A method and system for predicting ocean current energy in a target sea area
By acquiring and combining underwater dynamic and sonar sample images, building a current motion trajectory model and combining deep learning, the problem of insufficient accuracy of current energy prediction is solved, and more accurate and stable current energy prediction is achieved.
Patent Information
- Application Number
- CN202411154263.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-22
- Publication Date
- 2024-12-20
- Estimated Expiration
- 2044-08-22
AI Technical Summary
In the prior art, the accuracy of the current energy prediction is insufficient, resulting in difficulty in managing the marine energy system.
By obtaining the underwater dynamic sample images and sonar sample images of the target sea area, extracting the motion trajectory data, combining and constructing the current motion trajectory model, combining sea surface simulation data and deep learning technology, establishing a current energy prediction model, and performing multiple iterative updates Improve model stability and accuracy.
It improves the accuracy and stability of current energy prediction, enables more realistic simulation of ocean motion information, and determines the available depth of current energy based on the water depth.
Smart Images
Figure CN118674125B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of marine engineering technology, and in particular to a method and system for predicting ocean current energy in a target sea area. Background Art
[0002] With the continuous development of energy development, renewable energy has gradually come into people's view, and ocean current energy, as a kind of clean marine energy, has also been gradually applied to power generation and other fields.
[0003] In the prediction of ocean current energy, ocean current energy is often calculated and predicted based on the seawater movement characteristics of the target sea area. However, due to the complexity of the marine environment, there are large deviations in the simulation of ocean current movement, which makes it difficult to predict the existing ocean current energy generation efficiency and it is impossible to obtain accurate power generation prediction data, which is not conducive to the effective integrated management of the energy system.
[0004] It can be seen that how to make a more accurate prediction of ocean current energy has become a technical problem that needs to be solved urgently by those skilled in the art. Summary of the invention
[0005] The present invention provides a method and system for predicting ocean current energy in a target sea area, so as to solve the technical problem of insufficient accuracy in predicting ocean current energy in the prior art.
[0006] In order to solve the above technical problems, an embodiment of the present invention provides a method for predicting ocean current energy in a target sea area, the method comprising:
[0007] According to the acquired underwater dynamic sample images of the target sea area, the first motion trajectory data of the ocean current is obtained.
[0008] Feature extraction is performed on underwater sonar sample images of the target sea area at different time periods to obtain various time-sharing static motion features, and the second motion trajectory data of the ocean current is restored based on the various time-sharing static motion features.
[0009] The first motion trajectory data and the second motion trajectory data are combined to obtain the motion trajectory data to be optimized.
[0010] A current motion trajectory model of the target sea area is constructed based on the motion trajectory data to be optimized and the sea surface simulation data of the target sea area, wherein the current motion trajectory model is designed to correct the model itself according to the response results caused by the applied sea surface wind simulation disturbance.
[0011] Based on big data analysis technology, the historical ocean current energy sample data of the target sea area is analyzed, and the ocean current motion trajectory model is deep-learned and trained with the corresponding analysis results to obtain an ocean current energy prediction model.
[0012] In the real-time prediction process, the ocean current energy of the target sea area is predicted using the ocean current energy prediction model.
[0013] Furthermore, the historical ocean current energy sample data of the target sea area is analyzed based on the big data analysis technology, and the ocean current motion trajectory model is trained by deep learning with the corresponding analysis results to obtain an ocean current energy prediction model, including:
[0014] A first sample set consisting of the historical ocean current energy sample data is obtained.
[0015] The first sample set is split based on energy standards corresponding to typical daily ocean current energy data to obtain a sample set to be tested and a sample set to be optimized.
[0016] Typical Gaussian noises of different intensities are added to the sample set to be optimized to obtain a sample set to be trained.
[0017] The ocean current motion trajectory model is subjected to deep learning training using the sample set to be trained.
[0018] The trained ocean current motion trajectory model is tested based on the sample set to be tested to obtain the corresponding ocean current energy prediction model.
[0019] Furthermore, the deep learning training of the ocean current motion trajectory model using the sample set to be trained includes:
[0020] A corresponding relationship between the ocean current motion characteristic data in the ocean current set and the ocean current energy is established according to the ocean current set.
[0021] The corresponding relationship is transferred to the ocean current motion trajectory model to perform deep learning training on the ocean current motion trajectory model.
[0022] Furthermore, the sea surface simulation data at least includes wave data and temperature and humidity data.
[0023] The construction of the ocean current motion trajectory model of the target sea area includes:
[0024] Different sea surface wind simulation conditions are created and applied according to the wave data and the temperature and humidity data.
[0025] Based on the linear relationship between wave data and ocean current velocity, the ocean current response trajectory data under the current sea surface wind simulation conditions are obtained.
[0026] The ocean current motion trajectory of the target sea area is corrected based on the deviation result between the ocean current response trajectory data and the motion trajectory data to be optimized, and the ocean current motion trajectory model is constructed based on the corrected result.
[0027] Furthermore, the first motion trajectory data and the second motion trajectory data both include ocean current motion trajectories at different depths.
[0028] The step of merging the first motion trajectory data and the second motion trajectory data to obtain the motion trajectory data to be optimized includes:
[0029] The first trajectory feature points corresponding to the first motion trajectory data at different depths and the second trajectory feature points corresponding to the second motion trajectory data are extracted respectively.
[0030] The first trajectory feature points and the second trajectory feature points are weightedly combined, and corresponding trajectories at different depths are reconstructed according to the combined data to obtain the motion trajectory data to be optimized.
[0031] Furthermore, obtaining the first motion trajectory data of the ocean current according to the acquired underwater dynamic sample image of the target sea area includes:
[0032] A first preprocessing is performed on the underwater dynamic sample image.
[0033] Based on the inter-frame difference algorithm, the motion area is detected from the pre-processed underwater dynamic sample image.
[0034] Each ocean current profile feature point is extracted in the motion area.
[0035] The matching degree of each of the ocean current profile feature points is calculated, and the motion trajectory is initialized with the ocean current profile feature points whose matching degrees meet preset conditions, so as to obtain the first motion trajectory data.
[0036] Furthermore, the first preprocessing of the underwater dynamic sample image includes:
[0037] Grayscale correction and refraction compensation processing are performed on the underwater dynamic sample image.
[0038] Furthermore, the feature extraction of underwater sonar sample images of the target sea area at different time periods to obtain each time-sharing static motion feature includes:
[0039] A second preprocessing is performed on each of the underwater sonar sample images.
[0040] Based on the edge detection method, image segmentation is performed on each of the underwater sonar sample images.
[0041] Based on the segmentation result, the time-sharing static motion feature of each of the underwater sonar sample images is extracted, and the time-sharing static motion feature includes one or more of the following:
[0042] Grayscale features, geometric features, texture features and transform domain features.
[0043] Furthermore, the performing a second preprocessing on each of the underwater sonar sample images comprises:
[0044] Each of the underwater sonar sample images is subjected to median filtering and angle sharpening processing.
[0045] Another embodiment of the present invention provides a system for predicting ocean current energy in a target sea area, comprising:
[0046] The dynamic image acquisition module is used to obtain the first motion trajectory data of the ocean current based on the underwater dynamic sample image of the target sea area.
[0047] The sonar image acquisition module is used to extract features from underwater sonar sample images of the target sea area at different time periods to obtain various time-sharing static motion features, and restore the second motion trajectory data of the ocean current based on the various time-sharing static motion features.
[0048] The image merging module is used to merge the first motion trajectory data and the second motion trajectory data to obtain the motion trajectory data to be optimized.
[0049] The trajectory model construction module is used to construct a current motion trajectory model of the target sea area based on the motion trajectory data to be optimized and the sea surface simulation data of the target sea area, wherein the current motion trajectory model is designed to realize the correction of the model itself according to the response result caused by the applied sea surface wind simulation disturbance.
[0050] The prediction model building module is used to analyze the historical ocean current energy sample data of the target sea area based on big data analysis technology, and to perform deep learning training on the ocean current motion trajectory model with the corresponding analysis results to obtain an ocean current energy prediction model.
[0051] The ocean current energy prediction module is used to predict the ocean current energy of the target sea area using the ocean current energy prediction model during the real-time prediction process.
[0052] Compared with the prior art, the embodiments of the present invention have the following advantages:
[0053] (1) By performing multiple rounds of iterative updates on the seawater motion model based on the motion trajectory deviation under the same sea breeze disturbance, the stability of the seawater motion model in response to disturbances can be improved, thereby simulating more realistic ocean motion information;
[0054] (2) According to different water depths, the motion state of the seawater motion model under wind disturbance is obtained in layers and the model is constructed. This can not only simulate the seawater motion information more accurately, but also reasonably determine the underwater usable depth based on the changes in ocean current energy at different underwater depths. BRIEF DESCRIPTION OF THE DRAWINGS
[0055] Figure 1 A flow chart of the steps of a method for predicting ocean current energy in a target sea area provided by an embodiment of the present invention;
[0056] Figure 2 This is a structural block diagram of a target sea area ocean current energy prediction system provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0057] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0058] In the description of this application, the terms "first", "second", "third", etc. are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first", "second", "third", etc. may explicitly or implicitly include one or more of the feature. In the description of this application, unless otherwise specified, "plurality" means two or more.
[0059] In the description of the present application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be a connection between the two elements. The terms "vertical", "horizontal", "left", "right", "upper", "lower" and similar expressions used herein are only for illustrative purposes, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific circumstances.
[0060] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in the present invention have the same meaning as those commonly understood by those skilled in the art. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood by specific circumstances.
[0061] An embodiment of the present invention provides a method for predicting ocean current energy in a target sea area. For details, see Figure 1 , Figure 1 The flowchart of the method for predicting ocean current energy in a target sea area in one embodiment of the present invention includes steps S11 to S16:
[0062] Step S11, obtaining first motion trajectory data of the ocean current according to the acquired underwater dynamic sample image of the target sea area.
[0063] Ocean current energy is the kinetic energy generated by the flow of seawater in the ocean, which is closely related to the velocity and flow rate of seawater. Usually, the calculation formula of ocean current energy is as follows:
[0064]
[0065] in, is the density of seawater, is the area through which seawater passes, is the seawater flow rate.
[0066] However, due to the influence of various natural conditions, the flow rate of seawater changes frequently, and the density and flow rate of seawater at different depths also vary greatly. Therefore, the simulation of the movement state of seawater is particularly important in the prediction of ocean current energy. This embodiment collects underwater images in seawater as a data source to construct an ocean current motion trajectory model to achieve the simulation of the movement state of seawater.
[0067] Firstly, an underwater robot equipped with a camera system is used to collect underwater dynamic sample images at different depths in the target sea area, and then a first preprocessing operation is performed on the underwater dynamic sample images, wherein the first preprocessing operation includes grayscale correction and refraction compensation.
[0068] Specifically, the first preprocessing process is:
[0069] (1) Perform grayscale correction on the collected underwater dynamic sample images to eliminate the color distortion of the underwater images.
[0070] (2) Due to the influence of underwater light refraction, the position and shape of the ripples in the image may be distorted. This phenomenon is more obvious in seawater below a certain depth. Therefore, this embodiment performs refraction compensation processing on the underwater dynamic sample images of different depths after grayscale correction to restore the true state of the water ripples in the image.
[0071] The shore-based terminal is an important component of the shore-to-ship wireless communication system. It is mainly installed on land and is used for communication and data exchange with equipment on ships. The transmitted text information contains a large amount of descriptive information related to the target sea area. This embodiment uses the text description information obtained by the shore-based terminal near the target sea area to reconstruct the underwater image of the target sea area to obtain a reconstructed image, and integrates the reconstructed image and the underwater dynamic sample image to obtain integrated image data.
[0072] Based on the inter-frame difference algorithm, the motion area is detected from the integrated image data to quickly analyze and determine the motion characteristics in the dynamic image.
[0073] The characteristic points of each ocean current contour are extracted in the motion area, the matching degree of each ocean current contour characteristic point is calculated, and it is determined whether the matching degree meets the preset conditions, and the ocean current contour characteristic points that meet the conditions are screened out to initialize the motion trajectory, and the first motion trajectory data is obtained.
[0074] Step S12, extracting features from underwater sonar sample images of the target sea area at different time periods to obtain various time-sharing static motion features, and restoring second motion trajectory data of the ocean current based on the various time-sharing static motion features.
[0075] Due to the complex composition of seawater, the acquisition of underwater images by video acquisition equipment alone will result in unclear imaging due to the turbidity of the seawater. In addition, the imaging of the camera depends on light, while the underwater light is often insufficient and affected by the absorption and scattering of the water system, resulting in limited imaging effect. Therefore, this embodiment uses sonar equipment to collect underwater sonar images in the target sea area according to fixed time intervals. Specifically, this embodiment adopts a collection frequency of once every 300ms to collect underwater images at different depths, obtain a fixed-frequency time-sharing static underwater sonar sample image, and perform a second preprocessing on the underwater sonar sample image.
[0076] The second preprocessing specifically includes: performing median filtering on the image to remove salt and pepper noise in the image, and performing angle sharpening on the image to enhance the contrast of the contour in the image.
[0077] Based on the edge detection method, the preprocessed image is segmented to segment the ripple image generated by the water flow movement to obtain a segmented image.
[0078] Based on the segmented image, the time-sharing static motion features such as grayscale features, geometric features, texture features and transform domain features in each underwater sonar sample image are extracted, and based on the above time-sharing static motion features, the second motion trajectory data of the ocean current is obtained.
[0079] Since sonar equipment can produce high-resolution underwater images, these images can clearly capture the characteristics of water flow and have high real-time performance, which helps to extract the characteristics of seawater bodies.
[0080] Step S13: merging the first motion trajectory data and the second motion trajectory data to obtain motion trajectory data to be optimized.
[0081] Since sound waves have strong penetrating power when propagating underwater, they can penetrate complex seawater bodies and reach deeper and farther areas that are difficult for cameras to reach, and are not limited by light. Therefore, using sonar equipment to collect underwater sonar images can better obtain seawater characteristics. Sonar equipment can also detect information such as the geological structure and mineral resources of the seabed, providing an important basis for the development and utilization of marine resources.
[0082] The camera can capture rich details such as color, texture, and shape of underwater objects, and is more intuitive in image formation. Therefore, this embodiment merges the first motion trajectory data generated according to the image captured by the camera and the second motion trajectory data generated according to the sonar image captured by the sonar device to obtain the motion trajectory data to be optimized.
[0083] Specifically, the process of merging the first motion trajectory data and the second motion trajectory data is as follows:
[0084] The first motion trajectory data and the second motion trajectory data both include ocean current motion trajectories at different depths, and first trajectory feature points corresponding to the first motion trajectory data at different depths and second trajectory feature points corresponding to the second motion trajectory data are extracted respectively.
[0085] According to the actual measurement situation, the first trajectory feature point and the second trajectory feature point are weightedly merged, and the corresponding trajectories at different depths are reconstructed according to the merged data to obtain the motion trajectory data to be optimized. By merging images using the trajectory feature point correspondence method, it is possible to ensure that the merged image maintains high clarity and detail information at the edge.
[0086] Step S14, constructing a current motion trajectory model of the target sea area according to the motion trajectory data to be optimized and the sea surface simulation data of the target sea area, wherein the current motion trajectory model is designed to realize the correction of the model itself according to the response result caused by the applied sea surface wind simulation disturbance.
[0087] Underwater sensors are used to collect seawater data such as temperature, salinity, density, etc. at different depths of seawater, and a current motion trajectory model reflecting the motion state of seawater in the target sea area is constructed based on the collected seawater data and the motion trajectory data to be optimized obtained in step S13.
[0088] However, in the actual natural environment, the flow of seawater is affected by the wind on the sea surface. Since it is difficult to accurately collect the sea breeze in real time, this embodiment applies multiple wind disturbances to the model after the model is constructed, and updates the model according to the model's response to each wind disturbance to achieve model training. The specific process is as follows:
[0089] Different sea surface wind simulation conditions are created based on the sea surface wave data and the sea surface temperature and humidity data, and applied to the motion trajectory data to be optimized.
[0090] Based on the linear relationship between wave data and ocean current velocity, the ocean current response trajectory data under the current sea surface wind simulation conditions are obtained, where the linear relationship between wave data and ocean current velocity is:
[0091]
[0092] Where k is the wave number, is the wave cycle frequency, is the wave height, t is the time, y is the distance from the point to the sea level, and h is the depth of the sea water.
[0093] The deviation between the ocean current response trajectory data and the trajectory data to be optimized is used to correct the ocean current trajectory of the target sea area. Since the effect of wind on the seawater under the water surface is weakened with the increase of depth, the deviation result is the layered trajectory deviation at different water depths. Based on this layered trajectory deviation, the ocean current trajectory of the target sea area is corrected, and the corrected result is used to realize the construction of the ocean current trajectory model.
[0094] Through multiple rounds of iterative updates, the ocean current motion trajectory model can gradually approach the real state not affected by sea breeze factors. At this time, the ocean current motion trajectory model can also be used to apply external force conditions to simulate the movement of seawater under different natural weather conditions to complete other different experiments, which will not be described in detail in this embodiment.
[0095] In addition, since the seawater movement model simulated based on natural data is unstable compared to the real seawater movement, applying wind disturbances and training the model based on its response to the disturbances also increases the model's anti-interference and stability.
[0096] Step S15, analyzing the historical ocean current energy sample data of the target sea area based on big data analysis technology, performing deep learning training on the ocean current motion trajectory model with the corresponding analysis results, and obtaining an ocean current energy prediction model.
[0097] The first sample set consisting of historical ocean current energy sample data is obtained; however, although a certain amount of historical ocean current energy sample data is obtained, it still does not meet the requirement of deep learning algorithms for a large number of samples for training. Therefore, we need to expand the data set to a certain extent.
[0098] This embodiment collects 3122 original historical ocean current energy sample data, each sample data consists of 4 images, and the image resolution is 226×226, which is equivalent to a total of 12488 images. The first sample set is split according to the energy standard corresponding to the typical daily ocean current energy data, of which 1000 sample data are used as the sample set to be tested, and the remaining 2122 sample data are used as the sample set to be optimized for data expansion and participate in the training of the model.
[0099] In environmental science, a typical day of ocean current energy refers to a typical date that can represent ocean current energy in a period of time. On a typical day, key parameters such as ocean current speed, direction, and energy density can better reflect the overall level and characteristics of ocean current energy in that period. Therefore, this embodiment uses ocean current energy data of a typical day as the basis for splitting, which improves the reliability of sample splitting and can also improve the predictive ability of subsequent models.
[0100] Typical Gaussian noises of different intensities are added to the sample set to be optimized to obtain the sample set to be trained; and the corresponding relationship between the ocean current motion characteristic data and the ocean current energy in the sample set to be trained is established according to the sample set to be trained.
[0101] The above correspondence is transferred to the ocean current motion trajectory model to perform deep learning training on the ocean current motion trajectory model to obtain an ocean current energy prediction model, which can help the model quickly learn to predict ocean current energy and improve the prediction performance of the model.
[0102] The trained ocean current motion trajectory model is tested based on the sample set to be tested, and the prediction effect of ocean current energy is tested to obtain the corresponding ocean current energy prediction model.
[0103] Step S16, in the real-time prediction process, the ocean current energy of the target sea area is predicted using the ocean current energy prediction model.
[0104] In actual energy utilization of ocean current energy, it is often limited by the depth of sea water. Therefore, according to the relationship between the difficulty of developing and utilizing ocean current energy and the size of ocean current energy, an energy threshold line is established. Based on the predicted ocean current energy data, the ocean current energy change curve at different water depths is drawn. The depth range corresponding to the ocean current energy above the energy threshold line in the ocean current energy change curve is divided into the available range, thereby evaluating the availability of ocean current energy.
[0105] The method for predicting ocean current energy in the target sea area of the present invention performs multiple rounds of iterative updates on the seawater motion model according to the motion trajectory deviation under the same sea breeze disturbance, which can improve the stability of the seawater motion model in response to disturbances, thereby simulating more realistic ocean motion information; according to different water depths, the motion state of the seawater motion model under wind disturbance is obtained in layers and a model is constructed, which can not only simulate the seawater motion information more accurately, but also reasonably determine the underwater usable depth according to the changes in ocean current energy at different underwater depths.
[0106] The embodiment of the present invention further provides a target sea area ocean current energy prediction system, which is used to execute the target sea area ocean current energy prediction method as described above. Figure 2 This is a structural block diagram of a target sea area current energy prediction system according to an embodiment of the present invention, and the system includes:
[0107] The dynamic image acquisition module 21 is used to obtain the first motion trajectory data of the ocean current according to the underwater dynamic sample image of the target sea area acquired.
[0108] The sonar image acquisition module 22 is used to extract features from the underwater sonar sample images of the target sea area at different time periods to obtain various time-sharing static motion features, and restore the second motion trajectory data of the ocean current based on the various time-sharing static motion features.
[0109] The image merging module 23 is used to merge the first motion trajectory data and the second motion trajectory data to obtain the motion trajectory data to be optimized.
[0110] The trajectory model construction module 24 is used to construct a current motion trajectory model of the target sea area based on the motion trajectory data to be optimized and the sea surface simulation data of the target sea area, wherein the current motion trajectory model is designed to realize the correction of the model itself according to the response result caused by the applied sea surface wind simulation disturbance.
[0111] The prediction model building module 25 is used to analyze the historical ocean current energy sample data of the target sea area based on big data analysis technology, and to perform deep learning training on the ocean current motion trajectory model with the corresponding analysis results to obtain an ocean current energy prediction model.
[0112] The ocean current energy prediction module 26 is used to predict the ocean current energy of the target sea area using the ocean current energy prediction model during the real-time prediction process.
[0113] The technical features and technical effects of the system proposed in the embodiment of the present invention are the same as those of the method proposed in the embodiment of the present invention, and are not described in detail here. Each module in the above system can be implemented in whole or in part by software, hardware, and a combination thereof. Each of the above modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each of the above modules.
[0114] In summary, the method and system for predicting ocean current energy in a target sea area provided by the embodiments of the present invention have the following beneficial effects compared with the prior art:
[0115] (1) By performing multiple rounds of iterative updates on the seawater motion model based on the motion trajectory deviation under the same sea breeze disturbance, the stability of the seawater motion model in response to disturbances can be improved, thereby simulating more realistic ocean motion information.
[0116] (2) According to different water depths, the motion state of the seawater motion model under wind disturbance is obtained in layers and the model is constructed. This can not only simulate the seawater motion information more accurately, but also reasonably determine the underwater usable depth based on the changes in ocean current energy at different underwater depths.
[0117] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.
Claims
1. A method for predicting ocean current energy in a target sea area. It is characterized in that include: Obtaining first motion trajectory data of the ocean current according to the acquired underwater dynamic sample image of the target sea area; Extracting features from underwater sonar sample images of the target sea area at different time periods to obtain time-sharing static motion features, and restoring second motion trajectory data of the ocean current based on the time-sharing static motion features; Merging the first motion trajectory data and the second motion trajectory data to obtain motion trajectory data to be optimized; Constructing a current motion trajectory model of the target sea area according to the motion trajectory data to be optimized and the sea surface simulation data of the target sea area, wherein the current motion trajectory model is designed to realize the correction of the model itself according to the response result caused by the applied sea surface wind simulation disturbance; Analyze the historical ocean current energy sample data of the target sea area based on big data analysis technology, and use the corresponding analysis results to perform deep learning training on the ocean current motion trajectory model to obtain an ocean current energy prediction model; In the real-time prediction process, the ocean current energy of the target sea area is predicted using the ocean current energy prediction model; Among them, the historical ocean current energy sample data of the target sea area is analyzed based on the big data analysis technology, and the ocean current motion trajectory model is deep-learned and trained with the corresponding analysis results to obtain an ocean current energy prediction model, including: obtaining a first sample set composed of the historical ocean current energy sample data; splitting the first sample set based on the energy standard corresponding to the typical daily ocean current energy data to obtain a sample set to be tested and a sample set to be optimized; adding typical Gaussian noise of different intensities to the sample set to be optimized to obtain a sample set to be trained; deep-learning training is performed on the ocean current motion trajectory model with the sample set to be trained; and the trained ocean current motion trajectory model is tested based on the sample set to be tested to obtain the corresponding ocean current energy prediction model.
2. The method for predicting ocean current energy in a target sea area according to claim 1, It is characterized in that The deep learning training of the ocean current motion trajectory model using the sample set to be trained includes: Establishing a corresponding relationship between the ocean current motion characteristic data in the to-be-trained sample set and the ocean current energy according to the to-be-trained sample set; The corresponding relationship is transferred to the ocean current motion trajectory model to perform deep learning training on the ocean current motion trajectory model.
3. The method for predicting ocean current energy in a target sea area according to claim 1, It is characterized in that The sea surface simulation data at least includes wave data and temperature and humidity data; The construction of the ocean current motion trajectory model of the target sea area includes: Creating and applying different sea surface wind simulation conditions according to the wave data and the temperature and humidity data; Based on the linear relationship between wave data and ocean current velocity, the ocean current response trajectory data under the current sea surface wind simulation conditions is obtained; The ocean current motion trajectory of the target sea area is corrected based on the deviation result between the ocean current response trajectory data and the motion trajectory data to be optimized, and the ocean current motion trajectory model is constructed based on the corrected result.
4. The method for predicting ocean current energy in a target sea area according to claim 1, It is characterized in that The first motion trajectory data and the second motion trajectory data both include ocean current motion trajectories at different depths; The step of merging the first motion trajectory data and the second motion trajectory data to obtain the motion trajectory data to be optimized includes: Respectively extracting first trajectory feature points corresponding to the first motion trajectory data at different depths, and second trajectory feature points corresponding to the second motion trajectory data; The first trajectory feature points and the second trajectory feature points are weightedly combined, and corresponding trajectories at different depths are reconstructed according to the combined data to obtain the motion trajectory data to be optimized.
5. The method for predicting ocean current energy in a target sea area according to claim 1, It is characterized in that The method of obtaining the first motion trajectory data of the ocean current according to the underwater dynamic sample image of the target sea area includes: Performing a first preprocessing on the underwater dynamic sample image; Based on an inter-frame difference algorithm, detecting a motion area from the pre-processed underwater dynamic sample image; Extracting each current contour feature point in the motion area; The matching degree of each of the ocean current profile feature points is calculated, and the motion trajectory is initialized with the ocean current profile feature points whose matching degrees meet preset conditions, so as to obtain the first motion trajectory data.
6. The method for predicting ocean current energy in a target sea area according to claim 5, It is characterized in that The first preprocessing of the underwater dynamic sample image includes: Grayscale correction and refraction compensation processing are performed on the underwater dynamic sample image.
7. The method for predicting ocean current energy in a target sea area according to claim 1, It is characterized in that The feature extraction of underwater sonar sample images of the target sea area at different time periods to obtain each time-sharing static motion feature includes: Performing a second preprocessing on each of the underwater sonar sample images; Based on the edge detection method, performing image segmentation on each of the underwater sonar sample images; Based on the segmentation result, the time-sharing static motion feature of each of the underwater sonar sample images is extracted, and the time-sharing static motion feature includes one or more of the following: Grayscale features, geometric features, texture features and transform domain features.
8. The method for predicting ocean current energy in a target sea area according to claim 7, It is characterized in that The second preprocessing of each of the underwater sonar sample images comprises: Each of the underwater sonar sample images is subjected to median filtering and angle sharpening processing.
9. A current energy prediction system for a target sea area, It is characterized in that include: A dynamic image acquisition module, used to obtain first motion trajectory data of the ocean current according to the underwater dynamic sample image of the target sea area; A sonar image acquisition module is used to extract features from underwater sonar sample images of the target sea area at different time periods to obtain various time-sharing static motion features, and restore the second motion trajectory data of the ocean current based on the various time-sharing static motion features; An image merging module, used for merging the first motion trajectory data and the second motion trajectory data to obtain the motion trajectory data to be optimized; A trajectory model building module, used to build a current motion trajectory model of the target sea area according to the motion trajectory data to be optimized and the sea surface simulation data of the target sea area, wherein the current motion trajectory model is designed to realize the correction of the model itself according to the response result caused by the applied sea surface wind simulation disturbance; A prediction model building module is used to analyze the historical ocean current energy sample data of the target sea area based on big data analysis technology, and to perform deep learning training on the ocean current motion trajectory model with the corresponding analysis results to obtain an ocean current energy prediction model; An ocean current energy prediction module, used to predict the ocean current energy of the target sea area using the ocean current energy prediction model during real-time prediction; Among them, the historical ocean current energy sample data of the target sea area is analyzed based on the big data analysis technology, and the ocean current motion trajectory model is deep-learned and trained with the corresponding analysis results to obtain an ocean current energy prediction model, including: obtaining a first sample set composed of the historical ocean current energy sample data; splitting the first sample set based on the energy standard corresponding to the typical daily ocean current energy data to obtain a sample set to be tested and a sample set to be optimized; adding typical Gaussian noise of different intensities to the sample set to be optimized to obtain a sample set to be trained; deep-learning training is performed on the ocean current motion trajectory model with the sample set to be trained; and the trained ocean current motion trajectory model is tested based on the sample set to be tested to obtain the corresponding ocean current energy prediction model.