Multi-unmanned ship oil spill cleaning path planning method combined with SAR (Synthetic Aperture Radar) image
By combining SAR images and YOLOv8 models to identify oil spill areas, GNOME models predict oil spill diffusion, and using geodesic-Manhattan Voronoi division and STC algorithm for path planning, the coverage and accuracy of unmanned boat path planning methods are solved. The defects of the ocean oil spill have been achieved efficient cleaning.
Patent Information
- Application Number
- CN202411176973.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-26
- Publication Date
- 2024-11-26
AI Technical Summary
The existing unmanned surface vehicle path planning methods have shortcomings in coverage and accuracy, especially when dealing with marine oil spills, it is difficult to fully cover the oil spill area and take into account the impact of natural environmental factors.
SAR images combined with YOLOv8 model to identify oil spill areas, GNOME model is used to predict oil spill diffusion, generate grid maps through Gaussian hybrid model, and use geodesic-Manhattan Voronoi division and STC algorithm for path planning to ensure the work of unmanned boats Accuracy of area division and path planning.
It improves the full coverage capacity and cleaning efficiency of unmanned boats for marine oil spill areas, and can accurately identify and handle oil spill areas in a dynamically changing marine environment, enhancing the accuracy and coverage of the cleaning process.
Smart Images

Figure CN119022934A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of multi-unmanned boat path planning, and in particular to a multi-unmanned boat oil spill cleaning path planning method combined with SAR images. Background Art
[0002] Marine oil spill refers to a large amount of oil entering the marine environment due to unexpected circumstances. Marine oil spills are a global environmental problem, and their causes are mainly attributed to two categories: one is the leakage of drilling platforms during oil exploration and development activities; the other is the oil spill caused by the stranding or collision of ships during oil transportation. Such incidents not only pose a direct threat to the marine ecosystem, but also have a significant negative impact on economic activities, fishery resources and tourism. With the advancement of science and technology, the application of multi-unmanned boat systems in the treatment of marine oil spills has shown great potential and prospects. Multi-unmanned boat systems can effectively navigate in complex marine environments and accurately perform oil spill treatment operations, avoiding the risks faced by people during the treatment process.
[0003] Reference [1] proposed a regional coverage path planning method combined with Voronoi partitioning. By combining the move to centroid algorithm (CVT algorithm) with distributed sliding mode control, the UAV is navigated to the centroid of the Voronoi sub-region. However, if the move to centroid algorithm is to be used for oil spill cleaning, it is not suitable. The CVT algorithm may cause the UAV to only cover the path to the centroid and ignore other possible oil spill areas, which limits the comprehensiveness of the cleaning work. Reference [2]: 2021 6th International Conference on Robotics and Automation Engineering (ICRAE). IEEE, 2021: 40-45. In order to meet the needs of unmanned boats cleaning up sea surface oil pollution under time constraints, the oil pollution image is processed to find the target hull point, and the path planning of the unmanned surface vehicle is carried out, but it does not consider the influence of natural environment such as wind, waves, and currents on the distribution of oil pollution.
[0004] The existing literature is as follows: Document [1]: 2019 7th International Conference on Robotics and Mechatronics (ICRoM). IEEE, 2019: 17-22. Document[2]: 2021 6th International Conference on Robotics and AutomationEngineering (ICRAE). IEEE, 2021: 40-45. Summary of the invention The purpose of the present invention is to provide a multi-unmanned boat oil spill cleaning path planning method combined with SAR images in order to solve the problem that the existing path planning methods for unmanned surface vehicles have poor coverage and accuracy.
[0005] The above-mentioned purpose of the present application is achieved through the following technical solutions: S1: Obtain SAR images and combine them with the YOLOv8 model to determine the marine oil spill area; S2: Input the coordinates of the marine oil spill area into the GNOME model, make diffusion predictions for the marine oil spill area, and obtain the prediction results of the oil spill diffusion distribution; S3: Based on the prediction results of oil spill diffusion distribution, a grid map is obtained using the Gaussian mixture model; S4: The geodesic-Manhattan Voronoi partitioning method is used in combination with the grid map to divide the working area of the unmanned boat; S5: Use the STC algorithm to plan the paths of the unmanned boats in the working area to obtain the cleaning paths of each unmanned boat.
[0006] Optionally, step S1 includes: Based on the acquired SAR images, a data set is constructed; the YOLOv8 model is trained using the data set to obtain a recognition and segmentation model; By identifying the segmentation model, the oil spill area is identified in the SAR image with geographic information to determine the marine oil spill area.
[0007] Optionally, step S3 includes: According to the prediction results of oil spill diffusion distribution, the Gaussian mixture model is used to calculate the probability density of each point in the marine oil spill area to obtain a probability density distribution map; Discretize the probability density distribution map into a grid map.
[0008] 4. A method for planning a multi-unmanned boat oil spill cleaning path in combination with SAR images as claimed in claim 1, characterized in that the Gaussian mixture model is composed of a convex combination of N Gaussian densities, and the mathematical expression is as follows:
[0009]
[0010] in represents the density function of the Gaussian mixture model, It is The weights of the Gaussian components, It is The mean vector of the Gaussian components, It is The covariance matrix of the Gaussian components; Indicates is the mean, is the probability density function of the multivariate Gaussian distribution of the covariance; yes The determinant of ; Represents the dimensionality of the data; Indicates the location of the data point.
[0011] Optionally, step S4 includes: Using geodesic-Manhattan Voronoi partitioning, The working area is divided into two parts: By using the ocean oil spill area in a two-dimensional plane According to the number of unmanned boats, generate Generators, Represents the position of the generator, and performs geodesic-Manhattan Voronoi partitioning based on the generator position, which is mathematically defined as:
[0012] in Represents the divided Voronoi region; Indicates the location of points within the Voronoi region; Represents the oil spill area in a two-dimensional plane; Represents a point within the Voronoi region and the corresponding region generator The geodesic-Manhattan distance of .
[0013] Optionally, the Voronoi region Center of gravity , Centroid The definition is as follows:
[0014]
[0015] in, Indicates the probability density value of each point in the probability density distribution map.
[0016] An electronic device comprises a processor, a memory, a user interface and a network interface, wherein the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes a method for planning a multi-unmanned boat oil spill cleaning path combined with SAR images.
[0017] A computer-readable storage medium stores instructions. When the instructions are executed, a method for planning a multi-unmanned boat oil spill cleaning path combined with SAR images is executed.
[0018] The beneficial effects of the technical solution provided by this application are: The high resolution and penetration ability of synthetic aperture radar SAR images are used to overcome the limitations of traditional images due to weather and other conditions. The Yolov8 image segmentation technology is used to accurately identify and locate oil pollution, providing accurate oil pollution location information for path planning. The GNOME model is used to simulate the impact of natural environmental factors such as wind, waves, and currents on the spread of oil pollution to adapt to the dynamic changes and distribution of marine oil spill areas and improve the accuracy of cleaning work. The geodesic-Manhattan Voronoi partitioning method is used in combination with grid maps to divide the working area of the unmanned boat, so that each unmanned boat can operate in its own partition, and the unmanned boat can fully cover the entire oil spill area to improve the work efficiency of the robot during the cleaning process. The STC algorithm is used for path planning to ensure that the unmanned boat can fully cover the entire oil spill area and improve the coverage rate of oil spill cleanup. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] The present application will be further described below with reference to the accompanying drawings and embodiments, in which: Figure 1 is a step diagram in an embodiment of the present application; Figure 2 It is a training result diagram of the Yolov8 model in the embodiment of the present application; Figure 3 is an oil spill area identification result diagram in an embodiment of the present application; Figure 4 It is the oil spill distribution result diagram in the embodiment of the present application; Figure 5 is a probability density map in an embodiment of the present application; Figure 6 is a working area result diagram in an embodiment of the present application; Figure 7 It is a visualization result diagram of the spanning tree in the embodiment of the present application; Figure 8 It is a visualization result diagram of the spanning tree in the embodiment of the present application; Fig. 9 It is a detour generation path map in an embodiment of the present application. DETAILED DESCRIPTION
[0020] In order to have a clearer understanding of the technical features, purposes and effects of the present application, the specific implementation methods of the present application are now described in detail with reference to the accompanying drawings.
[0021] An embodiment of the present application provides a method for planning a path for oil spill cleanup using multiple unmanned boats in combination with SAR images.
[0022] Please refer to Figure 1 , Figure 1 This is a step diagram of a method for planning a multi-unmanned boat oil spill cleaning path in combination with SAR images in an embodiment of the present application, including: S1: Obtain SAR images and combine them with the YOLOv8 model to determine the marine oil spill area; S2: Input the coordinates of the marine oil spill area into the GNOME model, make diffusion predictions for the marine oil spill area, and obtain the prediction results of the oil spill diffusion distribution; Specifically, in order to achieve the integration between satellite imagery and oil spill modeling, it is necessary to realistically describe the scope of the oil spill area in the simulation. The GNOME (General NOAA Operational Modeling Environment) model can be used to predict the effects of processes such as wind and ocean currents on the spread of oil spills on the water surface. It can also be used to understand how predicted oil trajectories are affected by the uncertainty of ocean current and wind observations and forecasts, and to understand how chemical and physical changes occur in predicted oil spills while they remain on the water surface. After obtaining the geographic location information of the oil spill boundary, the coordinate file and related environmental factors (such as wind speed, wind direction, ocean currents, and temperature) are processed using the GNOME (General NOAA Operational Modeling Environment) model to obtain the distribution of the oil spill. The results are as follows: Figure 4 shown.
[0023] S3: Based on the prediction results of oil spill diffusion distribution, a grid map is obtained using the Gaussian mixture model; S4: The geodesic-Manhattan Voronoi partitioning method is used in combination with the grid map to divide the working area of the unmanned boat; S5: Use the STC algorithm to plan the paths of the unmanned boats in the working area to obtain the cleaning paths of each unmanned boat.
[0024] Specifically, the STC (spanning tree coverage) algorithm is used to generate a spanning tree for each Voronoi region using a depth-first search algorithm in the Voronoi diagram. Finally, the unmanned boats circle around each spanning tree clockwise, which is the cleaning path of each unmanned boat. The spanning tree visualization result is as follows: Figure 7 The detour path is shown in 8.
[0025] Step S1 includes: Based on the acquired SAR images, a data set is constructed; the YOLOv8 model is trained using the data set to obtain a recognition and segmentation model; By identifying the segmentation model, the oil spill area is identified in the SAR image with geographic information to determine the marine oil spill area.
[0026] Specifically, the Yolov8 model is first used to train the training set, which contains 6218 oil spill SAR images. The data set is randomly split into a training set and a test set, 75% for model training and 25% for testing. The training results are shown in Figure 2. Figure 2 As shown in the figure, after the recognition segmentation model is trained, the recognition segmentation model is used to identify the oil spill area on the SAR image with geographic information. The results are shown in Figure 3 As shown, after obtaining the boundary, the boundary geographical location information is extracted.
[0027] Step S3 includes: According to the prediction results of oil spill diffusion distribution, the Gaussian mixture model is used to calculate the probability density of each point in the marine oil spill area to obtain a probability density distribution map; Discretize the probability density distribution map into a grid map.
[0028] Specifically, after obtaining the spatiotemporal distribution of oil spills (the prediction results of oil spill diffusion distribution), the Gaussian mixture model is used to fit the distribution of oil spills, and then the probability density map of oil spills can be obtained. After discretization, a gridded probability density map can be obtained, such as Figure 5 shown.
[0029] The Gaussian mixture model is composed of a convex combination of N Gaussian densities, and the mathematical expression is as follows:
[0030]
[0031] in represents the density function of the Gaussian mixture model, It is The weights of the Gaussian components, It is The mean vector of the Gaussian components, It is The covariance matrix of the Gaussian components; Indicates is the mean, is the probability density function of the multivariate Gaussian distribution of the covariance; yes The determinant of ; Represents the dimensionality of the data; Indicates the location of the data point.
[0032] Step S4 includes: Using geodesic-Manhattan Voronoi partitioning, The working area is divided into two parts: By using the ocean oil spill area in a two-dimensional plane According to the number of unmanned boats, generate Generators, Represents the position of the generator, and performs geodesic-Manhattan Voronoi partitioning based on the generator position, which is mathematically defined as:
[0033] in Represents the divided Voronoi region; Indicates the location of points within the Voronoi region; Represents the oil spill area in a two-dimensional plane; Represents a point within the Voronoi region and the corresponding region generator The geodesic-Manhattan distance of .
[0034] Specifically, The generator always moves towards the Voronoi partition formed by itself at the current moment. The centroid position Direction movement; the final generator will stabilize at the center of mass , we get the geodesic-Manhattan Voronoi partitioning of each area.
[0035] The Voronoi region Center of gravity , Centroid The definition is as follows:
[0036]
[0037] in, Indicates the probability density value of each point in the probability density distribution map.
[0038] Specifically, the geodesic-Manhattan Voronoi partitioning is a spatial partitioning method that combines the concepts of geodesic distance and Manhattan distance and is used to partition regions in multidimensional space. The working area of each unmanned boat is obtained by partitioning, and the results are as follows: Figure 6 shown.
[0039] The present application also discloses an electronic device. Fig. 9 , Fig. 9The electronic device 500 may include: at least one processor 501 , at least one network interface 504 , a user interface 503 , a memory 505 , and at least one communication bus 502 .
[0040] The communication bus 502 is used to realize the connection and communication between these components.
[0041] The user interface 503 may include a display screen, and the optional user interface 503 may also include a standard wired interface or a wireless interface.
[0042] The network interface 504 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).
[0043] The present application also discloses a computer-readable storage medium storing a plurality of instructions suitable for loading by a processor to execute the above-mentioned method for planning a path for oil spill cleaning by multiple unmanned boats in combination with SAR images.
[0044] The above are only exemplary embodiments of the present disclosure and cannot be used to limit the scope of the present disclosure. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the disclosure of the specification and the truth of practice, those skilled in the art will easily think of other embodiments of the present disclosure.
[0045] This application is intended to cover any variation, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary technical means in the art not described in the present disclosure. The description and examples are to be regarded as exemplary only, and the scope and spirit of the present disclosure are defined by the claims.
Claims
1. A method for planning a multi-unmanned boat oil spill cleaning path based on SAR images, characterized in that: The method comprises the following steps: S1: Obtain SAR images and combine them with the YOLOv8 model to determine the marine oil spill area; S2: Input the coordinates of the marine oil spill area into the GNOME model, make diffusion predictions for the marine oil spill area, and obtain the prediction results of the oil spill diffusion distribution; S3: Based on the prediction results of oil spill diffusion distribution, a grid map is obtained using the Gaussian mixture model; S4: The geodesic-Manhattan Voronoi partitioning method is used in combination with the grid map to divide the working area of the unmanned boat; S5: Use the STC algorithm to plan the paths of the unmanned boats in the working area to obtain the cleaning paths of each unmanned boat.
2. The method for planning a multi-unmanned boat oil spill cleaning path in combination with SAR images as claimed in claim 1, characterized in that: Step S1 includes: Based on the acquired SAR images, a data set is constructed; the YOLOv8 model is trained using the data set to obtain a recognition and segmentation model; By identifying the segmentation model, the oil spill area is identified in the SAR image with geographic information to determine the marine oil spill area.
3. The method for planning a multi-unmanned boat oil spill cleaning path in combination with SAR images as claimed in claim 1, characterized in that: Step S3 includes: According to the prediction results of oil spill diffusion distribution, the Gaussian mixture model is used to calculate the probability density of each point in the marine oil spill area to obtain a probability density distribution map; Discretize the probability density distribution map into a grid map.
4. The method for planning a multi-unmanned boat oil spill cleaning path in combination with SAR images as claimed in claim 1, characterized in that: The Gaussian mixture model is composed of a convex combination of N Gaussian densities, and the mathematical expression is as follows: in represents the density function of the Gaussian mixture model, It is The weights of the Gaussian components, It is The mean vector of the Gaussian components, It is The covariance matrix of the Gaussian components; Indicates is the mean, is the probability density function of the multivariate Gaussian distribution of the covariance; yes The determinant of ; Represents the dimensionality of the data; Indicates the location of the data point.
5. The method for planning a multi-unmanned boat oil spill cleaning path in combination with SAR images as claimed in claim 3, characterized in that: Step S4 includes: Using geodesic-Manhattan Voronoi partitioning, The working area is divided into two parts: By using the ocean oil spill area in a two-dimensional plane According to the number of unmanned boats, generate Generators, Represents the position of the generator, and performs geodesic-Manhattan Voronoi partitioning based on the generator position, which is mathematically defined as: in Represents the divided Voronoi region; Indicates the location of points within the Voronoi region; Represents the oil spill area in a two-dimensional plane; Represents a point within the Voronoi region and the corresponding region generator The geodesic-Manhattan distance of .
6. The method for planning a multi-unmanned boat oil spill cleaning path in combination with SAR images as claimed in claim 5, characterized in that: The Voronoi region Center of gravity , Centroid The definition is as follows: in, Indicates the probability density value of each point in the probability density distribution map.
7. An electronic device, characterized in that: The electronic device comprises a processor (501), a memory (505), a user interface (503) and a network interface (504), wherein the memory (505) is used to store instructions, the user interface (503) and the network interface (504) are used to communicate with other devices, and the processor (501) is used to execute the instructions stored in the memory (505) so that the electronic device executes the method according to any one of claims 1 to 6.
8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed by a computer, the method according to any one of claims 1 to 6 is executed.
