Photovoltaic inspection method and system based on aircraft-boat cooperation

Abstract

The invention discloses a photovoltaic inspection method based on aircraft-boat cooperation. The photovoltaic inspection method comprises the following steps: dividing a photovoltaic panel in an inspection sea area into k inspection areas based on the number k of unmanned aerial vehicles for inspection; determining the inspection sequence of each photovoltaic panel in the inspection area; respectively determining a take-off point and a recovery point of the unmanned aerial vehicle based on the inspection starting point and the inspection terminal point of each inspection area; an unmanned plane is carried on an unmanned ship, the unmanned ship runs from a starting position to a take-off point based on a planned path, the unmanned plane is released at the take-off point, the unmanned plane flies from the take-off point to an inspection starting point, flies from the inspection starting point to an inspection terminal point and flies from the inspection terminal point to a recovery point based on a planned flight path, and the unmanned ship sails from the take-off point to the recovery point at the same time. The release point and the recovery point of the unmanned ship are dynamically determined based on the inspection start point and the inspection end point of the unmanned aerial vehicle group, the flight distance of the unmanned aerial vehicle group flying back to the parking point is reduced as much as possible, and the task execution efficiency of the unmanned aerial vehicle group is greatly improved.

Description

Technical Field

[0001] This invention relates to the field of unmanned driving, and more specifically, to a photovoltaic inspection method and system based on ship-vehicle collaboration. Background Technology

[0002] With the large-scale construction of floating photovoltaic panels, their characteristics such as wide distribution, complex environment, and high difficulty in inspection are becoming increasingly prominent.

[0003] Traditional manual inspection methods are inefficient and risky. Currently, most inspections are conducted using drones. The drone inspection methods for water-surface photovoltaic power stations include the following:

[0004] (1) Single UAV inspection: After performing the inspection task, the single UAV flies back to the fixed parking point to recharge. After recharging, it repeatedly performs the inspection task.

[0005] (2) Multiple drones perform inspection operations simultaneously. The inspection routes of multiple drones are planned based on the principle of the shortest overall flight distance. After completing the inspection task of the corresponding inspection route, a single drone flies back to a fixed parking point to recharge.

[0006] The aforementioned drone inspection methods are all limited by the short battery life of drones. When the battery is low, they need to return to a fixed docking point for charging, resulting in a low task completion rate. Summary of the Invention

[0007] This invention provides a photovoltaic inspection method based on ship-vehicle collaboration, which aims to improve at least one of the above-mentioned problems.

[0008] This invention is implemented as follows: a photovoltaic inspection method based on aircraft-ship collaboration, the method is as follows:

[0009] (1) Based on the number of drones used for inspection, k, the photovoltaic panels in the inspection area are divided into k clusters. The photovoltaic panels in the cluster form an inspection area. Each inspection area is assigned to a drone, and one drone corresponds to one inspection area.

[0010] (2) Determine the inspection sequence of each photovoltaic panel in the inspection area, take the location of the first photovoltaic panel in the inspection sequence as the inspection start point of the corresponding inspection area, and take the location of the last photovoltaic panel in the inspection sequence as the inspection end point of the corresponding inspection area.

[0011] (3) Determine the take-off point of the UAV based on the inspection start point of each inspection area, and determine the UAV recovery point based on the inspection end point of each inspection area;

[0012] (4) Plan the travel path of the unmanned surface vessel from the starting position to the drone take-off point, and from the drone take-off point to the drone recovery point; plan the flight path of the drone from the drone take-off point to the inspection start point, from the inspection start point to the inspection end point, and from the inspection end point to the drone recovery point.

[0013] (5) The UAV is mounted on the unmanned surface vessel. The unmanned surface vessel travels from the starting position to the take-off point according to the planned path. The UAV on the unmanned surface vessel is released at the take-off point. The UAV flies from the take-off point to the inspection start point according to the planned flight path. From the inspection start point to the inspection end point, the inspection end point flies to the recovery point. At the same time, the unmanned surface vessel travels from the take-off point to the recovery point.

[0014] Furthermore, the method for determining the inspection sequence of photovoltaic panels within the inspection area is as follows:

[0015] Chromosomes are encoded based on various inspection sequences of photovoltaic panels within the inspection area. The quality of each chromosome is evaluated based on fitness. The inspection sequence of photovoltaic panels corresponding to the optimal chromosome degree is taken as the inspection sequence of photovoltaic panels in the corresponding inspection area. The specific formula for calculating the fitness of the c-th chromosome is as follows:

[0016] ;

[0017] in, , Indicates the weight value; This represents the energy consumption required for the inspection sequence corresponding to the c-th chromosome. This indicates the smoothness of the inspection sequence corresponding to chromosome c.

[0018] Furthermore, the specific method for determining the takeoff point of the drone is as follows:

[0019] (31) Extract the inspection starting point of each inspection area and construct the smallest rectangular area containing all inspection starting points;

[0020] (32) Sampling of candidate points within the rectangular area;

[0021] (33) Calculate the sum of the distances from all inspection starting points to each candidate point, and take the candidate point with the smallest sum of distances as the take-off point of the UAV.

[0022] Furthermore, the specific method for determining drone recovery points is as follows:

[0023] (31) Extract the inspection endpoints of each inspection area and construct the smallest rectangular area containing all inspection endpoints;

[0024] (32) Sampling of candidate points within the rectangular area;

[0025] (33) Calculate the sum of the distances from all inspection endpoints to each candidate point, and take the candidate point with the smallest sum of distances as the UAV recovery point.

[0026] Furthermore, the unmanned surface vessel's (USV) travel path is planned using the artificial potential field method, and the resultant potential field is expressed as follows:

[0027] ;

[0028] in, Indicates the location of the unmanned surface vessel. The gravitational field at that location, Indicates the location of the unmanned surface vessel. The repulsive field at a given location, where the gravitational field is represented as follows:

[0029] ;

[0030] in, This indicates the set gravitational field gain coefficient. Indicates the target location. This indicates the distance between the current position of the unmanned surface vessel and the target position.

[0031] Furthermore, the repulsive field is represented as follows:

[0032] ;

[0033] in, These represent the repulsive potential field generated by a static obstacle, the time-varying repulsive potential field generated by a dynamic obstacle, and the flow resistance potential field, respectively.

[0034] Furthermore, the flight path planning methods are all the same, as follows:

[0035] Dynamically sample multiple velocity pairs in velocity space ,in, It measures the linear and angular velocities of the drone and simulates and predicts multiple candidate flight paths for the next time window;

[0036] Each candidate flight path is scored, and the candidate path with the highest score is selected as the flight path for the next time window.

[0037] Furthermore, the score of the i-th candidate flight path The specific calculation formula is as follows:

[0038] ;

[0039] in, , , All are weighting coefficients. This represents the alignment between the flight direction of the endpoint of the i-th candidate flight path and the direction of the line connecting the endpoint of the path and the target position. This represents the safety score of the i-th candidate flight path. Indicates the energy consumption of the i-th candidate flight path .

[0040] This invention is implemented as follows: a photovoltaic inspection system based on ship-vehicle collaboration, the system comprising:

[0041] One unmanned surface vessel and m unmanned aerial vehicles, where m is greater than or equal to k;

[0042] The control center, which is remotely connected to the unmanned surface vessel (USV) and unmanned aerial vehicle (UAV), controls the USV and UAV to perform inspection tasks in a coordinated manner based on the aforementioned photovoltaic inspection method of ship-vessel collaboration.

[0043] This invention uses unmanned surface vessels (USVs) as mobile docking points. It first determines the optimal inspection sequence for each inspection area, and then dynamically determines the release and recovery points of the USVs based on the inspection start and end points of the USV swarm. This minimizes the flight distance from the release point to the inspection start point and from the inspection end point to the recovery point, greatly improving the mission execution efficiency of the USV swarm. Attached Figure Description

[0044] Figure 1 A flowchart illustrating a photovoltaic inspection method based on ship-vehicle collaboration provided in an embodiment of the present invention;

[0045] Figure 2 This is a schematic diagram illustrating the inspection sequence of photovoltaic panels within an inspection area, provided in an embodiment of the present invention.

[0046] Figure 3 This is a schematic diagram of unmanned surface vessel path planning based on artificial potential field method provided in an embodiment of the present invention, wherein (a) is a schematic diagram of the unmanned surface vessel's travel path in plan view and (b) is a schematic diagram of the path with potential field in three-dimensional view.

[0047] Figure 4 This is a schematic diagram of the flight path planning of an unmanned aerial vehicle (UAV) provided in an embodiment of the present invention.

[0048] Figure 5 This is a schematic diagram of the coordination planning of unmanned ships and unmanned surface vessels provided in an embodiment of the present invention. Detailed Implementation

[0049] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings, so as to help those skilled in the art to have a more complete, accurate and in-depth understanding of the inventive concept and technical solution of the present invention.

[0050] Unmanned surface vessels (USVs) are capable of long-term surface operations and can serve as mobile take-off and landing platforms and charging stations for drones. Drones, on the other hand, are flexible, maneuverable, and have a superior field of view, enabling them to acquire high-definition images and infrared data of photovoltaic panels at close range. Combining the two forms a "USV-drone collaborative" intelligent inspection system, which can achieve efficient, accurate, and automated inspection of large-area photovoltaic panels.

[0051] Figure 1 A flowchart of a photovoltaic inspection method based on ship-vehicle collaboration provided in an embodiment of the present invention is shown below:

[0052] (1) Based on the number of drones used for inspection, k, the photovoltaic panels in the inspection area are divided into k clusters. The photovoltaic panels in the cluster form an inspection area. Each inspection area is assigned to a drone, with one drone corresponding to one inspection area.

[0053] In this embodiment of the invention, the sea area to be inspected is divided into several inspection sea areas. After the UAV completes the inspection task in the current inspection sea area, it flies back to the unmanned surface vessel (USV) to replenish the power of the UAV when it is low on power. The UAV's status includes: UAV in charging state, UAV in fault state, and UAV in normal state. UAV in normal state is used to perform the inspection task in the next inspection sea area.

[0054] In this embodiment of the invention, based on The clustering algorithm clusters the photovoltaic panels in the current inspection area. If there are k drones on the unmanned surface vessel for inspection, the number of clusters is set to k, and the photovoltaic panels in the current inspection area form k clusters.

[0055] (2) Determine the inspection sequence of each photovoltaic panel in the inspection area, take the location of the first photovoltaic panel in the inspection sequence as the inspection start point of the corresponding inspection area, and take the location of the last photovoltaic panel in the inspection sequence as the inspection end point of the corresponding inspection area.

[0056] In this embodiment of the invention, a genetic algorithm is used to determine the inspection order of each inspection panel in each inspection area. The genetic algorithm is applied to the photovoltaic panels within a cluster to solve for the optimal inspection order in each inspection area. Taking the Kth cluster as an example, this cluster contains n photovoltaic panels, which are numbered as follows: Each chromosome is coded using a photovoltaic panel inspection sequence. Different chromosomes have different codes, meaning that the inspection sequence of each photovoltaic panel in the inspection area is different. The quality of each chromosome is evaluated based on fitness, and the photovoltaic panel inspection sequence corresponding to the optimal chromosome degree is finally used as the photovoltaic panel inspection sequence for the corresponding inspection area.

[0057] In this embodiment of the invention, the formula for calculating the fitness of the c-th chromosome is as follows:

[0058] ;

[0059] in, , Indicates the weight value; Let c represent the energy consumption required for the inspection sequence corresponding to the c-th chromosome, and calculate the Euclidean distance of the UAV flight required for the inspection sequence corresponding to the c-th chromosome. European distance The larger it is, the more energy it requires. The larger; To demonstrate the smoothness of the inspection sequence corresponding to the c-th chromosome, using photovoltaic panels as stations, the turning angle of each station in the inspection sequence corresponding to the c-th chromosome is calculated. Turning angles greater than a turning angle threshold are extracted, and the extracted turning angles are summed to form the total turning angle of the inspection sequence corresponding to the c-th chromosome. General steering The smaller the value, the smoother the surface. The larger.

[0060] (3) Determine the take-off point of the UAV based on the inspection start point of each inspection area, and determine the UAV recovery point based on the inspection end point of each inspection area;

[0061] In this embodiment of the invention, the method for determining the takeoff point of the UAV is as follows:

[0062] (31) Extract the inspection starting point of each inspection area and construct the smallest rectangular area containing all inspection starting points;

[0063] (32) Sampling of candidate points within the rectangular area;

[0064] (33) Calculate the sum of the distances from all inspection starting points to each candidate point, and take the candidate point with the smallest sum of distances as the take-off point of the UAV.

[0065] In this embodiment of the invention, the method for determining the drone recovery point is as follows:

[0066] (31) Extract the inspection endpoints of each inspection area and construct the smallest rectangular area containing all inspection endpoints;

[0067] (32) Sampling of candidate points within the rectangular area;

[0068] (33) Calculate the sum of the distances from all inspection endpoints to each candidate point, and take the candidate point with the smallest sum of distances as the UAV recovery point.

[0069] (4) Plan the travel path of the unmanned surface vessel from the starting position to the drone take-off point, and from the drone take-off point to the drone recovery point; plan the flight path of the drone from the drone take-off point to the inspection start point, from the inspection start point to the inspection end point, and from the inspection end point to the drone recovery point.

[0070] In this embodiment of the invention, the current position of the unmanned surface vessel (USV) is taken as the starting position. The current position of the USV is the recovery point of the UAV in the previous inspection area. If the current inspection area is the first inspection area in the inspection process, then the current position of the USV is the docking point of the USV on the shore.

[0071] It should be noted that the method for planning the unmanned surface vessel's (USV) path from its starting position to the UAV's takeoff point is the same as the method for planning the USV's path from the UAV's takeoff point to the UAV's recovery point. This invention plans the above-mentioned USV path using the artificial potential field method, as detailed below:

[0072] In this embodiment of the invention, the combined potential field is expressed as follows: ;

[0073] Indicates the location of the unmanned surface vessel. The gravitational field at that location, Indicates the location of the unmanned surface vessel. The repulsive field at a given location, where the gravitational field is represented as follows:

[0074] ;

[0075] in, This indicates the set gravitational field gain coefficient. , Indicates the target location. This indicates the distance between the current position of the unmanned surface vessel and the target position. If the route is planned from the starting position to the drone's takeoff point, the target position is set as the drone's takeoff point. If the route is planned from the drone's takeoff point to the drone's recovery point, the target position is set as the drone's recovery point.

[0076] In this embodiment of the invention, the repulsive field is represented as follows:

[0077] ;

[0078] in, These represent the repulsive potential field generated by a static obstacle, the time-varying repulsive potential field generated by a dynamic obstacle, and the flow resistance potential field, respectively. Time-varying repulsive potential field generated by dynamic obstacles Flow resistance potential field The specific expression is as follows:

[0079] ;

[0080] in, This represents the repulsive field gain coefficient. Indicates the location of the unmanned surface vessel. Repulsive force This indicates the location of the i-th static obstacle on the map. repulsive force, Indicates the location of the unmanned surface vessel. The distance from the i-th static obstacle. This indicates the distance at which the repulsive field acts.

[0081] ;

[0082] in, This indicates the location of the j-th dynamic obstacle on the map. repulsive force, Indicates the location of the unmanned surface vessel. The distance from the j-th dynamic obstacle;

[0083] ;

[0084] in, Indicates the current heading of the unmanned surface vessel. Indicates the current position of the unmanned surface vessel. To the target location The course.

[0085] In this embodiment of the invention, the flight path planning methods from the UAV takeoff point to the inspection start point, from the inspection start point to the inspection end point, and from the inspection end point to the UAV recovery point are all the same, as follows:

[0086] Dynamically sample multiple velocity pairs in velocity space ,in, It measures the linear and angular velocities of the drone and simulates and predicts multiple candidate flight paths for the next time window;

[0087] Each candidate flight path is scored, and the candidate path with the highest score is selected as the flight path for the next time window.

[0088] In this embodiment of the invention, the score of the i-th candidate flight path The specific calculation formula is as follows:

[0089] ;

[0090] in, , , All are weighting coefficients. Let represent the alignment between the flight direction of the endpoint of the i-th candidate flight path and the direction of the line connecting the endpoint of the path and the target position. Calculate the angle between the flight direction of the endpoint of the i-th candidate flight path and the direction of the line connecting the endpoint of the path and the target position. included angle The smaller the value, the better the alignment. The larger; Represent the safety score of the i-th candidate flight path, and calculate the closest distance between the i-th candidate flight path and the obstacle. The candidate flight path consists of a series of flight points. The distance from each flight point on the candidate flight path to the obstacle is calculated, and the minimum distance is taken as the distance from the corresponding candidate trajectory to the obstacle. Obstacles include static and dynamic obstacles. When the minimum distance is less than the set safety distance, the smaller the minimum distance, the higher the corresponding safety score. The smaller; Indicates the energy consumption of the i-th candidate flight path Since the motor power of each drone is a constant during flight, the longer the flight distance corresponding to the i-th candidate flight path, the higher the score. The lower.

[0091] (5) The UAV is mounted on the unmanned surface vessel (USV). The USV travels from the starting position to the takeoff point according to the planned path. The UAV is released from the USV at the takeoff point. The UAV flies from the takeoff point to the inspection start point according to the planned flight path. From the inspection start point to the inspection end point, the inspection end point flies to the recovery point. At the same time, the USV travels from the takeoff point to the recovery point. The time it takes for the USV to travel from the takeoff point to the recovery point is shorter than the time it takes for the UAV to travel from the takeoff point to the recovery point.

[0092] This invention uses unmanned surface vessels (USVs) as mobile docking points. It first determines the optimal inspection sequence for each inspection area, and then dynamically determines the release and recovery points of the USVs based on the inspection start and end points of the USV swarm. This minimizes the flight distance from the USV release point to the inspection start point and from the inspection end point to the recovery point, greatly improving the mission execution efficiency of the USV swarm.

[0093] This invention also provides a photovoltaic inspection system based on ship-machine collaboration, the system comprising:

[0094] One unmanned surface vessel and m unmanned aerial vehicles, where m is greater than or equal to k;

[0095] The control center, which is remotely connected to the unmanned surface vessel (USV) and unmanned aerial vehicle (UAV), controls the USV and UAV to perform inspection tasks in a coordinated manner based on the aforementioned photovoltaic inspection method of ship-vessel collaboration.

[0096] The effectiveness of the method described in this invention was verified using the MATLAB R2022b simulation platform. The simulation experiment was conducted on a computer configured with an Intel Core i5-12600KF processor and 32GB of RAM.

[0097] Simulation scenario construction: In a 100×100m two-dimensional water simulation environment, m=10 photovoltaic arrays were randomly deployed. The maximum forward speed of the unmanned surface vessel (USV) was set to 1.5 m / s, and the maximum angular velocity was set to 0.8 rad / s. The UAV model was simplified to an omnidirectional moving point on a two-dimensional plane, and its cruising speed was set to 5 m / s to simulate its high maneuverability relative to the USV.

[0098] A genetic algorithm was used to solve the problem. The population size was set to 200, the number of iterations was 500, the crossover probability was 0.8, and the mutation probability was 0.05. The inspection order of the inspection area corresponding to each cluster was as follows: Figure 2 As shown, the current patrol area is divided into 5 clusters (represented by five colors), and an optimized access sequence is generated for each drone.

[0099] Artificial potential field method: The gravitational field gain coefficient is set to The repulsive field gain coefficient is set to The repulsive field's effective distance is set to Unmanned surface vessel path planning based on artificial potential field method, such as Figure 3 As shown, (a) is a planar schematic diagram of the unmanned surface vessel's (USV) path, with green dots representing static obstacles, and (b) is a three-dimensional schematic diagram of the path with a force field. Figure 3 The demonstration showed that the unmanned surface vessel (USV) generated a smooth and collision-free path under the guidance of the combined potential field (gravity and repulsion), safely avoided all static obstacles (green dots), and finally successfully arrived at the recovery point (blue).

[0100] Dynamic window method: The velocity sampling resolutions are respectively Simulated trajectory time length Five drones took off simultaneously from the release point and began their inspection tasks according to their assigned inspection areas and order. During the flight, each drone operated independently. The algorithm performs local obstacle avoidance and online planning, such as Figure 4 As shown, black dots represent obstacles, and red dots represent target locations. When the drone's sensors detect dynamic obstacles, The algorithm simulates multiple future trajectories in real time and quickly selects the optimal candidate path based on the evaluation function to avoid dynamic obstacles while maintaining the forward trend toward the target point.

[0101] When the first drone is detected to have completed more than 80% of its mission, the dynamic calculation of the recovery point location begins. After receiving the recovery point location, the unmanned surface vessel plans its flight path from the current takeoff point to the recovery point. After completing their respective inspection tasks, all drones use the recovery point as their target location and plan their flight paths from the inspection endpoint to the target location, ultimately forming... Figure 5 The diagram shows a coordinated planning scheme for unmanned ships and unmanned surface vessels.

[0102] The present invention has been described by way of example. Obviously, the specific implementation of the present invention is not limited to the above-described manner. Any non-substantial improvements made using the inventive concept and technical solution of the present invention, or the direct application of the inventive concept and technical solution of the present invention to other occasions without modification, are all within the protection scope of the present invention.

Claims

1. A photovoltaic inspection method based on ship-machine collaboration, characterized in that, The method is as follows: (1) Based on the number of drones used for inspection, k, the photovoltaic panels in the inspection area are divided into k inspection areas, and each inspection area is assigned to a drone. (2) Determine the inspection sequence of each photovoltaic panel in the inspection area; (3) Determine the take-off point of the UAV based on the inspection start point of each inspection area, and determine the UAV recovery point based on the inspection end point of each inspection area; (4) Plan the travel path of the unmanned surface vessel from the starting position to the drone take-off point, and from the drone take-off point to the drone recovery point; plan the flight path of the drone from the drone take-off point to the inspection start point, from the inspection start point to the inspection end point, and from the inspection end point to the drone recovery point. (5) The UAV is mounted on the unmanned surface vessel. The unmanned surface vessel travels from the starting position to the take-off point according to the planned path. The UAV on the unmanned surface vessel is released at the take-off point. The UAV flies from the take-off point to the inspection start point according to the planned flight path. From the inspection start point to the inspection end point, the inspection end point flies to the recovery point. At the same time, the unmanned surface vessel travels from the take-off point to the recovery point.

2. The photovoltaic inspection method based on ship-vehicle collaboration as described in claim 1, characterized in that, The method for determining the inspection sequence of photovoltaic panels within the inspection area is as follows: Chromosomes are encoded based on various inspection sequences of photovoltaic panels within the inspection area. The quality of each chromosome is evaluated based on fitness. The inspection sequence of photovoltaic panels corresponding to the optimal chromosome degree is taken as the inspection sequence of photovoltaic panels in the corresponding inspection area. The specific formula for calculating the fitness of the c-th chromosome is as follows: ； in, , Indicates the weight value; This represents the energy consumption required for the inspection sequence corresponding to the c-th chromosome. This indicates the smoothness of the inspection sequence corresponding to chromosome c.

3. The photovoltaic inspection method based on ship-machine collaboration as described in claim 1, characterized in that, The specific method for determining the takeoff point of a drone is as follows: (31) Extract the inspection starting point of each inspection area and construct the smallest rectangular area containing all inspection starting points; (32) Sampling of candidate points within the rectangular area; (33) Calculate the sum of the distances from all inspection starting points to each candidate point, and take the candidate point with the smallest sum of distances as the take-off point of the UAV.

4. The photovoltaic inspection method based on ship-vehicle collaboration as described in claim 1, characterized in that, The specific method for determining drone recovery points is as follows: (31) Extract the inspection endpoints of each inspection area and construct the smallest rectangular area containing all inspection endpoints; (32) Sampling of candidate points within the rectangular area; (33) Calculate the sum of the distances from all inspection endpoints to each candidate point, and take the candidate point with the smallest sum of distances as the UAV recovery point.

5. The photovoltaic inspection method based on ship-vehicle collaboration as described in claim 1, characterized in that, The unmanned surface vessel's (USV) path is planned using the artificial potential field method, and the resultant potential field is expressed as follows: ； in, Indicates the location of the unmanned surface vessel. The gravitational field at that location, Indicates the location of the unmanned surface vessel. The repulsive field at a given location, where the gravitational field is represented as follows: ； in, This indicates the set gravitational field gain coefficient. Indicates the target location. This indicates the distance between the current position of the unmanned surface vessel and the target position.

6. The photovoltaic inspection method based on ship-vehicle collaboration as described in claim 5, characterized in that, The repulsive field is represented as follows: ； in, These represent the repulsive potential field generated by a static obstacle, the time-varying repulsive potential field generated by a dynamic obstacle, and the flow resistance potential field, respectively.

7. The photovoltaic inspection method based on ship-machine collaboration as described in claim 1, characterized in that, The flight path planning methods are all the same, as follows: Dynamically sample multiple velocity pairs in velocity space ,in, It measures the linear and angular velocities of the drone and simulates and predicts multiple candidate flight paths for the next time window; Each candidate flight path is scored, and the candidate path with the highest score is selected as the flight path for the next time window.

8. The photovoltaic inspection method based on ship-vehicle collaboration as described in claim 7, characterized in that, Score of the i-th candidate flight path The specific calculation formula is as follows: ； in, , , All are weighting coefficients. This represents the alignment between the flight direction of the endpoint of the i-th candidate flight path and the direction of the line connecting the endpoint of the path and the target position. This represents the safety score of the i-th candidate flight path. Indicates the energy consumption of the i-th candidate flight path .

9. A photovoltaic inspection system based on ship-machine collaboration, characterized in that, The system includes: One unmanned surface vessel and m unmanned aerial vehicles, where m is greater than or equal to k; A control center that is remotely connected to unmanned surface vessels (USVs) and unmanned aerial vehicles (UAVs) controls USVs and UAVs to collaboratively perform inspection tasks based on the photovoltaic inspection method based on ship-vessel collaboration as described in any one of claims 1 to 8.

Images