Unmanned aerial vehicle cruise path planning method for ship emission supervision under complex sea conditions

Abstract

This application provides a method for planning drone patrol paths for ship emission monitoring under complex sea conditions. The method includes: constructing a drone patrol spatiotemporal network using ship spatiotemporal nodes, maritime base station spatiotemporal nodes, and feasible directed arcs for drone flight; calculating the actual energy consumption of the drone flying along any directed arc under relative wind direction; determining the actual monitoring benefits of drone visits to each spatiotemporal node under sea fog conditions; optimizing the drone patrol path under given battery capacity, time window, wind direction, and sea fog conditions, with the objective of maximizing the total actual monitoring benefits obtained by the drone swarm in a single patrol, while setting corresponding constraints for the target ship, drones, maritime base stations, and drone battery power based on actual energy consumption. This method is applicable to short-term, routine maritime patrol scenarios and can provide path optimization solutions for multi-drone ship emission monitoring tasks under limited power and global time window constraints.

Description

Technical Field

[0001] This application belongs to the field of intelligent maritime supervision technology, and in particular relates to a method for planning the cruise path of unmanned aerial vehicles (UAVs) for ship emission supervision under complex sea conditions. Background Technology

[0002] Traditional maritime regulatory methods mainly include on-site spot checks by maritime patrol boats, law enforcement officers boarding ships to inspect fuel documents or collect fuel samples, and deploying fixed exhaust gas monitoring stations around ports or waterways to monitor the sulfur content, nitrogen oxides, sulfur oxides, and particulate matter emissions from ship fuel. These methods suffer from high operating costs, limited coverage, insufficient proactive tracking capabilities, potential for creating regulatory blind spots, and possible disruption to normal ship navigation. Recently, drone-based patrol monitoring has become an important auxiliary means of ship emission regulation. Equipped with portable gas sensors, drones can quickly approach target ships and enter the exhaust plume area behind the ship's smokestack to collect samples, thus completing preliminary emission verification without boarding the ship or interfering with its normal navigation.

[0003] However, maritime drone patrols differ from conventional land-based path planning. On one hand, target vessels continuously navigate within emission control areas, their spatial coordinates changing over time; drones must reach the corresponding instantaneous position of the vessel at a specific time to complete monitoring. On the other hand, drones have limited endurance, and wind direction and force significantly alter flight energy consumption; localized sea fog can also reduce the effectiveness of sensor sampling data. If scheduling is still based on path planning methods for windless, stationary targets, or fixed unit energy consumption, problems such as excessively rapid power consumption of drones on upwind segments, inability to return safely, reduced sampling benefits in sea fog areas, or repeated monitoring of the same vessel can easily arise. Therefore, in complex sea conditions, how to comprehensively consider moving vessel targets, wind field energy consumption, sea fog benefit attenuation, drone power limitations, station deployment, and multi-drone collaboration to generate safe, efficient, and highly profitable patrol paths is a pressing technical problem that needs to be solved in the field of ship emission regulation.

[0004] Existing research on unmanned aerial vehicle (UAV) ship emission regulation has achieved certain results in areas such as exhaust gas telemetry, spatiotemporal network modeling of mobile vessels, multi-UAV collaborative scheduling, and path optimization algorithms. However, existing technologies still have three problems: First, most studies set the UAV flight energy consumption as a fixed constant, lacking consideration of the impact of real wind fields at sea. In headwind environments, planned paths may lead to excessively rapid power consumption and difficulty in safely returning to base. Second, some research models that consider wind field factors are quite complex, usually targeting long-term or extreme weather scheduling scenarios. This results in high computational costs and complex parameter tuning, making it difficult to meet the short-term, rapid response application needs of grassroots maritime supervision. Third, existing research does not adequately consider the attenuation of monitoring effectiveness under complex sea conditions such as sea fog, making it difficult to accurately reflect the effectiveness of UAV exhaust gas monitoring results under different meteorological conditions. Summary of the Invention

[0005] In view of this, the purpose of this application is to provide a UAV cruise path planning method for ship emission monitoring under complex sea conditions, which is applicable to short-term routine maritime cruise scenarios and can provide path optimization schemes for multi-UAV ship emission monitoring tasks under the constraints of limited power and global time window.

[0006] This application provides a method for UAV cruise path planning for ship emission regulation under complex sea conditions, including: A spatiotemporal network for UAV patrol is constructed by ship spatiotemporal nodes, maritime base station spatiotemporal nodes, and directed arcs that allow UAV flight. The directed arc refers to the ability of a UAV to depart from the time and space location corresponding to one spatiotemporal node and accurately arrive at its spatial location at the time corresponding to another spatiotemporal node. For any directed arc in the UAV's cruise spatiotemporal network, calculate the actual energy consumption of the UAV flying along the directed arc in the relative wind direction; Determine the actual monitoring benefits of UAV visits to various spatiotemporal nodes under sea fog conditions; The goal is to maximize the total actual monitoring benefits obtained by the drone swarm in a single patrol. At the same time, corresponding constraints are set for the target ship, drones, maritime base stations, and drone battery power based on actual energy consumption. Given battery capacity, time window, wind direction and sea fog conditions, the drone patrol path is optimized. Finally, the drone departure time, return time, target ship visit time, patrol path, relative wind direction of the flight segment, flight energy consumption, node remaining power and total actual monitoring benefits are output.

[0007] Furthermore, before constructing the UAV cruise spatiotemporal network jointly by ship spatiotemporal nodes, maritime base station spatiotemporal nodes, and feasible directed flight arcs of the UAV, the method also includes: Discretization is performed for single short-term planning periods, and the continuous planning time is divided into several equal-length steps; Based on the target ship's initial position, heading, and speed, the spatial position of the ship at each time step is deduced, and spatiotemporal nodes of the ship with time tags are generated. Based on the fixed spatial location of the maritime base station and its state at each time step, a spatiotemporal node of the maritime base station is generated, which serves as both the departure node and the return node of the drone. Based on the maximum flight speed of the UAV, the spatial distance between each spatiotemporal node, and the time tag, it is determined whether there is a feasible directed arc for UAV flight between each spatiotemporal node; the spatiotemporal nodes include ship spatiotemporal nodes and maritime base station spatiotemporal nodes.

[0008] Furthermore, the calculation of the actual energy consumption of the UAV flying along the directional arc in relative wind direction includes: For any directed arc in the spatiotemporal network of the UAV cruise The heading angle of the UAV flying along the directed arc is calculated based on the spatial coordinates of the two spatiotemporal nodes: ; In the formula, For the drone's heading angle, Representing spatiotemporal nodes coordinates Represents the coordinates of spatiotemporal node j; Obtain the ambient wind direction angle and calculate its relative angle with the UAV's flight heading angle, then map this relative angle to... Within the range: ; In the formula, For the ambient wind direction angle; Based on the relative angle after mapping Flight status is divided into three relative wind directions: tailwind, crosswind, and headwind, and corresponding segmented energy consumption coefficients are set. ; Based on segmented energy consumption coefficients Calculate the UAV along the directed arc Actual energy consumption of flight: ; In the formula, This represents the baseline power consumption per unit time under windless conditions. spatiotemporal nodes To spacetime node The flight time span.

[0009] Furthermore, the flight state is divided into three relative wind directions—tailwind, crosswind, and headwind—based on the mapped relative angles, and corresponding segmented energy consumption coefficients are set, including: like This is a tailwind condition, where the wind provides a boost to the drone's flight. ; like This is a crosswind condition, where the drone needs to consume additional energy to correct its yaw. ; like This is a headwind operating condition, where the drone needs to overcome wind resistance and set... Much greater than 1.

[0010] Furthermore, determining the actual monitoring benefits of UAV visits to various spatiotemporal nodes under sea fog conditions includes: Based on the vessel's tonnage, fuel type, main engine power, and historical violation records, basic monitoring benefits are assigned to each target vessel; When any ship spatiotemporal node falls into a pre-defined sea fog coverage area at a specific time step, a sea fog attenuation coefficient is assigned to the ship spatiotemporal node based on the fog concentration, so that the monitoring benefit attenuation mechanism is automatically triggered when the UAV visits the ship spatiotemporal node. The actual monitoring benefits of the UAV visiting the ship's spatiotemporal nodes are determined by the ship's basic monitoring benefits and the sea fog attenuation coefficient. And the actual monitoring benefit of setting the spatiotemporal node of the maritime base station is 0.

[0011] Furthermore, the constraints set for the target ship, drone, maritime base station, and drone battery power based on actual energy consumption include: (1) Each target vessel within the constrained emission control area shall be assigned a maximum of one UAV for emission monitoring during a single planning period: ; In the formula, If the decision variable is 0 / 1, then the drone Along the directed arc To carry out a flight mission, ,otherwise ,in, , Let be the set of directed arcs that allow UAV flight in the spatiotemporal network of UAV cruise, representing the flight paths that UAVs can execute; A collection of drones for performing monitoring tasks; This refers to the set of all spatiotemporal nodes in the spatiotemporal network of UAV patrols. , It is a set of spatiotemporal nodes for maritime base stations. A set of spatiotemporal nodes for ships; For the target vessel The corresponding set of ship spatiotemporal nodes, among which, , A collection of target vessels to be monitored within the emission control area; (2) Each UAV is restricted to leaving the maritime base station and performing only one patrol mission within a single planning period: ; (3) Each UAV must fly to the next target vessel or fly directly back to the maritime base station after completing the escort sampling upon reaching the target vessel: ; In the formula, Indicates if drone Along the directed arc To carry out a flight mission, ,otherwise ; (4) By constraining whether each UAV is within the maritime base station at each time step, the continuity between the UAV's standby, departure, and return states is ensured: ; In the formula, If the decision variable is 0-1, then the drone At time step Located at a maritime base station Waiting inside, then ,otherwise ,in, , A collection of maritime base stations for deploying drones. , This is the set of discretized planning time steps. express Maritime base station This point in time and space; (5) When the UAV flies along a directed arc, the actual energy consumption under the corresponding relative wind direction is deducted. By introducing a sufficiently large positive number M, the energy consumption of each UAV is constrained to be deducted only when it actually flies along the directed arc: ; In the formula, Let be a non-negative continuous decision variable, representing the unmanned aerial vehicle (UAV). Arrival at the spacetime node The remaining battery power at that time Indicates drone Arrival at the spacetime node The remaining battery power at that time; (6) Constrain each UAV to be in a fully charged state at the initial moment before exiting the cabin: ; In the formula, For drones Maximum rated battery capacity: (7) Constrain the remaining power of each UAV at any node to be between zero and the rated capacity.

[0012] Furthermore, the optimization of the drone's cruise path includes: Extract the selected spatiotemporal nodes and directed arcs for each UAV, and arrange the spatiotemporal nodes according to the time sequence to obtain the cruise path of each UAV; the cruise path consists of the starting spatiotemporal node of the maritime base station, the sequentially visited ship spatiotemporal nodes, and the returning spatiotemporal node of the maritime base station. Based on the cruise path, the actual energy consumption of each drone during each flight segment, the remaining power at each spatiotemporal node, and the remaining power after returning home are output to ensure that the drone can return home safely. It outputs whether each drone is on standby at the maritime base station at each time step, thus providing a reference for drone scheduling.

[0013] The UAV patrol path planning method for ship emission monitoring under complex sea conditions provided in this application has the following beneficial effects: 1) Aiming to maximize the total monitoring benefits, it can reasonably arrange the patrol paths of multiple UAVs within a single planning time window, avoiding duplicate monitoring and blind clustering, improving the verification coverage of high-risk ships, and thus serving the needs of ship emission control area supervision and green shipping development. 2) Improved flight safety: Considering the impact of offshore wind fields on UAV flight energy consumption, segmented energy consumption coefficients are set based on relative wind direction, enabling path planning to avoid high-energy-consuming headwind segments, reducing the risk of UAVs being unable to return safely due to rapid power consumption, and improving the safety of maritime patrol missions. 3) Enhanced environmental adaptability: The impact of sea fog on monitoring benefits is incorporated into the path planning process, and the changes in the effectiveness of monitoring data under different meteorological conditions are quantified through the sea fog attenuation coefficient, making the UAV patrol scheme more in line with the actual operational needs of ship exhaust gas monitoring under complex sea conditions. 4) Improved solution efficiency and scheme rationality: By transforming the dynamic tracking problem of moving ships into a static network optimization problem through the UAV patrol spatiotemporal network, a more accurate and efficient UAV patrol path scheme can be obtained under reasonable scenario simplification. Attached Figure Description

[0014] Figure 1 A flowchart of the UAV cruise path planning method for ship emission regulation under complex sea conditions provided in an embodiment of this application is shown; Figure 2 A schematic diagram of the spatiotemporal network for UAV patrol provided in an embodiment of this application is shown; Figure 3 This paper illustrates a schematic diagram of the drone's cruise path provided in an embodiment of this application. Figure 4 A schematic diagram of the drone cruise path projection provided in an embodiment of this application is shown. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of this technical solution clearer, the following detailed description, in conjunction with specific embodiments, further illustrates this technical solution. It should be understood that these descriptions are merely exemplary and not intended to limit the scope of this technical solution.

[0016] Please see as follows Figure 1 The flowchart shown illustrates a UAV cruise path planning method for ship emission regulation under complex sea conditions. Figure 1 As shown, the method includes: S101. A spatiotemporal network for UAV patrol is jointly constructed by ship spatiotemporal nodes, maritime base station spatiotemporal nodes, and directed arcs that allow for UAV flight.

[0017] The directed arc refers to the ability of a drone to start from a time and space location corresponding to one spatiotemporal node and accurately arrive at its spatial location at a time corresponding to another spatiotemporal node.

[0018] Before constructing the UAV cruise spatiotemporal network using ship spatiotemporal nodes, maritime base station spatiotemporal nodes, and feasible directed arcs of UAV flight, the ship spatiotemporal nodes, maritime base station spatiotemporal nodes, and feasible directed arcs of UAV flight are obtained through the following method: Step 201: Discretize the single short-term planning period and divide the continuous planning time into several equal-length steps.

[0019] Step 202: Based on the initial position, heading and speed of the target ship, deduce the spatial position of the ship at each time step, and generate the ship's spatiotemporal nodes with time tags.

[0020] Step 203: Based on the fixed spatial location of the maritime base station and its status at each time step, generate a spatiotemporal node for the maritime base station, which serves as both the departure node and the return node for the UAV.

[0021] Step 204: Based on the maximum flight speed of the UAV, the spatial distance between each spatiotemporal node, and the time tag, determine whether there is a feasible directed arc for UAV flight between each spatiotemporal node.

[0022] The spatiotemporal nodes include ship spatiotemporal nodes and maritime base station spatiotemporal nodes.

[0023] In summary, for short-duration maritime patrol scenarios, the continuous planning time is divided into several discrete time steps. Based on the target vessel's initial position, heading, and speed, its spatial position at each time step is deduced, thus generating time-labeled spatiotemporal nodes for the vessel. Simultaneously, the state of the maritime base station at each time step is generated as a station spatiotemporal node. The station spatiotemporal nodes, the vessel spatiotemporal nodes, and the feasible directed arcs of UAV flight together constitute a spatiotemporal network, transforming the dynamic tracking problem of moving vessel targets into a static network optimization problem.

[0024] As an example, please refer to Figure 2 The diagram shows a spatiotemporal network diagram of a drone cruise. It illustrates the relationships between maritime base station nodes, mobile vessel nodes, the drone's flight arc, and its accompanying sampling arc.

[0025] S102. For any directed arc in the UAV's cruise spatiotemporal network, calculate the actual energy consumption of the UAV flying along the directed arc in the relative wind direction.

[0026] In practical implementation, the actual energy consumption of the UAV flying along the directional arc under relative wind direction is calculated using the following method: For any directed arc in the spatiotemporal network of the UAV cruise The heading angle of the UAV flying along the directed arc is calculated based on the spatial coordinates of the two spatiotemporal nodes: (1) In the formula, For the drone's heading angle, Representing spatiotemporal nodes coordinates Represents the coordinates of spatiotemporal node j; Obtain the ambient wind direction angle and calculate its relative angle with the UAV's flight heading angle, then map this relative angle to... Within the range: (2) In the formula, For the ambient wind direction angle; Based on the relative angle after mapping Flight status is divided into three relative wind directions: tailwind, crosswind, and headwind, and corresponding segmented energy consumption coefficients are set. ; In practical implementation, the three relative wind directions (tailwind, crosswind, and headwind) and their corresponding segmented energy consumption coefficients are set as follows: like This is a tailwind condition, where the wind provides a boost to the drone's flight. ;like This is a crosswind condition, where the drone needs to consume additional energy to correct its yaw. ;like This is a headwind operating condition, where the drone needs to overcome wind resistance and set... Much greater than 1.

[0027] Based on segmented energy consumption coefficients Calculate the UAV along the directed arc Actual energy consumption of flight: (3) In the formula, This represents the baseline power consumption per unit time under windless conditions. spatiotemporal nodes To spacetime node The flight time span.

[0028] Here, wind direction is factored into the energy consumption coefficient. Change the energy consumption of drone flight This will further affect the cruise routes that drones can choose under battery capacity constraints, but will not directly change the benefits of ship monitoring.

[0029] S103. Determine the actual monitoring benefits of UAV visits to various spatiotemporal nodes under sea fog conditions.

[0030] In practical implementation, the actual monitoring benefits of UAV visits to various spatiotemporal nodes under sea fog conditions are determined through the following methods: Step 1031: Assign basic monitoring benefits to each target vessel based on its tonnage, fuel type, main engine power, and historical violation records.

[0031] In this step, a higher basic monitoring benefit indicates a higher priority monitoring value for the vessel in emissions regulation.

[0032] Step 1032: When any ship spatiotemporal node falls into a pre-defined sea fog coverage area at a specific time step, a sea fog attenuation coefficient is assigned to the ship spatiotemporal node according to the fog concentration, so that the monitoring benefit attenuation mechanism is automatically triggered when the UAV visits the ship spatiotemporal node.

[0033] In this step, a sea fog attenuation coefficient is assigned to the ship's spatiotemporal node based on the different fog concentrations. The value ranges from 0.3 to 0.8. At this point, the actual benefit of the drone visiting the ship's spatiotemporal node is... The corresponding reduction will be applied. Conversely, if the ship's spatiotemporal nodes... Located in a clear, fog-free sea area, set The drone can obtain the full amount of the ship's basic revenue.

[0034] Step 1033: Based on the ship's basic monitoring benefits and the sea fog attenuation coefficient, determine the actual monitoring benefits of the UAV visiting the ship's spatiotemporal node.

[0035] In this step, the actual monitoring benefits when the UAV visits the target ship are calculated using the following formula: (4) Step 1034, and setting the actual monitoring revenue of the maritime base station spatiotemporal node to 0.

[0036] In summary, this application considers the impact of offshore wind fields and localized sea fog on UAV patrol missions. Regarding the wind field impact, step S102 categorizes the flight state into three types—tailwind, crosswind, and headwind—based on the relative angle between the UAV's flight path and the ambient wind direction, and sets energy consumption gain or penalty coefficients for each type to calculate the actual energy consumption of the UAV flying along a directed arc. Regarding the sea fog impact, step S103 sets a sea fog attenuation coefficient based on whether the target ship's spatiotemporal node is located in a sea fog-covered area and the fog concentration level, reducing the basic monitoring benefits of the ship to obtain the actual effective monitoring benefits for that node.

[0037] S104. With the goal of maximizing the total actual monitoring benefits obtained by the UAV swarm in a single patrol, corresponding constraints are set for the target ship, UAVs, maritime base stations, and UAV battery power based on actual energy consumption. Given the battery capacity, time window, wind direction, and sea fog conditions, the UAV patrol path is optimized. Finally, the UAV departure time, return time, target ship visit time, patrol path, relative wind direction of the flight segment, flight energy consumption, node remaining power, and total actual monitoring benefits are output.

[0038] In this step, with the objective of maximizing the total actual monitoring benefits obtained by the drone swarm in a single patrol, the following objective function is established: (5) The aforementioned UAV cruise spatiotemporal network, monitoring benefits, sea fog attenuation, wind direction energy consumption, and pre-set constraints are then transformed into a mixed-integer linear programming model. The model is then solved using the Gurobi optimization solver in Python to obtain the optimized UAV cruise path under given battery capacity, time window, wind direction, and sea fog conditions.

[0039] In practical implementation, the following constraints are set for the target ship, drone, maritime base station, and drone battery power based on actual energy consumption: (1) Single visit constraint: Each target vessel within the emission control area is restricted to having a maximum of one UAV deployed for emission monitoring within a single planning period. (6) In the formula, If the decision variable is 0 / 1, then the drone Along the directed arc To carry out a flight mission, ,otherwise ,in, , Let be the set of directed arcs that allow UAV flight in the spatiotemporal network of UAV cruise, representing the flight paths that UAVs can execute; A collection of drones for performing monitoring tasks; This refers to the set of all spatiotemporal nodes in the spatiotemporal network of UAV patrols. , It is a set of spatiotemporal nodes for maritime base stations. A set of spatiotemporal nodes for ships; For the target vessel The corresponding set of ship spatiotemporal nodes, among which, , A collection of target vessels to be monitored within the emission control area; (2) Single Exit Constraint: Each UAV is restricted to leaving the maritime base station and performing a patrol mission at most once within a single planning period. (7) (3) Traffic balance constraint: Each UAV must fly to the next target vessel or fly directly back to the maritime base station after completing the escort sampling upon reaching the target vessel. (8) In the formula, Indicates if drone Along the directed arc To carry out a flight mission, ,otherwise ; (4) State constraints corresponding to maritime base stations: By constraining whether each UAV is within a maritime base station at each time step, the continuity between the UAV's standby, departure, and return states is ensured: (9) In the formula, If the decision variable is 0-1, then the drone At time step Located at a maritime base station Waiting inside, then ,otherwise ,in, , A collection of maritime base stations for deploying drones. , This is the set of discretized planning time steps. express Maritime base station This point in time and space; (5) Battery power recursive constraint: When the UAV flies along a directed arc, the actual energy consumption under the corresponding relative wind direction is deducted. By introducing a sufficiently large positive number M, the energy consumption of each UAV is constrained to be deducted only when it is actually flying along the directed arc: (10) In the formula, Let be a non-negative continuous decision variable, representing the unmanned aerial vehicle (UAV). Arrival at the spacetime node The remaining battery power at that time Indicates drone Arrival at the spacetime node The remaining battery power at that time; (6) Initial power constraint: Each UAV is constrained to be in a fully charged state at the initial moment before exiting the cabin: (11) In the formula, For drones Maximum rated battery capacity: (7) Power safety upper and lower limit constraints: The remaining power of each UAV at any node should be between zero and the rated capacity.

[0040] After obtaining the optimal solution from the objective function, the mathematical solution needs to be transformed into a practically executable drone navigation plan. In implementation, the drone navigation path is optimized using the following methods: Step 301: Extract the selected spatiotemporal nodes and directed arcs for each UAV, and arrange the spatiotemporal nodes according to the time sequence to obtain the cruise path of each UAV; the cruise path consists of the starting spatiotemporal node of the maritime base station, the sequentially visited ship spatiotemporal nodes, and the returning spatiotemporal node of the maritime base station.

[0041] Step 302: Based on the cruise path, output the actual energy consumption of each drone during each flight segment, the remaining power at each spatiotemporal node, and the remaining power after returning to base, to ensure that the drone can return safely.

[0042] Step 303: Output whether each drone is on standby at the maritime base station at each time step, thus providing a reference for drone scheduling.

[0043] The above information is integrated into a patrol plan for maritime regulatory authorities, including UAV departure time, return time, target vessel visit time, patrol route, wind direction and conditions along the route, flight energy consumption, remaining battery power at each node, and total monitoring revenue. This result can be directly used by maritime regulatory authorities to direct multiple UAVs to conduct ship emission monitoring operations in complex sea conditions.

[0044] In summary, a mixed-integer linear programming model is constructed to maximize the total monitoring benefits of the UAV fleet in a single patrol mission. Variables for UAV flight arc selection, station standby status, and node remaining battery power are set. Under constraints such as each ship being monitored at most once, each UAV performing at most one out-of-cabin mission during the planning period, UAV paths satisfying flow conservation, station state continuity, and battery power not exceeding rated capacity and decreasing with flight energy consumption, the optimal patrol path for each UAV is solved. This model can be globally solved accurately using Python combined with the Gurobi optimization solver, outputting a patrol plan for the UAVs starting from the station, sequentially visiting the spatiotemporal nodes of the target ships, and safely returning to the station.

[0045] To more intuitively observe the dynamic tracking process of the drone on the time axis and spatial coordinates, a graph was drawn based on specific path node information, as shown below. Figure 3 The diagram shows the cruise paths of multiple drones. The Z-axis represents the time step, and the X and Y axes represent the actual geographic coordinates of the sea surface, illustrating the cruise paths of multiple drones within a spatiotemporal network. Furthermore, [the diagram will also include...] Figure 3 The three-dimensional cruise path is projected onto the two-dimensional sea surface to obtain the following: Figure 4 The diagram shown illustrates the drone's cruise path projection, used to display the mission area and cruise route of each drone.

[0046] The above content is only a preferred embodiment of the present invention. For those skilled in the art, many changes can be made in the specific implementation and application scope based on the ideas of the present invention. As long as these changes do not depart from the concept of the present invention, they all fall within the protection scope of the present invention.

Claims

1. A method for planning unmanned aerial vehicle (UAV) cruise paths for ship emission monitoring under complex sea conditions, characterized in that, The method includes: A spatiotemporal network for UAV patrol is constructed by ship spatiotemporal nodes, maritime base station spatiotemporal nodes, and directed arcs that allow UAV flight. The directed arc refers to the ability of a UAV to depart from the time and space location corresponding to one spatiotemporal node and accurately arrive at its spatial location at the time corresponding to another spatiotemporal node. For any directed arc in the UAV's cruise spatiotemporal network, calculate the actual energy consumption of the UAV flying along the directed arc in the relative wind direction; Determine the actual monitoring benefits of UAV visits to various spatiotemporal nodes under sea fog conditions; The goal is to maximize the total actual monitoring benefits obtained by the drone swarm in a single patrol. At the same time, corresponding constraints are set for the target ship, drones, maritime base stations, and drone battery power based on actual energy consumption. Given battery capacity, time window, wind direction and sea fog conditions, the drone patrol path is optimized. Finally, the drone departure time, return time, target ship visit time, patrol path, relative wind direction of the flight segment, flight energy consumption, node remaining power and total actual monitoring benefits are output.

2. The method as described in claim 1, characterized in that, Before constructing the UAV cruise spatiotemporal network jointly by ship spatiotemporal nodes, maritime base station spatiotemporal nodes, and feasible directed flight arcs of the UAV, the method further includes: Discretization is performed for single short-term planning periods, and the continuous planning time is divided into several equal-length steps; Based on the target ship's initial position, heading, and speed, the spatial position of the ship at each time step is deduced, and spatiotemporal nodes of the ship with time tags are generated. Based on the fixed spatial location of the maritime base station and its state at each time step, a spatiotemporal node of the maritime base station is generated, which serves as both the departure node and the return node of the drone. Based on the maximum flight speed of the UAV, the spatial distance between each spatiotemporal node, and the time tag, it is determined whether there is a feasible directed arc for UAV flight between each spatiotemporal node; the spatiotemporal nodes include ship spatiotemporal nodes and maritime base station spatiotemporal nodes.

3. The method as described in claim 1, characterized in that, The calculation of the actual energy consumption of the UAV flying along the oriented arc in relative wind direction includes: For any directed arc in the spatiotemporal network of the UAV cruise The heading angle of the UAV flying along the directed arc is calculated based on the spatial coordinates of the two spatiotemporal nodes: ； In the formula, For the drone's heading angle, Representing spatiotemporal nodes coordinates Represents the coordinates of spatiotemporal node j; Obtain the ambient wind direction angle and calculate its relative angle with the UAV's flight heading angle, then map this relative angle to... Within the range: ； In the formula, For the ambient wind direction angle; Based on the relative angle after mapping Flight status is divided into three relative wind directions: tailwind, crosswind, and headwind, and corresponding segmented energy consumption coefficients are set. ; Based on segmented energy consumption coefficients Calculate the UAV along the directed arc Actual energy consumption of flight: ； In the formula, This represents the baseline power consumption per unit time under windless conditions. spatiotemporal nodes To spacetime node The flight time span.

4. The method as described in claim 3, characterized in that, The flight state is divided into three relative wind directions—tailwind, crosswind, and headwind—based on the mapped relative angle, and corresponding segmented energy consumption coefficients are set, including: like This is a tailwind condition, where the wind provides a boost to the drone's flight. ; like This is a crosswind condition, where the drone needs to consume additional energy to correct its yaw. ; like This is a headwind operating condition, where the drone needs to overcome wind resistance and set... Much greater than 1.

5. The method as described in claim 1, characterized in that, The determination of the actual monitoring benefits of UAV visits to various spatiotemporal nodes under sea fog conditions includes: Based on the vessel's tonnage, fuel type, main engine power, and historical violation records, basic monitoring benefits are assigned to each target vessel; When any ship spatiotemporal node falls into a pre-defined sea fog coverage area at a specific time step, a sea fog attenuation coefficient is assigned to the ship spatiotemporal node based on the fog concentration, so that the monitoring benefit attenuation mechanism is automatically triggered when the UAV visits the ship spatiotemporal node. The actual monitoring benefits of the UAV visiting the ship's spatiotemporal nodes are determined by the ship's basic monitoring benefits and the sea fog attenuation coefficient. And the actual monitoring benefit of setting the spatiotemporal node of the maritime base station is 0.

6. The method as described in claim 1, characterized in that, The constraints set for the target ship, drone, maritime base station, and drone battery power based on actual energy consumption include: (1) Each target vessel within the constrained emission control area shall be assigned a maximum of one UAV for emission monitoring during a single planning period: ； In the formula, If the decision variable is 0 / 1, then the drone Along the directed arc To carry out a flight mission, ,otherwise ,in, , Let be the set of directed arcs that allow UAV flight in the spatiotemporal network of UAV cruise, representing the flight paths that UAVs can execute; A collection of drones for performing monitoring tasks; This refers to the set of all spatiotemporal nodes in the spatiotemporal network of UAV patrols. , It is a set of spatiotemporal nodes for maritime base stations. A set of spatiotemporal nodes for ships; For the target vessel The corresponding set of ship spatiotemporal nodes, among which, , A collection of target vessels to be monitored within the emission control area; (2) Each UAV is restricted to leaving the maritime base station and performing only one patrol mission within a single planning period: ； (3) Each UAV must fly to the next target vessel or fly directly back to the maritime base station after completing the escort sampling upon reaching the target vessel: ； In the formula, Indicates if drone Along the directed arc To carry out a flight mission, ,otherwise ; (4) By constraining whether each UAV is within the maritime base station at each time step, the continuity between the UAV's standby, departure, and return states is ensured: ； In the formula, If the decision variable is 0-1, then the drone At time step Located at a maritime base station Waiting inside, then ,otherwise ,in, , A collection of maritime base stations for deploying drones. , This is the set of discretized planning time steps. express Maritime base station This point in time and space; (5) When the UAV flies along a directed arc, the actual energy consumption under the corresponding relative wind direction is deducted. By introducing a sufficiently large positive number M, the energy consumption of each UAV is constrained to be deducted only when it actually flies along the directed arc: ； In the formula, Let be a non-negative continuous decision variable, representing the unmanned aerial vehicle (UAV). Arrival at the spacetime node The remaining battery power at that time Indicates drone Arrival at the spacetime node The remaining battery power at that time; (6) Constrain each UAV to be in a fully charged state at the initial moment before exiting the cabin: ； In the formula, For drones Maximum rated battery capacity: (7) Constrain the remaining power of each UAV at any node to be between zero and the rated capacity.

7. The method as described in claim 1, characterized in that, The optimization of the drone's cruise path includes: Extract the selected spatiotemporal nodes and directed arcs for each UAV, and arrange the spatiotemporal nodes according to the time sequence to obtain the cruise path of each UAV; the cruise path consists of the starting spatiotemporal node of the maritime base station, the sequentially visited ship spatiotemporal nodes, and the returning spatiotemporal node of the maritime base station. Based on the cruise path, the actual energy consumption of each drone during each flight segment, the remaining power at each spatiotemporal node, and the remaining power after returning home are output to ensure that the drone can return home safely. It outputs whether each drone is on standby at the maritime base station at each time step, thus providing a reference for drone scheduling.

Images