) in the United States or the European Centre for Medium-Range Weather Forecasts (ECWMF). The downloaded weather data files are in netCDF format and parsed using a C++ parser or functions from the MATLAB built-in Mapping toolbox. The data is then added to the latitude and longitude grids of the recommended route, or search area. Depending on the accuracy of the weather data, the latitude and longitude grid sizes are 1°×1°, 0.5°×0.5°, or 0.25°×0.25°. The specific grid size is determined by the vessel operator.

[0073] S3: generating navigation grid nodes with obstacles and wind, wave and current with obstacle attributes in the recommended route belt;

[0074] Specifically, in this embodiment, the wind, wave, and current attributes that represent obstacles are set based on the specific user. Each ship user has different standards for wind, wave, and current obstacle attributes. For example, ships of different tonnages have different wind speeds, wave heights, swell heights, current directions, and current velocities that can withstand safe navigation. Combining meteorological data with a latitude and longitude grid generates a meteorological data grid for the entire voyage. Critical values for wind speed, wave height, current speed, and current direction are set based on the ship's seakeeping performance. Objects exceeding these thresholds are considered obstacles and automatically avoided during navigation.

[0075] S4: Based on the navigation grid nodes and the meteorological data grid nodes, based on the A-star algorithm, obtain the ship position from the departure port P0 to the destination port P in the recommended route belt. N The weather routes between the two countries include the lowest fuel consumption routes, the shortest time routes and the shortest distance routes;

[0076] Specifically, in this embodiment, obstacle areas such as islands and reefs, high-wave areas, and high-latitude restricted areas are set in the recommended route belt to prevent the optimized route from passing through unnavigable areas. The longitude and latitude coordinates of the island and reef areas can be derived from electronic nautical charts. The A-star algorithm can determine the coordinates of the island and reef areas as obstacles. Navigators can set latitude thresholds based on the latitudes traversed by the voyage to avoid ships sailing into high-latitude areas with harsh sea conditions. Finally, according to different optimization objectives, weather routes including the lowest fuel consumption route, the shortest time route, and the shortest distance route are solved:

[0077] Preferably, the shortest distance route in this embodiment is solved by a cost function based on the sailing distance;

[0078] Let L be the total distance from the departure port to the destination port, l i is the distance between two adjacent meteorological data grid nodes, then the optimization goal of the shortest distance problem is to minimize the total distance L of the entire navigation process, that is:

[0079]

[0080] Where, L represents the total distance from the departure port to the destination port; l i represents the distance between the i-1th meteorological data grid node and the ith meteorological data grid node; n represents the number of meteorological data grid nodes; i represents the meteorological data grid node number;

[0081] Specifically, when the traditional A-star algorithm searches for the shortest path, the cost represents the cost (i.e., the distance) already incurred to reach the current weather data grid node, and the heuristic value represents the remaining cost from the current node to the destination node. The cost function and heuristic function are used to calculate the cost and heuristic value. The traditional A-star algorithm often uses Manhattan distance, Euclidean distance, and diagonal distance to represent the cost and heuristic value, namely:

[0082] f=g+h (2)

[0083] Where f is the total cost; g is the cost already incurred to reach the current meteorological data grid node; and h is the cost required to reach the destination meteorological data grid node from the current meteorological data grid node, also known as the heuristic value. When a ship reaches the current meteorological data grid node, it compares the f values of the next candidate meteorological data grid nodes and finds the meteorological data grid node with the minimum f value, which becomes the next meteorological data grid node for navigation. Specifically, in this embodiment, the cost in calculating the shortest time route refers to time, the cost in calculating the shortest distance route refers to distance, and the cost in calculating the lowest fuel consumption route refers to fuel consumption.

[0084] Improve the cost function to a function representing the spherical distance, that is:

[0085]

[0086] Where g i is the spherical distance from the current meteorological data grid node to the next meteorological data grid node; is the latitude value of the current meteorological data grid node; is the latitude value of the next meteorological data grid node; i is the longitude value of the current meteorological data grid node; i+1 is the longitude value of the next meteorological data grid node;

[0087] Improve the heuristic function to spherical distance and keep it consistent with the cost function for easy calculation, that is:

[0088]

[0089] Where h i is the distance from the next node to the destination node; is the latitude value of the destination node; is the latitude value of the next node; goal is the longitude value of the current node; i+1 The longitude value of the next node.

[0090] Preferably, the minimum fuel consumption route is the one with the lowest fuel consumption from the departure port to the destination port, that is, the minimum fuel consumption route is solved based on the fuel consumption as the cost function;

[0091] The optimization target of the minimum fuel consumption route is the total fuel consumption of the entire voyage. total Minimum, that is:

[0092]

[0093] Where: Fuel total The total fuel consumption from the departure port to the destination port; fuel i is the fuel consumption between the i-1th meteorological data grid node and the i-th meteorological data grid node;

[0094] A ship's fuel consumption depends on the main engine output power and sailing time, and sailing time is determined by the range and speed. During the actual sailing process of a ship, the ship will be affected by the wind, waves, and ocean currents in the area, and the speed will change. The output power of the ship's main engine can correspond to the corresponding speed, and the fuel consumption also has a corresponding relationship with the output power of the ship's main engine. This correspondence is reflected in the ship's speed and fuel consumption table. The speed of the ship moving from the current node to the next node is constant, so the fuel consumption of the ship in this section can be calculated. The algorithm cost function with the lowest fuel consumption as the optimization goal can be designed as:

[0095] g fuel =v i ·S(v) (6)

[0096] Where g fuel is the fuel consumption between the i-th node and the i+1-th node; S(v) is the function of fuel consumption to ship speed v, where ship speed v is the independent variable and fuel consumption S(v) is the function value; v i is the ship speed between the i-th meteorological data grid node and the i+1-th meteorological data grid node;

[0097] To facilitate calculation, the heuristic function is also modified to a function expressed in fuel consumption, namely:

[0098] h fuel =v mean ·S(v)·t rest (7)

[0099] Where: h fuel is the heuristic function with the minimum fuel consumption as the optimization objective; v mean is the average speed from the i+1th node to the destination node, which can be obtained by formula (8); restis the estimated sailing time from the i+1th node to the destination node, which can be obtained by formula (9);

[0100]

[0101] Where s rest is the distance from the i+1th meteorological data grid node to the destination, expressed as spherical distance;

[0102]

[0103] Where, t total is the estimated total time of the entire voyage; t i is the required flight time for the i-th leg; t i+1 is the flight time required from the current meteorological data grid node to the next meteorological data grid node, that is, the flight time required for the i+1th flight segment.

[0104] Preferably, the shortest time route is the one that takes the least time from the departure port to the destination port, that is, based on the sailing time as the cost function, the shortest time route can be solved; then the optimization goal of the shortest sailing time problem is to solve the minimum total sailing time of the entire sailing process, that is:

[0105]

[0106] Where: T is the total sailing time from the departure port to the destination port; s i is the distance of the i-th segment, v i is the speed of the i-th segment;

[0107] It can be seen that the navigation time is determined by the distance and speed. The distance s that a ship travels from the current node to the next node is i is fixed, so the greater the speed, the shorter the voyage time. If the speed is to be as large as possible, the stall caused by weather conditions must be minimized. Therefore, the stall function can be used as the cost function for the shortest time route, that is:

[0108] V lost =V w +V current (11)

[0109] Where V lost V is the speed loss of the ship sailing from the current meteorological data grid node to the next meteorological data grid node; w is the speed loss caused by wind and waves during the voyage between the current meteorological data grid node and the next meteorological data grid node; obtained by formula (12); V currentis the speed loss caused by ocean currents during the voyage between the current meteorological data grid node and the next meteorological data grid node; obtained by formula (13); the above variables are all positive and negative, a positive value increases the navigation speed, and a negative value reduces the navigation speed.

[0110]

[0111] Where b0 and bj are the coefficients of the ship stall equation; m is the number of stall factors; x j is the stall factor; j is the number of the stall factor; the stall factor takes into account wind direction, wind speed, wave height, swell direction, swell height, displacement, heading and speed;

[0112]

[0113] Where V oc is the ocean current velocity, It is the sheer angle, that is, the angle between the flow direction and the bow direction of the ship.

[0114] The heuristic function is also modified to a function with time as the unit, which is consistent with the cost function, that is:

[0115]

[0116] Where g t is the heuristic value in time units; t total is the estimated total time of the entire voyage; t i is the required flight time for the i-th leg; t i+1 is the flight time required from the current meteorological data grid node to the next meteorological data grid node, that is, the flight time required for the i+1th flight segment.

[0117] S5: Determine the probability of selecting the lowest fuel consumption route, the probability of selecting the shortest time route, and the probability of selecting the shortest distance route based on the fuel cost and the voyage time, so as to determine the optimal weather route for the current voyage;

[0118] In S5, the method for determining the optimal weather route for this voyage is as follows:

[0119] A ternary logit model is established to calculate the probability of the lowest fuel consumption route, the shortest time route, and the shortest distance route being selected:

[0120]

[0121]

[0122]

[0123] Where: P(y=0) represents the probability of the shortest distance route being selected; P(y=1) represents the probability of the shortest time route being selected; P(y=2) represents the probability of the lowest fuel consumption route being selected; y represents an event, specifically the shortest distance event, the shortest time event, and the lowest fuel consumption test piece; X k represents the selection support characteristic variables, namely X1 and X2; which can be calculated from the navigation speed, navigation mileage and fuel consumption of the lowest fuel consumption route, the shortest time route and the shortest distance route; wherein X1 represents the navigation time; β1 is a parameter representing the navigation time; X2 represents the fuel cost; β2 is a parameter representing the fuel cost; β k Indicates that the maximum likelihood estimation method is used to solve the optimal estimate; k is the number of the influencing factors of route selection, where the influencing factors in this embodiment are fuel cost and voyage time;

[0124] So we get the optimal weather route for this voyage:

[0125]

[0126] Specifically, in another embodiment of the present invention, the optimal weather route determines the route optimization strategy based on the operational needs of the ship operator. Furthermore, based on actual operational needs, the estimated arrival time is used as a constraint condition for route optimization to improve the practicality of the optimized route.

[0127] S6: Calculate the ship position P1 starting from the departure port and passing the weather update time threshold according to the optimal weather route;

[0128] Specifically, the method of calculating the ship position P1 when the weather update time threshold is passed according to the optimal weather route in this embodiment is a common method in the field and will not be described in detail here.

[0129] S7: Repeat S2 to S3, and according to the optimal weather route of the current voyage, based on the A-star algorithm, obtain the next weather update time threshold time from the ship position P1 to the destination port position P in the recommended route band. N The updated optimal weather route between the two intervals is obtained; the ship position P2 is obtained when the next weather update time threshold is passed; and so on, the ship position P in the optimal weather route is obtained. i , i∈1,……,N-1, thereby obtaining the entire optimal weather route.

[0130] Specifically, the weather update time threshold in this embodiment is 12 hours, that is, the ship starts calculating the optimal weather route from the departure port, and calculates the ship's position 12 hours later based on the optimal weather route at that time; 12 hours later, the weather data is updated, and the updated optimal weather route from the current ship position to the destination port is calculated. Based on the updated optimal weather route, the ship's position 12 hours later from the current position is obtained, and so on, to obtain the optimal weather route for the entire route.

[0131] The simulation experiments in this embodiment of the present invention are based on the MATLAB simulation platform and the Electronic Chart and Information Display System (ECDIS). The simulation example in the accompanying figure shows an ocean route with Tokyo Port as the departure port and Los Angeles Port as the destination port. This embodiment optimizes only anchorage-to-anchorage transoceanic routes. The starting and destination coordinates are (35.5°N, 140°E) and (33.5°N, 118.5°W), respectively. The departure time is May 1, 2022. The vessel parameters in this embodiment are shown in Table 1.

[0132] Table 1 Simulation experiment ship parameters

[0133]

[0134] 1. Generate a navigation environment model covering the entire voyage based on predicted meteorological data

[0135] (1) Obtain the recommended climate route for this voyage, and extend the recommended climate route to the north and south by the first distance threshold to form a recommended route belt to limit the search area and improve the calculation efficiency, such as Figure 1 This is a diagram showing the effect of restricting the search area using the climate route from Tokyo Port, Japan to Los Angeles Port, USA as the initial route;

[0136] (2) Obtain the predicted wind, wave, and current data, parse them, and add them to the meteorological attributes of the grid nodes in the search area, such as Figure 2 This is a schematic diagram of meteorological data and longitude and latitude grids for a certain area in the Pacific Ocean;

[0137] (3) Generate navigation grid nodes with wind, wave, current and obstacle attributes within the restricted search area.

[0138] 2. Search for the optimal route from the departure port to the destination port based on the A-star algorithm.

[0139] (1) Based on the sailing distance as the cost function, the shortest distance route can be solved;

[0140] The schematic diagram of the optimized route with the shortest sailing distance as the optimization goal is as follows Figure 3 Since the cost function is the spherical distance, the optimal weather route is basically consistent with the great circle route while ensuring navigation safety.

[0141] (2) Based on fuel consumption as the cost function, the route with the lowest fuel consumption can be solved;

[0142] The optimized route with the lowest fuel consumption as the optimization goal is as follows Figure 4 . Figure 5 The weather conditions for the local area east of Tokyo Port on May 1st are given. Comparing the optimized route shows that the ship avoided the headwind and wave areas in the local area east of Tokyo Port, which is also fuel-efficient in navigation practice. The experimental data for the lowest fuel consumption route is shown in Table 2. As we all know, fuel consumption is proportional to the cube of speed. Therefore, the experimental assumption is that the relationship between fuel consumption and speed is:

[0143] F=a·v 3

[0144] Where F is the fuel consumption per unit time, a is the proportionality coefficient, and v is the ship speed. The optimization algorithm of the present invention also uses the same method to calculate fuel consumption when compared with the initial recommended climate route. Therefore, the accuracy of the fuel consumption calculation does not affect the optimization effect. Therefore, the formula can be used to approximate the calculation of fuel consumption.

[0145] Table 2 Minimum fuel route test results

[0146]

[0147] (3) Based on the navigation time as the cost function, the shortest time route can be solved.

[0148] The optimized route with the shortest time as the optimization goal is as follows Figure 6 , due to the presence of strong wind and wave areas in the central Pacific Ocean, such as Figure 2 Continuing to follow the great circle route will inevitably affect the safety of the ship and cargo. Therefore, the algorithm avoids the areas with headwind and waves and searches for the node with the smallest stall, thus generating a route with the shortest time. The experimental data of the shortest time optimization route are shown in Table 3.

[0149] Table 3 Shortest time route experimental results

[0150]

[0151] 3. Dynamically optimize routes based on dynamic programming

[0152] (1) Assuming that the update time of meteorological data is 12 hours, based on the first optimized route, the ship position 12 hours after departure is predicted, which is set as P1;

[0153] (2) Update the weather data with P1 as the initial position and re-optimize the route;

[0154] (3) Repeat steps (1) and (2) until P N-1The ship position is obtained, and the entire route (P0, P1, P2, ..., P N-1 , P N ).

[0155] This embodiment fully considers meteorological conditions, solves the cumulative errors in predicted speed, ship position and estimated arrival time caused by meteorological conditions, and improves the practicality of the optimal weather route. At the same time, it takes into account the constraints in navigation practice. For example, in relatively crowded ports, ship operators often want ships to arrive at a specific time, rather than arriving too early or too late, which imposes the estimated arrival time constraint on route optimization. In addition, since the voyage time of transoceanic ships often exceeds the time range of weather forecasts, it is necessary to continuously update meteorological data during the voyage to perform dynamic optimization of the route. This embodiment fully considers the changes in weather forecast time and has good dynamic optimization performance.

[0156] This embodiment has the following beneficial effects:

[0157] 1. To improve the optimization performance of weather routes, this invention considers the impact of wind, waves, and ocean currents on ship navigation, establishes a weather data grid, and obtains the optimal weather route based on the A-star algorithm, thereby improving the accuracy of speed prediction, ship position calculation, and estimated arrival time, ensuring navigation safety;

[0158] 2. This embodiment establishes a recommended route belt to limit the search direction and search area of the A-Star algorithm, improving the algorithm's time efficiency. First, the present invention limits the algorithm's search direction based on the relative positions of the voyage's departure and destination ports, reducing the number of candidate nodes and significantly improving the algorithm's computational efficiency. Second, the present invention uses the recommended route belt to limit the search area based on the initial recommended climate route, rationally restricting the search area and improving the algorithm's operational efficiency and optimization effect.

[0159] 3. To apply the A-star algorithm to navigation practice, this paper designs a route optimization strategy based on three optimization objectives: shortest time, shortest distance, and lowest fuel consumption, in accordance with the operational needs of ship operators. Furthermore, based on actual operational needs, the estimated arrival time is used as a constraint in route optimization to improve the practicality of the optimized route.

[0160] 4. This paper proposes a ternary logit model to evaluate the total navigation cost of the shortest distance route, the shortest time route, and the lowest fuel consumption route. Without human intervention, the optimized route with the lowest total cost can still be selected, laying the foundation for the application of this algorithm in intelligent ship navigation systems.

[0161] 5. This embodiment is based on the concept of dynamic programming. It uses the estimated ship position and the time it takes to reach that position to update the weather data in real time. The route is then optimized again using the estimated ship position as the new starting point. Repeating the above steps will yield a dynamic optimal weather route.

[0162] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A weather route dynamic optimization method based on the A-star algorithm, characterized by: The following steps are involved: S1: Obtain a recommended climate route for the current voyage, and shift the recommended climate route to both sides by a first distance threshold to obtain a recommended route belt; S2: Based on the predicted meteorological data and the longitude and latitude grid, a meteorological data grid including the entire voyage within a meteorological update time threshold is obtained to obtain a meteorological data grid node; the predicted meteorological data includes wind direction, wind speed, wave height, swell direction, swell height, ocean current direction, and ocean current speed; S3: generating navigation grid nodes with obstacles and wind, wave and current with obstacle attributes in the recommended route belt; S4: Based on the navigation grid nodes and the meteorological data grid nodes, based on the A-star algorithm, obtain the ship position from the departure port P0 to the destination port P in the recommended route belt. N The weather routes between the two countries include the lowest fuel consumption routes, the shortest time routes and the shortest distance routes; S5: Determine the probability of selecting the lowest fuel consumption route, the probability of selecting the shortest time route, and the probability of selecting the shortest distance route based on the fuel cost and the voyage time, so as to determine the optimal weather route for the current voyage; S6: Calculate the ship position P1 starting from the departure port and passing the weather update time threshold according to the optimal weather route; S7: Repeat S2 to S3, and according to the optimal weather route of the current voyage, based on the A-star algorithm, obtain the next weather update time threshold time from the ship position P1 to the destination port position P in the recommended route band. N The updated optimal weather route between the two intervals is obtained; the ship position P2 is obtained when the next weather update time threshold is passed; and so on, the ship position P in the optimal weather route is obtained. i , i∈1,……,N-1, thereby obtaining the entire optimal weather route.

2. The method for dynamic optimization of weather routes based on the A-star algorithm according to claim 1, characterized in that: The shortest distance route is the shortest total distance from the departure port to the destination port, that is Where, L represents the total distance from the departure port to the destination port; l i represents the distance between the i-1th meteorological data grid node and the ith meteorological data grid node; n represents the number of meteorological data grid nodes; i represents the meteorological data grid node number; The cost function of the A-star algorithm is improved to a function representing the spherical distance, namely: Where g i is the spherical distance from the current meteorological data grid node to the next meteorological data grid node; is the latitude value of the current meteorological data grid node; is the latitude value of the next meteorological data grid node; i is the longitude value of the current meteorological data grid node; i+1 is the longitude value of the next meteorological data grid node; Where h i is the distance from the next weather data grid node to the destination node; is the latitude value of the destination node; is the latitude value of the next meteorological data grid node; goal is the longitude of the destination node; i+1 The longitude value of the next meteorological data grid node.

3. The method for dynamic optimization of weather routes based on the A-star algorithm according to claim 1, characterized in that: The lowest fuel consumption route has the lowest fuel consumption from the departure port to the destination port, that is Where: Fuel total The total fuel consumption from the departure port to the destination port; fuel i is the fuel consumption between the i-1th meteorological data grid node and the i-th meteorological data grid node; The cost function with minimum fuel consumption as the optimization goal is: g fuel =v i ·S(v) (6) Where g fuel is the fuel consumption between the ith meteorological data grid node and the i+1th meteorological data grid node; S(v) is the function of fuel consumption versus ship speed v, with ship speed v being the independent variable; v i is the ship speed between the i-th meteorological data grid node and the i+1-th meteorological data grid node; The heuristic function with minimum fuel consumption as the optimization objective is: h fuel =v mean ·S(v)·t rest (7) Where: h fuel is the heuristic function with the minimum fuel consumption as the optimization objective; v mean is the average speed from the i+1th node to the destination node; t rest is the estimated flight time from the i+1th node to the destination node; where: Where s rest is the distance from the i+1th meteorological data grid node to the destination point; Where, t total is the estimated total time of the entire voyage; t i is the required flight time for the i-th leg; t i+1 It is the navigation time required from the current meteorological data grid node to the next meteorological data grid node.

4. The method for dynamic optimization of weather routes based on the A-star algorithm according to claim 1, characterized in that: The shortest time route is the one that takes the least time from the departure port to the destination port, that is Where: T is the total sailing time from the departure port to the destination port; s i is the distance of the i-th segment, v i is the speed of the i-th segment; Use the stall function as the cost function of the shortest time route, that is: V lost =V w +V current (11) Where V lost V is the speed loss of the ship sailing from the current meteorological data grid node to the next meteorological data grid node; w V is the speed loss caused by wind and waves during the voyage between the current meteorological data grid node and the next meteorological data grid node; current is the speed loss caused by ocean currents during the voyage between the current meteorological data grid node and the next meteorological data grid node; where: Where b0 and bj are the coefficients of the ship stall equation; m is the number of stall factors; x j is the stall factor; j is the number of the stall factor; Where V oc is the ocean current velocity, The sheer angle is the angle between the current direction and the bow direction of the ship. The heuristic function for the shortest time route is as follows: Where g t is the heuristic value in time units; t total is the estimated total time of the entire voyage; t i is the required flight time for the i-th leg; t i+1 It is the navigation time required from the current meteorological data grid node to the next meteorological data grid node.

5. The method for dynamic optimization of weather routes based on the A-star algorithm according to claim 1, characterized in that: In S5, the method for determining the optimal weather route for this voyage is as follows: Where: P(y=0) represents the probability of the shortest distance route being selected; P(y=1) represents the probability of the shortest time route being selected; P(y=2) represents the probability of the lowest fuel consumption route being selected; y represents the event; X k Indicates the selection support characteristic variables, namely X1 and X2; among them, X1 represents the voyage time; β1 is the parameter representing the voyage time; X2 represents the fuel cost; β2 is the parameter representing the fuel cost; β k Indicates that the maximum likelihood estimation method is used to solve the optimal estimate; k is the number of the influencing factors of route selection; So we get the optimal weather route for this voyage:

Images