CN117975074A - Corrosion state monitoring method and device for offshore wind turbine - Google Patents

Abstract

The application provides a corrosion state monitoring method and device of an offshore wind turbine, which relate to the field of intelligent monitoring of offshore wind turbines and comprise the following steps: acquiring a side image of a single pile foundation of the offshore wind turbine, which is acquired by an underwater camera; the side face image is passed through a first convolution neural network serving as a filter to obtain a side face corrosion feature vector, and the side face corrosion feature vector is passed through a multi-scale neighborhood feature extraction module to obtain a multi-scale side face corrosion feature vector; calculating a transfer matrix among the multi-scale side corrosion feature vectors to obtain a corrosion transfer matrix, and correcting the feature values of the corrosion transfer matrix to obtain a corrected corrosion transfer matrix; arranging the corrected corrosion transfer matrix into a three-dimensional input tensor, and then passing through a second convolution neural network serving as a feature extractor to obtain a classification feature map; and (5) passing the classification characteristic diagram through a classifier to obtain a classification result. The application is based on artificial intelligent monitoring and analysis of the corrosion state of the fan, and early warning is carried out in advance to avoid corrosion hazard.

Description

Corrosion state monitoring method and device for offshore wind turbine

Technical Field

The application relates to the field of intelligent monitoring of offshore wind turbines, in particular to a corrosion state monitoring method and device of an offshore wind turbine.

Background

The single pile foundation is the most commonly used type of fan foundation in shallow to medium water deep sea wind farms. Under the policy background of global coping with climate change, along with the increasing maturity of offshore wind power technology and the great reduction of offshore wind power electricity cost, particularly the proposal of a carbon peak and carbon neutralization double-carbon target plan, more and more offshore wind farm construction projects are expected to be put into practice, and the application scenes of single pile foundations are also more and more.

However, the offshore wind turbines are in a much more complex and harsher environment than onshore wind turbines. The high humidity and high salinity of the sea surface atmosphere area, the dry-wet alternation of the splash area and the tidal range area, the seawater soaking of the immersed area, the attachment of marine organisms and other severe corrosion environments can cause serious corrosion hazard to the single pile foundation of the offshore wind turbine.

Therefore, the corrosion state of the offshore wind power single pile foundation needs to be monitored for risk early warning.

Disclosure of Invention

Aiming at the problems, the corrosion state monitoring method and device of the offshore wind turbine are provided, by adopting an artificial intelligence monitoring technology, local implicit correlation characteristics of a plurality of side images of a single pile foundation of the offshore wind turbine are excavated by using a convolutional neural network model based on deep learning, and economic decomposition of high-dimensional manifolds is promoted by calculating distance combination of symbolized functions, so that dimension monotonicity among the high-dimensional manifolds of each corrosion transfer matrix is improved, classification effect is further improved, and the corrosion state of the single pile foundation is accurately monitored to early warn to avoid corrosion hazard.

The first aspect of the application provides a corrosion state monitoring method of an offshore wind turbine, comprising the following steps:

Acquiring a side image of a single pile foundation of the offshore wind turbine, which is acquired by an underwater camera;

The side image is passed through a first convolution neural network serving as a filter to obtain a side corrosion feature vector, and the side corrosion feature vector is passed through a multi-scale neighborhood feature extraction module to obtain a multi-scale side corrosion feature vector;

calculating a transfer matrix among the multi-scale side corrosion feature vectors to obtain a corrosion transfer matrix, and correcting the feature values of the corrosion transfer matrix to obtain a corrected corrosion transfer matrix;

arranging the corrected corrosion transfer matrix into a three-dimensional input tensor, and then passing through a second convolution neural network serving as a feature extractor to obtain a classification feature map;

and the classification characteristic diagram is passed through a classifier to obtain a classification result, wherein the classification result is used for representing the corrosion state of the single pile foundation.

Optionally, the step of passing the side image through a first convolutional neural network as a filter to obtain a side erosion feature vector includes:

carrying out convolution processing on input data to obtain a first convolution characteristic diagram;

Carrying out local feature matrix-based mean pooling on the first convolution feature map to obtain a first pooled feature map;

Non-linear activation is carried out on the first pooling feature map so as to obtain an activation feature map;

The output of the last layer of the first convolutional neural network is the side corrosion characteristic vector, and the input of the first layer of the first convolutional neural network is the side image.

Optionally, the step of passing the side corrosion feature vector through a multi-scale neighborhood feature extraction module to obtain a multi-scale side corrosion feature vector includes:

Checking the side corrosion feature vector by using a first convolution layer of the multi-scale neighborhood feature extraction module through a first one-dimensional convolution, and performing one-dimensional convolution coding to obtain a first-scale side corrosion feature vector;

Checking the side corrosion feature vector by using a second convolution layer of the multi-scale neighborhood feature extraction module through a second one-dimensional convolution, and performing one-dimensional convolution coding to obtain a second-scale side corrosion feature vector, wherein the size of the first one-dimensional convolution kernel is different from that of the second one-dimensional convolution kernel;

And cascading the first scale side surface corrosion feature vector and the second scale side surface corrosion feature vector to obtain the multi-scale side surface corrosion feature vector.

Optionally, the transfer matrix between the multiscale lateral corrosion feature vectors is calculated according to the following formula to obtain a corrosion transfer matrix, wherein the formula is as follows:

V₁＝M*V₂，

Wherein M represents the corrosion transfer matrix, and V ₁ and V ₂ represent each two of the multi-scale side corrosion feature vectors, respectively.

Optionally, the performing eigenvalue correction on the corrosion transfer matrix to obtain a corrected corrosion transfer matrix includes:

calculating an object-friendly decomposition incentive factor for the corrosion transfer matrix, wherein the object-friendly decomposition incentive factor is related to a weighted sum of natural exponential function values raised to a power by negative values of eigenvalues at respective positions in respective column vectors in the corrosion transfer matrix;

And weighting each corrosion transfer matrix in the corrosion transfer matrix by taking the object centipede decomposition encouraging factor of the corrosion transfer matrix as a weight so as to obtain the corrected corrosion transfer matrix.

Optionally, the object-level economic decomposition encouragement factor of the corrosion transfer matrix is calculated by the formula:

Wherein m _j,k∈M_i,m_j,k is a characteristic value of an ith corrosion transfer matrix in the corrosion transfer matrices, τ _i is the penalty factor for the ith one of the corrosion transfer matrices, ||·| ₂ denotes a feature the two norms of the matrix.

Optionally, the arranging the corrected erosion transfer matrix into a three-dimensional input tensor and then passing through a second convolutional neural network serving as a feature extractor to obtain a classification feature map, which includes:

And each layer of the second convolutional neural network carries out convolution processing, pooling processing and activation processing on input data in forward transmission of the layers, and the last layer of the second convolutional neural network generates the classification characteristic diagram, wherein the input of the first layer of the second convolutional neural network is the three-dimensional input tensor.

Optionally, the classifying feature map is passed through a classifier to obtain a classification result, where the classification result is used to represent a single pile foundation corrosion state, and the method includes:

The classifier processes the classification feature map to generate a classification result with the following formula:

softmax{(W_n,B_n):…:(W₁,B₁)|Project(F)}，

Where Project (F) represents projection of the classification feature map as a vector, W ₁ to W _n are weight matrices of each fully connected layer, and B ₁ to B _n represent bias matrices of each fully connected layer.

The second aspect of the present application provides a corrosion state monitoring device for an offshore wind turbine, comprising:

the image data acquisition unit is used for acquiring side images of the offshore wind turbine single pile foundation acquired by the underwater camera;

A first feature extraction unit, configured to pass the side image through a first convolutional neural network serving as a filter to obtain a side corrosion feature vector;

The multiscale domain unit is used for enabling the side corrosion feature vector to pass through a multiscale neighborhood feature extraction module to obtain a multiscale side corrosion feature vector;

The transfer matrix calculation unit is used for calculating a transfer matrix among the multi-scale side corrosion characteristic vectors to obtain a corrosion transfer matrix;

the correction unit is used for correcting the characteristic value of the corrosion transfer matrix to obtain a corrected corrosion transfer matrix;

a second feature extraction unit, configured to arrange the corrected erosion transfer matrix into a three-dimensional input tensor, and then obtain a classification feature map through a second convolutional neural network serving as a feature extractor;

And the classification unit is used for passing the classification characteristic diagram through a classifier to obtain a classification result, wherein the classification result is used for representing the single pile foundation corrosion state.

In a third aspect the application provides a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing a method as in any of the first aspects above when executing the computer program.

The technical scheme provided by the embodiment of the application at least has the following beneficial effects:

The method has the advantages that the local implicit correlation characteristics of a plurality of side images of the single pile foundation of the offshore wind turbine are excavated by adopting an artificial intelligence monitoring technology and using a convolutional neural network model based on deep learning, the economic decomposition of high-dimensional manifold is promoted by calculating the distance combination of symbolization functions, the dimension monotonicity among the high-dimensional manifold of each corrosion transfer matrix is improved, the classification effect is improved, the corrosion state of the single pile foundation is accurately monitored, and early warning is performed in advance to avoid corrosion hazard.

Additional aspects and advantages of the application will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the application.

Drawings

The foregoing and/or additional aspects and advantages of the application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings, in which:

FIG. 1 is a flow chart illustrating a method of monitoring corrosion status of an offshore wind turbine according to an exemplary embodiment of the application;

FIG. 2 is a flow chart illustrating a method of monitoring corrosion status of an offshore wind turbine according to an exemplary embodiment of the application;

FIG. 3 is a flow chart illustrating a method of monitoring corrosion status of an offshore wind turbine according to an exemplary embodiment of the application;

FIG. 4 is a flow chart illustrating a method of monitoring corrosion status of an offshore wind turbine according to an exemplary embodiment of the application;

FIG. 5 is an application scenario diagram illustrating a method of monitoring corrosion status of an offshore wind turbine according to an exemplary embodiment of the application;

FIG. 6 is a block diagram illustrating a corrosion status monitoring device for an offshore wind turbine according to an exemplary embodiment of the application;

fig. 7 is a block diagram of an electronic device.

Detailed Description

Embodiments of the present application are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are illustrative and intended to explain the present application and should not be construed as limiting the application.

Due to the severe and complex marine environment, the corrosion type of the offshore wind power single pile foundation presents various new characteristics. The type of single pile foundation corrosion can be morphologically classified into uniform corrosion and localized corrosion. Wherein the uniform corrosion is caused by microcell effect, the corrosion rate is uniformly distributed, and cathode and anode portions of the metal surface alternate during the corrosion process, without fixed cathode and anode. Localized corrosion includes pitting, crevice corrosion, stress corrosion, impact corrosion, galvanic corrosion, bioerosion, and the like. The local corrosion is hidden and difficult to detect, and often causes disastrous accidents and has great hazard. The conditions for generating the corrosion are related to the material and structure of the single pile foundation, and the environment of the single pile foundation is more inseparable.

Based on this, the present application considers that whether it is uniform corrosion or partial corrosion of the monopile foundation, detection judgment can be made by the side image of the monopile foundation, and in order to detect the overall corrosion condition of the monopile foundation, it is desirable to use a plurality of cameras to take a plurality of side images of the monopile foundation from a plurality of angles. This is essentially a classification problem, namely, deep feature mining of multiple side images using deep neural network models, and then classification judgment of whether or not a single pile foundation corrosion state pre-warning is generated using a classifier.

FIG. 1 is a method for monitoring corrosion status of an offshore wind turbine according to an exemplary embodiment of the application, comprising:

And step 101, acquiring side images of the offshore wind turbine single pile foundation acquired by the underwater camera.

In the embodiment of the application, whether uniform corrosion or partial corrosion of the single pile foundation is considered, detection and judgment can be carried out through the side images of the single pile foundation, and a plurality of cameras are used for shooting a plurality of side images of the single pile foundation from a plurality of angles for detecting the whole corrosion condition of the single pile foundation.

Step 102, the side image is passed through a first convolution neural network serving as a filter to obtain a side corrosion feature vector, and the side corrosion feature vector is passed through a multi-scale neighborhood feature extraction module to obtain a multi-scale side corrosion feature vector.

In the embodiment of the application, a process of obtaining the side corrosion characteristic vector and a processing process of obtaining the multi-scale side corrosion characteristic vector are respectively specifically described.

As shown in fig. 2, the process of obtaining the side corrosion feature vector in step 102 specifically includes:

step 201, performing convolution processing on input data to obtain a first convolution feature map;

step 202, carrying out local feature matrix-based mean pooling on a first convolution feature graph to obtain a first pooled feature graph;

Step 203, performing nonlinear activation on the first pooled feature map to obtain an activated feature map;

the output of the last layer of the first convolutional neural network is a side corrosion characteristic vector, and the input of the first layer of the first convolutional neural network is a side image.

In the embodiment of the application, each side image in the plurality of side images is processed through the first convolution neural network to determine a plurality of side corrosion feature vectors, and in addition, each layer of the first convolution neural network serving as a filter performs on input data in forward transfer of the layer.

As shown in fig. 3, the processing for obtaining the multiscale lateral corrosion feature vector in step 102 is specifically:

Step 301, checking the side corrosion feature vector by using a first convolution layer of a multi-scale neighborhood feature extraction module through a first one-dimensional convolution and performing one-dimensional convolution coding to obtain a first-scale side corrosion feature vector;

Step 302, checking the side corrosion feature vector with a second one-dimensional convolution layer by using a second convolution layer of the multi-scale neighborhood feature extraction module and performing one-dimensional convolution coding to obtain a second scale side corrosion feature vector, wherein the size of a first one-dimensional convolution kernel is different from that of a second one-dimensional convolution kernel;

Step 303, cascading the first scale side etching feature vector and the second scale side etching feature vector to obtain a multi-scale side etching feature vector.

In the embodiment of the present application, for example, a time sequence convolution structure with a convolution kernel size of 3, for time sequence data input, the convolution kernel moves along a time dimension in the form of a sliding window, and outputs a weighted sum of data in each time sequence segment. Each convolution unit stacks a plurality of convolution kernels to output a multi-dimensional feature. The large convolution kernel extracts features from the large-scale time sequence neighborhood, wherein the influence of each numerical value in the neighborhood is smaller, so that fluctuation of input data is weakened, and the influence of noise points on the output features is lightened. However, the large-scale convolution kernel weakens the difference of numerical variation, and easily causes the problem of smooth transition, so that the output characteristics lose the discrimination capability. In contrast, small scale convolution kernels are better able to preserve information in the input data, but are also more susceptible to interference from noise therein. Therefore, the characteristics of convolution of different scales are considered, and the convolution units of different sizes are used in combination to extract the characteristics of different time sequence scales. And then, feature fusion is completed in a feature splicing mode, so that multi-scale neighborhood features are obtained, and in this way, multi-scale neighborhood relations among different corrosion features of each side face are extracted, so that the output features not only comprise the smoothed features, but also preserve the features of the original input, avoid information loss, and improve the accuracy of subsequent classification.

And 103, calculating a transfer matrix among the multi-scale side corrosion feature vectors to obtain a corrosion transfer matrix, and correcting the feature values of the corrosion transfer matrix to obtain a corrected corrosion transfer matrix.

In the embodiment of the application, a transfer matrix between every two multiscale side corrosion feature vectors in the multiscale side corrosion feature vectors is further calculated to obtain a plurality of corrosion transfer matrices, so that transfer features, namely differential features, between every two multiscale side corrosion feature vectors with multilevel scales in the multiscale side corrosion feature vectors are expressed.

And calculating a transfer matrix and obtaining a corrosion transfer matrix according to the following formula:

V₁＝M*V₂，

where M represents the corrosion transfer matrix and V ₁ and V ₂ represent each two of the multi-scale side corrosion feature vectors.

In the embodiment of the present application, as shown in fig. 4, the process of correcting the eigenvalue of the corrosion transfer matrix to obtain the corrected corrosion transfer matrix specifically includes:

Step 401, calculating an object-friendly decomposition incentive factor of the corrosion transfer matrix, wherein the object-friendly decomposition incentive factor is related to a weighted sum of natural exponential function values of which the negative value of the eigenvalue of each position in each column vector in the corrosion transfer matrix is a power;

and step 402, weighting each corrosion transfer matrix in the corrosion transfer matrix by taking the centipede decomposition encouraging factors of the corrosion transfer matrix as weights so as to obtain a corrected corrosion transfer matrix.

In the embodiment of the application, for a plurality of corrosion transfer matrices, each corrosion transfer matrix represents transfer characteristics between two multi-scale side corrosion characteristic vectors with multi-level scales, so that each corrosion transfer matrix is easy to cause poor dimensional monotonicity of a high-dimensional manifold due to scale variability, and the obtained classification characteristic map is difficult to converge into a preset classification target domain.

Thus, the present application further calculates the object-level economic decomposition encouragement factor for each corrosion transfer matrix as a weight. It will be appreciated that here the object cursory decomposition encouragement factor is able to group features in a predetermined direction and to impose a penalty on the overlap of elements within the group, thereby promoting cursory decomposition of the high-dimensional manifolds by computing a distance union of symbolizing functions, by weighting each corrosion transfer matrix with it as a weighting factor, it can be understood geometrically that the geometry of the high-dimensional manifolds of each corrosion transfer matrix is constructed based on a predetermined convex polygon set, thus improving the dimensional monotonicity between the high-dimensional manifolds of each corrosion transfer matrix and improving the classification effect.

Optionally, the object-savable decomposition encouragement factor for each of the plurality of corrosion transfer matrices is calculated as:

Wherein m _j,k∈M_i is a characteristic value of an ith corrosion transfer matrix of the plurality of corrosion transfer matrices, and τ _i is a penalty factor for an ith one of the plurality of corrosion transfer matrices, ||·| ₂ denotes a feature the two norms of the matrix.

And 104, arranging the corrected corrosion transfer matrix into a three-dimensional input tensor, and then passing through a second convolution neural network serving as a feature extractor to obtain a classification feature map.

In the embodiment of the application, after a plurality of corrected corrosion transfer matrixes are obtained, the corrected corrosion transfer matrixes are arranged into three-dimensional input tensors to integrate hidden association features of the corrected corrosion transfer matrixes, and the hidden association features are processed in a second convolution neural network serving as a feature extractor to extract association features of differences among corrosion features among different sides, so that a classification feature graph is obtained.

Optionally, each layer of the second convolutional neural network performs convolutional processing, pooling processing and activating processing on input data in forward transfer of the layer, and a last layer of the second convolutional neural network generates a classification characteristic map, wherein the input of a first layer of the second convolutional neural network is a three-dimensional input tensor.

And 105, passing the classification characteristic diagram through a classifier to obtain a classification result, wherein the classification result is used for representing the corrosion state of the single pile foundation.

The classifier processes the classification feature map to generate a classification result with the following formula:

softmax{(W_n,B_n):…:(W₁,B₁)|Project(F)}，

Where Project (F) represents projection of the classification feature map as a vector, W ₁ to W _n are weight matrices of each fully connected layer, and B ₁ to B _n represent bias matrices of each fully connected layer.

The method shown in the embodiment of the application is applied to an application scene graph shown in fig. 5, and as shown in fig. 5, a plurality of side images of a single pile foundation T of an offshore wind turbine F are acquired through a plurality of underwater cameras C. And then, inputting the obtained multiple side images of the single pile foundation T of the offshore wind turbine F into a server S provided with a corrosion state monitoring algorithm of the offshore wind turbine F, wherein the server S can process the obtained multiple side images of the single pile foundation T of the offshore wind turbine F by using the corrosion state monitoring algorithm of the offshore wind turbine F to generate a classification result for indicating whether early warning of the corrosion state of the single pile foundation is generated.

According to the embodiment of the application, the local implicit correlation characteristics of a plurality of side images of the single pile foundation of the offshore wind turbine are excavated by adopting an artificial intelligence monitoring technology and using a convolutional neural network model based on deep learning, and the economic decomposition of high-dimensional manifolds is promoted by calculating the distance combination of symbolic functions, so that the dimensional monotonicity among the high-dimensional manifolds of each corrosion transfer matrix is improved, the classification effect is further improved, the corrosion state of the single pile foundation is accurately monitored, and early warning is performed in advance to avoid corrosion hazard.

Fig. 6 is a block diagram illustrating a corrosion state monitoring apparatus 600 of an offshore wind turbine according to an exemplary embodiment of the present application, which includes an image data acquisition unit 610, a first feature extraction unit 620, a multi-scale domain unit 630, a transfer matrix calculation unit 640, a correction unit 650, a second feature extraction unit 660, and a classification unit 670.

An image data acquisition unit 610, configured to acquire a side image of a single pile foundation of an offshore wind turbine acquired by an underwater camera;

a first feature extraction unit 620 for passing the side image through a first convolutional neural network as a filter to obtain a side erosion feature vector;

a multi-scale domain unit 630, configured to pass the side corrosion feature vector through a multi-scale neighborhood feature extraction module to obtain a multi-scale side corrosion feature vector;

A transfer matrix calculation unit 640 for calculating a transfer matrix between the multi-scale side corrosion feature vectors to obtain a corrosion transfer matrix;

A correction unit 650 for performing eigenvalue correction on the corrosion transfer matrix to obtain a corrected corrosion transfer matrix;

A second feature extraction unit 660 for arranging the corrected erosion transfer matrices into three-dimensional input tensors and then passing through a second convolutional neural network as a feature extractor to obtain a classification feature map;

And the classification unit 670 is used for passing the classification characteristic diagram through a classifier to obtain a classification result, wherein the classification result is used for representing the single pile foundation corrosion state.

The specific manner in which the various modules perform the operations in the apparatus of the above embodiments have been described in detail in connection with the embodiments of the method, and will not be described in detail herein.

Fig. 7 illustrates a schematic block diagram of an example electronic device 700 that may be used to implement embodiments of the present disclosure. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the disclosure described and/or claimed herein.

As shown in fig. 7, the apparatus 700 includes a computing unit 701 that can perform various appropriate actions and processes according to a computer program stored in a Read Only Memory (ROM) 702 or a computer program loaded from a storage unit 708 into a Random Access Memory (RAM) 703. In the RAM 703, various programs and data required for the operation of the device 700 may also be stored. The computing unit 701, the ROM 702, and the RAM 703 are connected to each other through a bus 704. An input/output (I/O) interface 705 is also connected to bus 704.

Various components in device 700 are connected to I/O interface 707, including: an input unit 706 such as a keyboard, a mouse, etc.; an output unit 707 such as various types of displays, speakers, and the like; a storage unit 708 such as a magnetic disk, an optical disk, or the like; and a communication unit 709 such as a network card, modem, wireless communication transceiver, etc. The communication unit 709 allows the device 700 to exchange information/data with other devices via a computer network, such as the internet, and/or various telecommunication networks.

The computing unit 701 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of computing unit 701 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various computing units running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 701 performs the respective methods and processes described above, such as a voice instruction response method. For example, in some embodiments, the voice instruction response method may be implemented as a computer software program tangibly embodied on a machine-readable medium, such as storage unit 708. In some embodiments, part or all of the computer program may be loaded and/or installed onto device 700 via ROM 702 and/or communication unit 709. When the computer program is loaded into RAM 703 and executed by computing unit 701, one or more steps of the voice instruction response method described above may be performed. Alternatively, in other embodiments, the computing unit 701 may be configured to perform the voice instruction response method by any other suitable means (e.g., by means of firmware).

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.

Program code for carrying out methods of the present disclosure may be written in any combination of one or more programming languages. These program code may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus such that the program code, when executed by the processor or controller, causes the functions/operations specified in the flowchart and/or block diagram to be implemented. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of this disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. The machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and pointing device (e.g., a mouse or trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), the internet, and blockchain networks.

The computer system may include a client and a server. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also called a cloud computing server or a cloud host, and is a host product in a cloud computing service system, so that the defects of high management difficulty and weak service expansibility in the traditional physical hosts and VPS service ("Virtual PRIVATE SERVER" or simply "VPS") are overcome. The server may also be a server of a distributed system or a server that incorporates a blockchain.

It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps recited in the present disclosure may be performed in parallel or sequentially or in a different order, provided that the desired results of the technical solutions of the present disclosure are achieved, and are not limited herein.

The above detailed description should not be taken as limiting the scope of the present disclosure. It will be apparent to those skilled in the art that various modifications, combinations, sub-combinations and alternatives are possible, depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present disclosure are intended to be included within the scope of the present disclosure.

Claims (10)

1. A method for monitoring corrosion status of an offshore wind turbine, comprising:

Acquiring a side image of a single pile foundation of the offshore wind turbine, which is acquired by an underwater camera;

The side image is passed through a first convolution neural network serving as a filter to obtain a side corrosion feature vector, and the side corrosion feature vector is passed through a multi-scale neighborhood feature extraction module to obtain a multi-scale side corrosion feature vector;

calculating a transfer matrix among the multi-scale side corrosion feature vectors to obtain a corrosion transfer matrix, and correcting the feature values of the corrosion transfer matrix to obtain a corrected corrosion transfer matrix;

arranging the corrected corrosion transfer matrix into a three-dimensional input tensor, and then passing through a second convolution neural network serving as a feature extractor to obtain a classification feature map;

and the classification characteristic diagram is passed through a classifier to obtain a classification result, wherein the classification result is used for representing the corrosion state of the single pile foundation.

2. The method of claim 1, wherein passing the side image through a first convolutional neural network as a filter to obtain a side erosion feature vector comprises:

carrying out convolution processing on input data to obtain a first convolution characteristic diagram;

Carrying out local feature matrix-based mean pooling on the first convolution feature map to obtain a first pooled feature map;

Non-linear activation is carried out on the first pooling feature map so as to obtain an activation feature map;

The output of the last layer of the first convolutional neural network is the side corrosion characteristic vector, and the input of the first layer of the first convolutional neural network is the side image.

3. The method of claim 1, wherein passing the side etch feature vector through a multi-scale neighborhood feature extraction module to obtain a multi-scale side etch feature vector comprises:

Checking the side corrosion feature vector by using a first convolution layer of the multi-scale neighborhood feature extraction module through a first one-dimensional convolution, and performing one-dimensional convolution coding to obtain a first-scale side corrosion feature vector;

Checking the side corrosion feature vector by using a second convolution layer of the multi-scale neighborhood feature extraction module through a second one-dimensional convolution, and performing one-dimensional convolution coding to obtain a second-scale side corrosion feature vector, wherein the size of the first one-dimensional convolution kernel is different from that of the second one-dimensional convolution kernel;

And cascading the first scale side surface corrosion feature vector and the second scale side surface corrosion feature vector to obtain the multi-scale side surface corrosion feature vector.

4. The method of claim 1, wherein the transfer matrix between the multi-scale side etch feature vectors is calculated to obtain an etch transfer matrix according to the formula:

V₁＝M*V₂，

Wherein M represents the corrosion transfer matrix, and V ₁ and V ₂ represent each two of the multi-scale side corrosion feature vectors, respectively.

5. The method of claim 1, wherein performing eigenvalue correction on the corrosion transfer matrix to obtain a corrected corrosion transfer matrix comprises:

calculating an object-friendly decomposition incentive factor for the corrosion transfer matrix, wherein the object-friendly decomposition incentive factor is related to a weighted sum of natural exponential function values raised to a power by negative values of eigenvalues at respective positions in respective column vectors in the corrosion transfer matrix;

And weighting each corrosion transfer matrix in the corrosion transfer matrix by taking the object centipede decomposition encouraging factor of the corrosion transfer matrix as a weight so as to obtain the corrected corrosion transfer matrix.

6. The method of claim 5, wherein the object-level factorization encouragement factor of the corrosion transfer matrix is calculated by the formula:

Wherein m _j,k∈M_i,m_j,k is a characteristic value of an ith corrosion transfer matrix in the corrosion transfer matrices, τ _i is the penalty factor for the ith one of the corrosion transfer matrices, ||·| ₂ denotes a feature the two norms of the matrix.

7. The method of claim 1, wherein the arranging the corrected erosion transfer matrix into a three-dimensional input tensor followed by a second convolutional neural network as a feature extractor to obtain a classification feature map comprises:

And each layer of the second convolutional neural network carries out convolution processing, pooling processing and activation processing on input data in forward transmission of the layers, and the last layer of the second convolutional neural network generates the classification characteristic diagram, wherein the input of the first layer of the second convolutional neural network is the three-dimensional input tensor.

8. The method of claim 1, wherein the passing the classification feature map through a classifier to obtain a classification result, wherein the classification result is used to represent a monopile foundation corrosion state, comprises:

The classifier processes the classification feature map to generate a classification result with the following formula:

softmax{(W_n,B_n):…:(W₁,B₁)|Project(F)}，

Where Project (F) represents projection of the classification feature map as a vector, W ₁ to W _n are weight matrices of each fully connected layer, and B ₁ to B _n represent bias matrices of each fully connected layer.

9. A corrosion state monitoring device for an offshore wind turbine, comprising:

the image data acquisition unit is used for acquiring side images of the offshore wind turbine single pile foundation acquired by the underwater camera;

A first feature extraction unit, configured to pass the side image through a first convolutional neural network serving as a filter to obtain a side corrosion feature vector;

The multiscale domain unit is used for enabling the side corrosion feature vector to pass through a multiscale neighborhood feature extraction module to obtain a multiscale side corrosion feature vector;

The transfer matrix calculation unit is used for calculating a transfer matrix among the multi-scale side corrosion characteristic vectors to obtain a corrosion transfer matrix;

the correction unit is used for correcting the characteristic value of the corrosion transfer matrix to obtain a corrected corrosion transfer matrix;

a second feature extraction unit, configured to arrange the corrected erosion transfer matrix into a three-dimensional input tensor, and then obtain a classification feature map through a second convolutional neural network serving as a feature extractor;

And the classification unit is used for passing the classification characteristic diagram through a classifier to obtain a classification result, wherein the classification result is used for representing the single pile foundation corrosion state.

10. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the method of any of claims 1-8 when executing the computer program.