A method and system for identifying hull coating damage based on machine learning

Abstract

The present invention provides a method and system for identifying hull coating damage based on machine learning, which belongs to the field of machine learning technology, and includes: obtaining ship physical field modeling based on electrochemical corrosion; setting the corrosion area and corrosion degree on the hull surface to obtain underwater potential data of the hull surface, and processing the underwater potential data using the ship physical field modeling to obtain a data set; preprocessing the data set to obtain a preprocessed data set; using the preprocessed data set to train a random forest model to obtain a target ship hull coating damage identification model; inputting the target ship underwater potential signal into the target ship hull coating damage identification model, and outputting the hull coating damage identification result. The present invention can realize automated hull coating damage identification through the training and optimization of the machine learning model, reducing the subjectivity and time-consuming nature of traditional manual inspections.

Description

Technical Field

[0001] The present invention relates to the field of machine learning technology, and in particular to a method and system for identifying hull coating damage based on machine learning. Background Art

[0002] In the history of ship development, most ships use metal materials. The shell based on metal materials not only takes into account the characteristics of being sturdy, but also makes the ship more beautiful. However, it is inevitable that steel materials will corrode due to various environmental conditions, especially ships, which are in seawater for a long time. The speed of electrochemical corrosion far exceeds the various chemical corrosion in general use scenarios. Even metal blocks with better performance cannot fundamentally prevent ships from being corroded by seawater. At present, the most effective and widely used protection method for ship corrosion is coating anti-corrosion. It prevents most steel materials from direct contact with seawater, thereby solving the corrosion problem from the root. However, in harsh marine environments, the hull anti-corrosion coating is easily damaged. Once the coating is damaged, the anti-corrosion function of the area will be completely lost, so the identification of hull coating damage is very critical for ship anti-corrosion technology.

[0003] There are currently two mainstream identification methods: (1) Coating corrosion defect feature identification and determination method based on electrochemical impedance spectroscopy and infrared thermal wave images; (2) Large-area visualization anti-corrosion coating failure detection technology and device based on infrared thermal imaging. Based on the above methods, the electrochemical key parameters of the coating corrosion stage and the corresponding visualization image defect features are extracted, and the anti-corrosion coating corrosion failure monitoring and evaluation method is established, and the anti-corrosion coating corrosion failure judgment criteria are established.

[0004] At present, the detection technology of infrared thermal wave map and infrared thermal image has problems such as high cost, bulky equipment, and long detection cycle. At present, the main treatment method for the above problems is to improve the infrared end equipment, which cannot fully take into account a series of problems such as equipment cost, time cost, and labor cost. Summary of the invention

[0005] The present invention provides a method and system for identifying hull coating damage based on machine learning, so as to solve the defects of the methods for identifying hull damage in the prior art, such as high cost, great implementation difficulty and large manpower cost.

[0006] In a first aspect, the present invention provides a method for identifying hull coating damage based on machine learning, comprising:

[0007] Get ship physics modeling based on electrochemical corrosion;

[0008] Setting the corrosion area and corrosion degree on the hull surface to obtain underwater potential data on the hull surface, and processing the underwater potential data using the ship physical field modeling to obtain a data set;

[0009] Preprocessing the data set to obtain a preprocessed data set;

[0010] The random forest model is trained using the preprocessed data set to obtain a target ship hull coating damage recognition model;

[0011] The underwater potential signal of the target ship is input into the target ship hull coating damage identification model, and the hull coating damage identification result is output.

[0012] According to a method for identifying hull coating damage based on machine learning provided by the present invention, a ship physical field modeling based on electrochemical corrosion is obtained, including:

[0013] Using modeling software to obtain geometric modeling parameters and basic physical field parameters of the ship, and forming the ship physical field modeling by the geometric modeling parameters and the basic physical field parameters;

[0014] The basic parameters of the physical field include electrical conductivity, magnetic permeability and material polarization characteristic curve.

[0015] According to a method for identifying hull coating damage based on machine learning provided by the present invention, a corrosion area and a corrosion degree are set on the hull surface to obtain underwater potential data of the hull surface, and the underwater potential data is processed by using the ship physical field modeling to obtain a data set, including:

[0016] The ship surface is divided into multiple areas according to the length and width of the ship and the preset positioning accuracy;

[0017] Acquire the location information of the corrosion areas of the same type of ships in the multiple areas and the underwater potential data corresponding to the average of multiple points at a specified distance directly below the keel of the ship to form the data set;

[0018] The data set is divided into a training set and a test set.

[0019] According to a method for identifying hull coating damage based on machine learning provided by the present invention, the corrosion area and corrosion degree are set on the hull surface to obtain underwater potential data of the hull surface, and the underwater potential data is processed by using the ship physical field modeling to obtain a data set, which also includes:

[0020] The ship physical field modeling is used to perform cyclic simulation on the location information of the corrosion area of ​​the same type of ship and on average multiple points at a specified distance directly below the keel of the ship to obtain simulated underwater potential data to supplement the data set.

[0021] According to a method for identifying hull coating damage based on machine learning provided by the present invention, the data set is preprocessed to obtain a preprocessed data set, including:

[0022] Splitting the data set into sequences according to the three-dimensional axial components, the length of the sequence being the number of data sets multiplied by 3;

[0023] The corrosion area information and the corrosion degree information are arranged before the sequence to form the preprocessing data set.

[0024] According to a method for identifying hull coating damage based on machine learning provided by the present invention, the preprocessed data set is used to train a random forest model to obtain a target ship hull coating damage identification model, including:

[0025] Inputting the feature data and label data in the preprocessed data set into the random forest model;

[0026] The random forest model constructs multiple decision trees based on the feature data, and determines the decision tree results by voting or averaging;

[0027] The hyperparameters of the random forest model are adjusted to perform model optimization, thereby obtaining the target ship hull coating damage identification model.

[0028] In a second aspect, the present invention also provides a hull coating damage identification system based on machine learning, comprising:

[0029] Acquisition module, used to acquire ship physical field modeling based on electrochemical corrosion;

[0030] A data processing module, used for setting the corrosion area and corrosion degree on the hull surface to obtain underwater potential data of the hull surface, and processing the underwater potential data by using the ship physical field modeling to obtain a data set;

[0031] A preprocessing module, used for preprocessing the data set to obtain a preprocessed data set;

[0032] A training module, used to train a random forest model using the preprocessed data set to obtain a target ship hull coating damage recognition model;

[0033] The identification module is used to input the underwater potential signal of the target ship into the hull coating damage identification model of the target ship and output the hull coating damage identification result.

[0034] In a third aspect, the present invention also provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a method for identifying hull coating damage based on machine learning as described above is implemented.

[0035] In a fourth aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-described methods for identifying hull coating damage based on machine learning.

[0036] In a fifth aspect, the present invention also provides a computer program product, comprising a computer program, which, when executed by a processor, implements any of the above-mentioned methods for identifying hull coating damage based on machine learning.

[0037] The hull coating damage identification method and system based on machine learning provided by the present invention can lock the hull damage area within a certain range by adding potential testing during the ship exit test phase and combining it with physical field modeling, thereby minimizing the workload of infrared testing at the current stage. As the number of uses increases, the equipment cost, time cost and labor cost required by the prior art will be reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0039] Figure 1 It is one of the flow charts of the method for identifying hull coating damage based on machine learning provided by the present invention;

[0040] Figure 2 This is the second flow chart of the method for identifying hull coating damage based on machine learning provided by the present invention;

[0041] Figure 3 It is a schematic diagram of modeling and segmentation of a target hull provided by the present invention;

[0042] Figure 4 It is a structural schematic diagram of a hull coating damage identification system based on machine learning provided by the present invention;

[0043] Figure 5 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0045] In view of the limitations of hull coating damage detection in the prior art, the present invention proposes to start from the corrosion electric field caused by hull coating damage, analyze the abnormal electromagnetic field generated by hull corrosion, and combine it with the current good deep learning technology to establish a suitable coating damage area identification model to minimize the scope of damage identification in order to save manpower and material resources.

[0046] Figure 1 FIG. 1 is one of the flow charts of the method for identifying hull coating damage based on machine learning provided in an embodiment of the present invention, such as Figure 1 As shown, including:

[0047] Step 100: Obtaining ship physical field modeling based on electrochemical corrosion;

[0048] Step 200: setting the corrosion area and corrosion degree on the hull surface to obtain underwater potential data on the hull surface, and processing the underwater potential data using the ship physical field modeling to obtain a data set;

[0049] Step 300: preprocessing the data set to obtain a preprocessed data set;

[0050] Step 400: training a random forest model using the preprocessed data set to obtain a target ship hull coating damage recognition model;

[0051] Step 500: Input the underwater potential signal of the target ship into the hull coating damage identification model of the target ship, and output the hull coating damage identification result.

[0052] like Figure 2 The overall process shown in the figure is a new hull coating damage identification method based on machine learning method provided by an embodiment of the present invention, which aims to improve the efficiency and accuracy of ship maintenance and safety management. The overall technical route includes: underwater potential signal acquisition; hull corrosion electric field modeling; data set acquisition; machine learning model establishment; evaluation of model loss degree to select a suitable model; and use of the model to obtain target features.

[0053] Specifically, it is first necessary to obtain physical field modeling of the ship based on electrochemical corrosion (using software such as COMSOL and HFSS that contain physical field modeling functions), collect simulation data sets based on the model (the model needs to fit the potential conditions when the ship is intact), establish training sets and test sets to complete the training of the deep learning model, and use the model to obtain the damaged area of ​​the hull based on the input detection electric field. The underwater potential signal of the ship required in the overall process includes multiple types of data: measured data, laboratory data, and simulation data. All of the above data can be implemented using this technical solution.

[0054] Through the training and optimization of machine learning models, the present invention can realize automatic identification of hull coating damage, reducing the subjectivity and time-consuming nature of traditional manual inspections; at the same time, the method can process a large amount of underwater potential data of the same type of hulls and improve the efficiency of maintenance work.

[0055] Based on the above embodiment, a ship physical field modeling based on electrochemical corrosion is obtained, including:

[0056] Using modeling software to obtain geometric modeling parameters and basic physical field parameters of the ship, and forming the ship physical field modeling by the geometric modeling parameters and the basic physical field parameters;

[0057] The basic parameters of the physical field include electrical conductivity, magnetic permeability and material polarization characteristic curve.

[0058] Specifically, the embodiment of the present invention uses modeling software (COMSOL, HFSS, etc.) to model the electrochemical corrosion physical field of the ship. Among them, COMSOL Multiphysics is a large-scale advanced numerical simulation software, which is widely used in scientific research and engineering calculations in various fields to simulate various physical processes in science and engineering. Based on the finite element method, the simulation of real physical phenomena is achieved by solving partial differential equations (single field) or partial differential equations (multiple fields), and mathematical methods are used to solve the physical phenomena of the real world. HFSS is a high-frequency structure simulation software that can quickly and accurately solve 3D electromagnetic problems, such as calculating S parameters, resonant frequencies and corresponding fields. HFSS uses the finite element solution method, that is, decomposing the object into many tiny tetrahedral finite elements for solution. All tetrahedral units in the model constitute a network, and the solutions of each finite element in the grid are related. Finally, the solution of the entire model is obtained through the association between finite elements.

[0059] The modeling process requires not only geometric modeling of the ship, but also the setting of basic parameters for solving the physical field (such as conductivity, magnetic permeability, polarization characteristic curve of the material, etc.). The model is adjusted based on the input initial potential signal so that the model can characterize the underwater potential signal of the target ship (actual ship data, laboratory data, and simulation data can all be used as target ship data). Figure 3 As shown, the main corrosion areas of the hull are divided into 28, which can be represented by serial numbers 1-28. Taking the No. 10 corrosion area as an example, the area is divided into 5×5 uniform units, and different corrosion degrees are given to a single area through these units.

[0060] On the basis of the above embodiment, the corrosion area and the corrosion degree are set on the hull surface to obtain the underwater potential data of the hull surface, and the underwater potential data is processed by using the ship physical field modeling to obtain a data set, including:

[0061] The ship surface is divided into multiple areas according to the length and width of the ship and the preset positioning accuracy;

[0062] Acquire the location information of the corrosion areas of the same type of ships in the multiple areas and the underwater potential data corresponding to the average of multiple points at a specified distance directly below the keel of the ship to form the data set;

[0063] The data set is divided into a training set and a test set.

[0064] The method further comprises setting the corrosion area and the corrosion degree on the hull surface to obtain the underwater potential data on the hull surface, and processing the underwater potential data by using the ship physical field modeling to obtain a data set.

[0065] The ship physical field modeling is used to perform cyclic simulation on the location information of the corrosion area of ​​the same type of ship and on average multiple points at a specified distance directly below the keel of the ship to obtain simulated underwater potential data to supplement the data set.

[0066] Specifically, in the embodiment of the present invention, it is firstly clarified that the abnormal electric field generated by electrochemical corrosion is caused by the change of current density on the corroded metal surface. In COMSOL, for all coating damage, 100% of the current density of the steel polarization curve is applied to simulate a completely uncoated area, and the calculation is performed by the following formula:

[0067]

[0068] in, represents the normal vector of the electrode surface, represents the surface current density vector, which is the component normal to the electrode surface, represents the total current density on the electrode surface, represents the local current density associated with different reactions on the electrode surface, Represents the number of different reactive parts on the electrode surface.

[0069] There is no way to modify the above formula in a weighted manner, but by dividing a single area more evenly and setting a certain percentage of random areas to be completely corroded within a single area, the effect of percentage corrosion of the entire area can be simulated, and this situation is closer to the actual situation.

[0070] by Figure 3 For example, the research surface is divided into multiple regions (usually related to the length-width ratio of the ship and the required positioning accuracy. The present invention takes 28 regions divided into 5×5 uniform units as an example. In actual situations, it is necessary to balance the relationship between network accuracy and model segmentation degree), and obtain the location information of the corrosion area of ​​the same type of ship and the underwater electric potential information (UEP) corresponding to the average of multiple points (taking 200 points as an example) 10 meters directly below the keel; if there is no large amount of corrosion hull data and potential data, simulation can be used to obtain the data set, and simulation is performed for different surface current density percentages (i.e. different corrosion degrees). Similarly, multiple points (200 points) 10 meters directly below the keel of each area are averaged to obtain the corresponding UEP, and multiple simulations are performed to obtain a large number of data sets of potential signals corresponding to the corrosion degree of the region. Finally, the data set is divided into a training set and a test set (the training set accounts for 70% and the test set accounts for 30%).

[0071] Based on the above embodiment, the data set is preprocessed to obtain a preprocessed data set, including:

[0072] Splitting the data set into sequences according to the three-dimensional axial components, the length of the sequence being the number of data sets multiplied by 3;

[0073] The corrosion area information and the corrosion degree information are arranged before the sequence to form the preprocessing data set.

[0074] The embodiment of the present invention preprocesses the acquired data set, and the potential vector data of 200 points 10 meters below the keel of any corrosion area with various corrosion degrees can be obtained by cyclic simulation, and a large-scale data set is obtained by multiple cyclic simulations. The data set is processed, and its x, y, and z axis components are arranged into a 1×600 sequence, and the corrosion area information and corrosion degree information are placed in the first and second columns, and an n×1×602 sequence (n is the number of samples) can be obtained, and the data processing is completed.

[0075] On the basis of the above embodiment, the random forest model is trained using the preprocessed data set to obtain a target ship hull coating damage recognition model, including:

[0076] Inputting the feature data and label data in the preprocessed data set into the random forest model;

[0077] The random forest model constructs multiple decision trees based on the feature data, and determines the decision tree results by voting or averaging;

[0078] The hyperparameters of the random forest model are adjusted to perform model optimization, thereby obtaining the target ship hull coating damage identification model.

[0079] Specifically, the embodiment of the present invention uses a random forest algorithm to train a data set. First, the feature data and label data in the data set are input into the random forest model. The random forest algorithm constructs multiple decision trees based on the input feature data, and synthesizes the results of all decision trees by voting or averaging to obtain the final prediction result. During the training process, the hyperparameters of the random forest algorithm, such as the number of trees, the depth of the trees, etc., can be adjusted to optimize the model performance. Techniques such as cross-validation can be used to evaluate the performance of the model, and the hyperparameters can be further adjusted to improve the accuracy and generalization ability of the model.

[0080] After training is completed, the trained random forest model can be used to predict new data. Taking the underwater potential signal of the target ship as input, the model will predict the corresponding corrosion area information and corrosion degree information. By comparing with the actual situation, the accuracy and reliability of the model can be evaluated. For the shortcomings of the model, the algorithm can be further optimized or more feature data can be added to improve the model performance.

[0081] According to the random forest model obtained from the above work, the underwater potential signal of the target ship is input and the model is run to obtain the corrosion area information and corrosion degree information on the hull surface.

[0082] It can be understood that the new hull coating damage identification scheme proposed in the present invention can optimize or replace the existing scheme to a certain extent; the scheme is not restricted by spatial conditions and does not require the ship to be parked in a specific area for detection. It only needs to use underwater potential signals that are more convenient to detect. In theory, all launched ships that are not in a sailing state can be detected; the scheme uses a combination of simulation and practice to reduce the actual workload, and batch detection can be achieved for ships of the same type to improve work efficiency; the data and the entire set of processes of the scheme can be obtained and implemented in a simulated state, which enhances the feasibility of optimizing the scheme according to actual conditions.

[0083] The following is a description of the hull coating damage identification system based on machine learning provided by the present invention. The hull coating damage identification system based on machine learning described below and the hull coating damage identification method based on machine learning described above can be referenced to each other.

[0084] Figure 4 is a structural schematic diagram of a hull coating damage identification system based on machine learning provided by an embodiment of the present invention, such as Figure 4 As shown, it includes: an acquisition module 41, a data processing module 42, a preprocessing module 43, a training module 44 and a recognition module 45, wherein:

[0085] The acquisition module 41 is used to obtain the ship physical field modeling based on electrochemical corrosion; the data processing module 42 is used to set the corrosion area and corrosion degree on the hull surface to obtain the underwater potential data of the hull surface, and use the ship physical field modeling to process the underwater potential data to obtain a data set; the preprocessing module 43 is used to preprocess the data set to obtain a preprocessed data set; the training module 44 is used to train the random forest model using the preprocessed data set to obtain a target ship hull coating damage recognition model; the recognition module 45 is used to input the target ship underwater potential signal into the target ship hull coating damage recognition model, and output the hull coating damage recognition result.

[0086] Figure 5 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 5 As shown, the electronic device may include: a processor 510, a communication interface 520, a memory 530 and a communication bus 540, wherein the processor 510, the communication interface 520 and the memory 530 communicate with each other through the communication bus 540. The processor 510 may call the logic instructions in the memory 530 to execute the ship coating damage identification method based on machine learning, the method comprising: obtaining ship physical field modeling based on electrochemical corrosion; setting the corrosion area and corrosion degree on the hull surface to obtain underwater potential data on the hull surface, processing the underwater potential data using the ship physical field modeling to obtain a data set; preprocessing the data set to obtain a preprocessed data set; using the preprocessed data set to train the random forest model to obtain a target ship hull coating damage identification model; inputting the target ship underwater potential signal into the target ship hull coating damage identification model, and outputting the hull coating damage identification result.

[0087] In addition, the logic instructions in the above-mentioned memory 530 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0088] On the other hand, the present invention also provides a computer program product, which includes a computer program, and the computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the machine learning-based hull coating damage identification method provided by the above-mentioned methods, and the method includes: obtaining ship physical field modeling based on electrochemical corrosion; setting the corrosion area and corrosion degree on the hull surface to obtain underwater potential data of the hull surface, and using the ship physical field modeling to process the underwater potential data to obtain a data set; preprocessing the data set to obtain a preprocessed data set; using the preprocessed data set to train a random forest model to obtain a target ship hull coating damage identification model; inputting the target ship underwater potential signal into the target ship hull coating damage identification model, and outputting the hull coating damage identification result.

[0089] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the machine learning-based hull coating damage identification method provided by the above-mentioned methods, the method comprising: obtaining ship physical field modeling based on electrochemical corrosion; setting corrosion areas and corrosion degrees on the hull surface to obtain underwater potential data on the hull surface, and using the ship physical field modeling to process the underwater potential data to obtain a data set; preprocessing the data set to obtain a preprocessed data set; using the preprocessed data set to train a random forest model to obtain a target ship hull coating damage identification model; inputting the target ship underwater potential signal into the target ship hull coating damage identification model, and outputting the hull coating damage identification result.

[0090] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0091] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0092] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for identifying hull coating damage based on machine learning, characterized in that: include: Get ship physics modeling based on electrochemical corrosion; Setting the corrosion area and corrosion degree on the hull surface to obtain underwater potential data on the hull surface, and processing the underwater potential data using the ship physical field modeling to obtain a data set; Preprocessing the data set to obtain a preprocessed data set; The random forest model is trained using the preprocessed data set to obtain a target ship hull coating damage recognition model; Inputting the underwater potential signal of the target ship into the target ship hull coating damage identification model, and outputting the hull coating damage identification result; Get access to electrochemical corrosion-based ship physics modeling, including: Using modeling software to obtain geometric modeling parameters and basic physical field parameters of the ship, and forming the ship physical field modeling by the geometric modeling parameters and the basic physical field parameters; The basic parameters of the physical field include electrical conductivity, magnetic permeability and material polarization characteristic curve; The corrosion area and the corrosion degree are set on the hull surface to obtain underwater potential data of the hull surface, and the underwater potential data is processed by using the ship physical field modeling to obtain a data set, including: The ship surface is divided into multiple areas according to the length and width of the ship and the preset positioning accuracy; Acquire the location information of the corrosion areas of the same type of ships in the multiple areas and the underwater potential data corresponding to the average of multiple points at a specified distance directly below the keel of the ship to form the data set; Dividing the data set into a training set and a test set; The corrosion area and the corrosion degree are set on the hull surface to obtain underwater potential data of the hull surface, and the underwater potential data is processed by using the ship physical field modeling to obtain a data set, which also includes: The ship physical field modeling is used to perform cyclic simulation on the location information of the corrosion area of ​​the same type of ship and on average multiple points at a specified distance directly below the keel of the ship to obtain underwater potential data after simulation to supplement the data set; The cycle simulation includes: The total current density on the electrode surface is calculated from the local current density related to different reactions on the electrode surface and the number of different reaction parts on the electrode surface; The completely uncovered area is simulated with 100% of the total current density on the electrode surface until the entire area is covered; The hull surface is divided into multiple areas. By dividing a single area more evenly, a certain percentage of random areas are set to be completely corroded in a single area to simulate the effect of percentage corrosion of the entire area, thereby achieving the setting of the degree of corrosion; The random forest model is trained using the preprocessed data set to obtain a target ship hull coating damage recognition model, including: Inputting the feature data and label data in the preprocessed data set into the random forest model; The random forest model constructs multiple decision trees based on the feature data, and determines the decision tree results by voting or averaging; The hyperparameters of the random forest model are adjusted to perform model optimization, thereby obtaining the target ship hull coating damage identification model.

2. The method for identifying hull coating damage based on machine learning according to claim 1 is characterized in that: Preprocessing the data set to obtain a preprocessed data set includes: Splitting the data set into sequences according to the three-dimensional axial components, the length of the sequence being the number of data sets multiplied by 3; The corrosion area information and the corrosion degree information are arranged before the sequence to form the preprocessing data set.

3. A hull coating damage identification system based on machine learning, based on the hull coating damage identification method based on machine learning as claimed in claim 1 or 2, characterized in that: include: Acquisition module, used to acquire ship physical field modeling based on electrochemical corrosion; A data processing module, used for setting the corrosion area and corrosion degree on the hull surface to obtain underwater potential data of the hull surface, and processing the underwater potential data by using the ship physical field modeling to obtain a data set; A preprocessing module, used for preprocessing the data set to obtain a preprocessed data set; A training module, used to train a random forest model using the preprocessed data set to obtain a target ship hull coating damage recognition model; The identification module is used to input the underwater potential signal of the target ship into the hull coating damage identification model of the target ship and output the hull coating damage identification result.

4. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the method for identifying hull coating damage based on machine learning as described in claim 1 or 2 is implemented.

5. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for identifying hull coating damage based on machine learning as described in claim 1 or 2 is implemented.

6. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method for identifying hull coating damage based on machine learning as described in claim 1 or 2 is implemented.

Images