Offshore wind power operation risk prediction method and device based on multi-source data, terminal equipment and storage medium

Abstract

The invention discloses an offshore wind power operation risk prediction method and device based on multi-source data, terminal equipment and a storage medium, and relates to the technical field of offshore wind power operation risk analysis. The operation risk prediction of the offshore wind turbine generator is carried out according to the equipment operation data and the electrical data of each wind turbine generator in an offshore wind power plant, and specifically, the method comprises the following steps: sequentially carrying out time sequence alignment, unified dimension conversion and standardization on the equipment operation characteristics, the environment characteristics and the electrical characteristics to generate a risk characteristic sequence in each time period; prediction distortion caused by dimensional deviation is avoided, the operation risk of the wind turbine generator at each moment in the future is analyzed from multiple aspects of equipment factors, environmental factors and electrical factors by adopting the risk prediction model, and the problem that the fault risk of the wind turbine generator is difficult to predict timely and accurately at present is solved.

Description

Technical Field

[0001] This invention relates to the field of offshore wind power operation risk analysis technology, and in particular to an offshore wind power operation risk prediction method, device, terminal equipment and storage medium based on multi-source data. Background Technology

[0002] Offshore wind power is an important development direction for renewable energy. As a clean and renewable energy source, it has experienced rapid development in recent years. However, offshore wind power operation and maintenance face many challenges, including frequent turbine failures, high maintenance costs, and poor operational stability. Therefore, fault risk analysis of offshore wind power equipment has become a key aspect of ensuring the safe and stable operation of wind farms.

[0003] In the analysis and prediction of offshore wind power failure risks, most methods rely on traditional data statistical analysis and machine learning models, such as support vector machines and random forests. However, traditional methods depend primarily on wind turbine operating data as a single data source for failure detection and risk analysis, making it difficult to predict wind turbine failure risks in a timely and accurate manner. Summary of the Invention

[0004] This invention provides a method, device, terminal equipment, and storage medium for predicting the operational risks of offshore wind power based on multi-source data. The method can solve the current problem of difficulty in predicting the failure risks of wind turbine units in a timely and accurate manner.

[0005] An embodiment of the present invention provides a method for predicting the operational risks of offshore wind power based on multi-source data, including: Acquire environmental data of offshore wind farms over several time periods, as well as equipment operation data and electrical data of each wind turbine unit in the offshore wind farm; Feature extraction is performed on the equipment operation data, the environmental data, and the electrical data respectively to generate equipment operation features, environmental features, and electrical features for each time period; The equipment operation characteristics, environmental characteristics, and electrical characteristics are sequentially aligned in time sequence, transformed in a unified dimension, and standardized to generate risk characteristic sequences for each time period; The risk feature sequence is input into a preset risk prediction model so that the risk prediction model can analyze the operation risk of the wind turbine at various future times based on the risk feature sequence, from equipment factors, environmental factors and electrical factors, and output the operation risk prediction result based on the operation risk.

[0006] Furthermore, the equipment operating data includes: equipment vibration frequency and equipment operating temperature; The step involves extracting features from the equipment operation data, the environmental data, and the electrical data to generate equipment operation features, environmental features, and electrical features for each time period, including: The vibration frequency and operating temperature of the equipment are used as feature data for different channels to construct input samples for a multi-channel tensor. The input sample is input into a preset equipment feature extraction model, so that the equipment feature extraction model uses different convolutional layers to extract features from the feature data in different channels of the input sample, generating vibration frequency features and temperature change features. Then, through a pooling layer, the vibration frequency features and temperature change features are integrated to generate equipment operation features for each time period.

[0007] Furthermore, the environmental data includes: wind data, temperature data, and marine environmental data; The step of extracting features from the equipment operation data, the environmental data, and the electrical data to generate equipment operation features, environmental features, and electrical features for each time period also includes: The wind data, temperature data, and marine environment data are input into a preset environmental feature extraction model, so that the environmental feature extraction model can extract the local fluctuation features of the wind data, temperature data, and marine environment data respectively. Then, a bidirectional gated loop unit is used to capture the correlation features of the wind data, temperature data, and marine environment data in each time period. The correlation features and the local fluctuation features are integrated to generate the environmental features in each time period.

[0008] Furthermore, the electrical data includes: power generation capacity; The step of extracting features from the equipment operation data, the environmental data, and the electrical data to generate equipment operation features, environmental features, and electrical features for each time period also includes: A two-dimensional feature map is constructed based on the power generation and wind power data within a preset time window. The two-dimensional feature map is then input into a preset electrical feature extraction model. The electrical feature extraction model extracts local correlation features of the power generation and wind power data in the short-term time series based on the two-dimensional feature map. According to the preset wind turbine operation rules that characterize the correlation pattern between power generation and wind power, the local correlation features are aggregated and compressed to generate electrical features that characterize whether the power generation in each time period conforms to the wind turbine operation rules.

[0009] Furthermore, the equipment operating characteristics, environmental characteristics, and electrical characteristics are sequentially aligned in time sequence, transformed to a unified dimension, and standardized to generate risk characteristic sequences for each time period, including: Based on a preset time dimension, the equipment operation characteristics, environmental characteristics and electrical characteristics under the same time period are time-aligned, and the time-aligned equipment operation characteristics, environmental characteristics and electrical characteristics are converted into one-dimensional vectors to generate corresponding one-dimensional equipment operation characteristics, one-dimensional environmental characteristics and one-dimensional electrical characteristics. The standard fraction method is used to standardize the one-dimensional equipment operation characteristics, one-dimensional environmental characteristics, and one-dimensional electrical characteristics to generate standard equipment operation characteristics, standard environmental characteristics, and standard electrical characteristics. By splicing together the standard equipment operation characteristics, standard environmental characteristics, and standard electrical characteristics under the same time period, a risk characteristic sequence for each time period is generated.

[0010] Furthermore, inputting the risk feature sequence into a preset risk prediction model includes: Construct a high-dimensional feature matrix based on the risk characteristic sequences of each time period; Principal component analysis is used to reduce the dimensionality of the high-dimensional feature matrix to generate a low-dimensional feature matrix, which is then input into the risk prediction model.

[0011] An embodiment of the present invention also provides an offshore wind power operation risk prediction device based on multi-source data, comprising: The data acquisition module is used to acquire environmental data of offshore wind farms over several time periods, as well as equipment operation data and electrical data of each wind turbine unit in the offshore wind farm. The feature extraction module is used to extract features from the equipment operation data, the environmental data, and the electrical data respectively, and generate equipment operation features, environmental features, and electrical features for each time period. The feature integration module is used to sequentially align, unify, and standardize the device operation features, environmental features, and electrical features to generate risk feature sequences for each time period. The risk prediction module is used to input the risk feature sequence into a preset risk prediction model, so that the risk prediction model can analyze the operation risk of the wind turbine at various future times based on the risk feature sequence, from equipment factors, environmental factors and electrical factors, and output the operation risk prediction result based on the operation risk.

[0012] Furthermore, the equipment operating data includes: equipment vibration frequency and equipment operating temperature; The feature extraction module extracts features from the equipment operation data, the environmental data, and the electrical data respectively, generating equipment operation features, environmental features, and electrical features for each time period, including: The vibration frequency and operating temperature of the equipment are used as feature data for different channels to construct input samples for a multi-channel tensor. The input sample is input into a preset equipment feature extraction model, so that the equipment feature extraction model uses different convolutional layers to extract features from the feature data in different channels of the input sample, generating vibration frequency features and temperature change features. Then, through a pooling layer, the vibration frequency features and temperature change features are integrated to generate equipment operation features for each time period.

[0013] This application also provides a terminal device, including: One or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement a method for predicting the operational risks of offshore wind power based on multi-source data as described in the above embodiments of the invention.

[0014] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for predicting the operational risks of offshore wind power based on multi-source data as described in the above embodiments of the invention.

[0015] The following benefits can be obtained by implementing the present invention: This invention provides a method, apparatus, terminal equipment, and storage medium for predicting the operational risks of offshore wind power based on multi-source data. The method considers that the failures of offshore wind turbines are related to the complex environmental conditions they face during operation. It acquires environmental data, equipment operation data, and electrical data of each wind turbine in the offshore wind farm over several time periods to predict the operational risks of the offshore wind turbines. Specifically, the equipment operation characteristics, environmental characteristics, and electrical characteristics are first sequentially aligned, uniformly transformed, and standardized to integrate the features in terms of dimensions, generating a risk feature sequence for each time period. This avoids prediction distortion caused by dimensional bias. Then, a risk prediction model is used to analyze the operational risks of the wind turbines at various future times from multiple perspectives, including equipment, environmental, and electrical factors. This enables accurate and timely prediction of offshore wind power operational risks, improving the safety of offshore wind power operation. Attached Figure Description

[0016] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating a method for predicting the operational risks of offshore wind power based on multi-source data, provided in a certain embodiment of this application. Figure 2 This is a schematic diagram of the structure of an offshore wind power operation risk prediction device based on multi-source data provided in a certain embodiment of this application; Figure 3 This is a schematic diagram of the structure of a terminal device provided in a certain embodiment of this application; Figure 4 This is another schematic diagram of a method for predicting the operational risks of offshore wind power based on multi-source data, provided in a certain embodiment of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0020] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0021] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0022] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0023] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0024] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0025] See Figure 1 To address the problems in the prior art, an embodiment of the present invention provides a method for predicting the operational risks of offshore wind power based on multi-source data, comprising: S1. Obtain environmental data of offshore wind farms at several time periods, as well as equipment operation data and electrical data of each wind turbine unit in the offshore wind farm. In a preferred embodiment of the present invention, various sensors and monitoring devices are used to collect equipment operation data, electrical data and environmental data of the wind turbine.

[0026] Specifically, wind turbine operating data includes: power output; environmental data such as wind speed, wind direction, air temperature, humidity, salt spray concentration, ocean waves, ocean currents, and tides; and equipment operating data such as vibration signals and equipment operating temperature.

[0027] The collected data undergoes cleaning, noise reduction, and normalization to improve its quality and consistency. This involves removing outliers and filling in missing values. Different normalization methods are used for different data types. For example, for data with clear physical meaning and range (such as temperature and humidity), the Min-Max normalization method is used. This approach ensures data comparability while preserving the original characteristics of the data.

[0028] S2. Extract features from the equipment operation data, the environmental data, and the electrical data respectively to generate equipment operation features, environmental features, and electrical features for each time period; Preferably, the equipment operating data includes: equipment vibration frequency and equipment operating temperature; The step involves extracting features from the equipment operation data, the environmental data, and the electrical data to generate equipment operation features, environmental features, and electrical features for each time period, including: S21. Using the vibration frequency of the equipment and the operating temperature of the equipment as feature data for different channels, respectively, input samples for a multi-channel tensor are constructed. S22. Input the input sample into a preset equipment feature extraction model, so that the equipment feature extraction model uses different convolutional layers to extract features from the feature data in different channels of the input sample, generate vibration frequency features and temperature change features, and integrates the vibration frequency features and temperature change features through a pooling layer to generate equipment operation features for each time period.

[0029] In a preferred embodiment of the present invention, such as Figure 4 As shown, the device feature extraction model is the AlexNet model. The device vibration frequency and operating temperature are treated as different "channels" and combined into a multi-channel input sample (similar to the "RGB three channels" of an image). Specifically, using a fixed time window (e.g., 100 time points) as a unit, each input sample can be constructed as a 2×100 tensor, with channels corresponding to the device vibration frequency and operating temperature, and each channel containing data from 100 time points.

[0030] Furthermore, AlexNet's convolutional layers support multiple input channels. By matching the number of input channels of the convolutional kernel with the channels of the input sample, different convolutional kernels can extract features for different channels. Some convolutional kernels are sensitive to the frequency patterns of vibration channels and extract vibration frequency features; other convolutional kernels focus on the trend changes of temperature channels and extract temperature change features.

[0031] Furthermore, through AlexNet's multi-layer convolutional + pooling structure, features undergo a process of localization, abstraction, and integration. Through shallow convolutional layers, local details of each type of feature are extracted: instantaneous frequency of vibration and short-term temperature fluctuations. Through deep convolutional layers, shallow features are further abstracted to extract the periodicity of vibration frequency and the long-term trend of temperature change. Through fully connected layers, features from different channels are implicitly integrated into device operation features.

[0032] Preferably, the environmental data includes: wind data, temperature data, and marine environmental data; The step of extracting features from the equipment operation data, the environmental data, and the electrical data to generate equipment operation features, environmental features, and electrical features for each time period also includes: S23. Input the wind data, temperature data, and marine environment data into a preset environmental feature extraction model so that the environmental feature extraction model can extract the local fluctuation features of the wind data, temperature data, and marine environment data respectively. Then, a bidirectional gated loop unit is used to capture the correlation features of the wind data, temperature data, and marine environment data in each time period. The correlation features and the local fluctuation features are integrated to generate the environmental features in each time period.

[0033] In a preferred embodiment of the present invention, the environmental feature extraction model is a model based on a convolution-BiGRU-convolution structure. Specifically, the first convolutional layer can initially extract local fluctuation features from wind force data, temperature data, and marine environmental data. The BiGRU (Bidirectional Gated Recurrent Unit) utilizes its bidirectional structure—Forward GRU and Backward GRU—to capture the temporal correlation of environmental factors, such as the chain reaction of wind speed change → wave height change. The BiGRU can simultaneously utilize information such as "wind speed changes first, wave height changes later" (forward) and "wave changes infer wind speed trends" (backward) to extract correlated features more accurately. Finally, the model passes through another convolutional layer to generate the environmental features for each time period.

[0034] Preferably, the electrical data includes: power generation capacity; The step of extracting features from the equipment operation data, the environmental data, and the electrical data to generate equipment operation features, environmental features, and electrical features for each time period also includes: S24. Using a preset time window, construct a two-dimensional feature map based on the power generation and wind power data. Input the two-dimensional feature map into a preset electrical feature extraction model so that the electrical feature extraction model can extract local correlation features of the power generation and wind power data in the short-term time series trend based on the two-dimensional feature map. Based on the preset wind turbine operation law representing the correlation pattern between power generation and wind power, aggregate and compress the local correlation features to generate electrical features used to represent whether the power generation in each time period conforms to the wind turbine operation law.

[0035] In a preferred embodiment of the present invention, the electrical feature extraction model is LeNet. Specifically, during training, LeNet learns the wind turbine operation patterns that characterize the correlation between power generation and wind speed from historical data of wind turbine power and wind speed. The core of LeNet is the convolutional layer, which excels at capturing the "dependencies in local regions" in the input data. For wind turbine data, the power generation and wind speed data are first combined into a two-dimensional feature map, with rows representing time windows, column 1 representing "wind speed value" and column 2 representing "power value", forming a matrix of [number of time steps, 2]. Then, the matrix is ​​adapted to LeNet's input format by dimensional expansion. LeNet's convolutional kernels slide on this two-dimensional feature map, capturing the direct numerical correlation between wind speed and power within local time windows as local correlation features.

[0036] Understandably, although LeNet is not a dedicated time-series model, by arranging wind speed and power data in chronological order (i.e., rows in the two-dimensional feature map corresponding to time sequence), its convolutional layers can indirectly capture correlation patterns in short-term time-series trends. These features reflect the power response patterns when wind speed dynamically changes. Finally, LeNet's pooling layers "aggregate and compress" the local correlation features extracted by the convolutional layers. This process indirectly extracts the statistical attributes of the wind speed-power relationship within a local time window, serving as electrical features to characterize whether the power generation at each time period conforms to the wind turbine's operating patterns.

[0037] S3. The equipment operation characteristics, environmental characteristics, and electrical characteristics are sequentially aligned in time sequence, transformed in a unified dimension, and standardized to generate risk characteristic sequences for each time period; Preferably, the equipment operating characteristics, environmental characteristics, and electrical characteristics are sequentially aligned in time sequence, transformed to a unified dimension, and standardized to generate risk characteristic sequences for each time period, including: S31. Based on the preset time dimension, the equipment operation characteristics, environmental characteristics and electrical characteristics under the same time period are time-aligned, and the time-aligned equipment operation characteristics, environmental characteristics and electrical characteristics are converted into one-dimensional vectors to generate corresponding one-dimensional equipment operation characteristics, one-dimensional environmental characteristics and one-dimensional electrical characteristics. S32. Using the standard fraction method, the one-dimensional equipment operation characteristics, one-dimensional environmental characteristics, and one-dimensional electrical characteristics are standardized to generate standard equipment operation characteristics, standard environmental characteristics, and standard electrical characteristics. S33. Combine the standard equipment operation characteristics, standard environmental characteristics, and standard electrical characteristics under the same time period to generate a risk characteristic sequence for each time period.

[0038] In a preferred embodiment of the present invention, the equipment operation features, environmental features, and electrical features need to be aligned based on the same time dimension to ensure that the "equipment features, unit features, and environmental features" belong to the same moment and avoid feature correlation distortion caused by time misalignment. Furthermore, the output features of the three models need to be converted into one-dimensional vectors. Specifically, the feature map before the last fully connected layer of AlexNet, with shape [C, H, W], is converted into a one-dimensional vector with length C×H×W through global average pooling; LeNet's output features are typically a one-dimensional vector before the fully connected layer; Convolution-BiGRU-Convolution: The output of BiGRU is converted into a one-dimensional vector through flattening after the last convolution layer.

[0039] The features of different models have different numerical scales. The feature value range of AlexNet is [0,10], while the feature value range of LeNet is [100,1000]. If they are directly integrated, the features with larger values ​​will dominate the subsequent analysis. Therefore, the feature vectors output by each model are individually Z-score standardized to generate standard equipment operation features, standard environmental features, and standard electrical features. Finally, the three types of one-dimensional feature vectors after standardization are concatenated according to the sample order to form the risk feature sequence for each time period.

[0040] S4. Input the risk feature sequence into a preset risk prediction model so that the risk prediction model can analyze the operation risk of the wind turbine at various future times based on the risk feature sequence, from equipment factors, environmental factors and electrical factors, and output the operation risk prediction result based on the operation risk.

[0041] Preferably, inputting the risk feature sequence into a preset risk prediction model includes: S41. Construct a high-dimensional feature matrix based on the risk characteristic sequences of each time period; S42. Principal component analysis is used to reduce the dimensionality of the high-dimensional feature matrix to generate a low-dimensional feature matrix, and the low-dimensional feature matrix is ​​then input into the risk prediction model.

[0042] In a preferred embodiment of the present invention, risk feature sequences are integrated to form a high-dimensional feature matrix. The covariance between standardized features in each risk feature sequence of the high-dimensional feature matrix is ​​calculated to construct a covariance matrix. Then, the eigenvalues ​​and eigenvectors of the covariance matrix are calculated. All eigenvalues ​​are sorted from largest to smallest, and the corresponding eigenvectors are also sorted accordingly. According to the preset dimensionality reduction dimension, the first k eigenvectors are taken to construct a principal component matrix. Finally, the high-dimensional feature matrix is ​​multiplied by the principal component matrix to obtain the dimensionality-reduced low-dimensional feature matrix.

[0043] Furthermore, such as Figure 4 As shown, the risk prediction model includes a fully connected layer and an activation layer. The low-dimensional feature matrix first passes through the fully connected layer, which establishes a mapping between features and operational risks. Finally, the activation function maps the features to the output space of the risk results, i.e., the Sigmoid function is used to compress the output to [0,1] to represent the failure probability. Specifically, the operational risk prediction results of wind turbine units at various future times are in the form of, but are not limited to, risk level (low / medium / high), failure probability value, risk score, etc.

[0044] See Figure 2 This invention provides an embodiment of an offshore wind power operation risk prediction device based on multi-source data, comprising: The data acquisition module is used to acquire environmental data of offshore wind farms over several time periods, as well as equipment operation data and electrical data of each wind turbine unit in the offshore wind farm. The feature extraction module is used to extract features from the equipment operation data, the environmental data, and the electrical data respectively, and generate equipment operation features, environmental features, and electrical features for each time period. The feature integration module is used to sequentially align, unify, and standardize the device operation features, environmental features, and electrical features to generate risk feature sequences for each time period. The risk prediction module is used to input the risk feature sequence into a preset risk prediction model, so that the risk prediction model can analyze the operation risk of the wind turbine at various future times based on the risk feature sequence, from equipment factors, environmental factors and electrical factors, and output the operation risk prediction result based on the operation risk.

[0045] Furthermore, the equipment operating data includes: equipment vibration frequency and equipment operating temperature; The feature extraction module extracts features from the equipment operation data, the environmental data, and the electrical data respectively, generating equipment operation features, environmental features, and electrical features for each time period, including: The vibration frequency and operating temperature of the equipment are used as feature data for different channels to construct input samples for a multi-channel tensor. The input sample is input into a preset equipment feature extraction model, so that the equipment feature extraction model uses different convolutional layers to extract features from the feature data in different channels of the input sample, generating vibration frequency features and temperature change features. Then, through a pooling layer, the vibration frequency features and temperature change features are integrated to generate equipment operation features for each time period.

[0046] It is understood that the above-described device embodiments correspond to the method embodiments of the present invention, and can realize the method for predicting offshore wind power operation risks based on multi-source data provided by any of the above-described method embodiments of the present invention.

[0047] It should be noted that the device embodiments described above are merely illustrative, and some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can specifically be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0048] See Figure 3 One embodiment of this application also provides a terminal device, including: One or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the offshore wind power operation risk prediction method based on multi-source data as described above.

[0049] The processor controls the overall operation of the terminal device to complete all or part of the steps of the aforementioned method for predicting offshore wind power operation risks based on multi-source data. The memory stores various types of data to support the operation of the terminal device. This data may include, for example, instructions for any application or method used to operate on the terminal device, as well as application-related data. The memory can be implemented using any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0050] In an exemplary embodiment, the terminal device may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to execute a method for predicting the operational risk of offshore wind power based on multi-source data as described in any of the foregoing embodiments, and to achieve the same technical effects as the methods described above.

[0051] In another exemplary embodiment, a computer-readable storage medium including a computer program is also provided. When executed by a processor, the computer program implements the steps of the offshore wind power operation risk prediction method based on multi-source data as described in any of the foregoing embodiments. For example, the computer-readable storage medium may be the aforementioned memory including the computer program, which may be executed by a processor of a terminal device to complete the offshore wind power operation risk prediction method based on multi-source data as described in any of the foregoing embodiments, and achieve the same technical effects as the aforementioned method.

[0052] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for predicting operational risks of offshore wind power based on multi-source data, characterized in that, include: Acquire environmental data of offshore wind farms over several time periods, as well as equipment operation data and electrical data of each wind turbine unit in the offshore wind farm; Feature extraction is performed on the equipment operation data, the environmental data, and the electrical data respectively to generate equipment operation features, environmental features, and electrical features for each time period; The equipment operation characteristics, environmental characteristics, and electrical characteristics are sequentially aligned in time sequence, transformed in a unified dimension, and standardized to generate risk characteristic sequences for each time period; The risk feature sequence is input into a preset risk prediction model so that the risk prediction model can analyze the operation risk of the wind turbine at various future times based on the risk feature sequence, from the perspectives of equipment factors, environmental factors, and electrical factors, and output the operation risk prediction result based on the operation risk.

2. The method for predicting offshore wind power operation risks based on multi-source data as described in claim 1, characterized in that, The equipment operating data includes: equipment vibration frequency and equipment operating temperature; The step involves extracting features from the equipment operation data, the environmental data, and the electrical data to generate equipment operation features, environmental features, and electrical features for each time period, including: The vibration frequency and operating temperature of the equipment are used as feature data for different channels to construct input samples for a multi-channel tensor. The input sample is input into a preset equipment feature extraction model, so that the equipment feature extraction model uses different convolutional layers to extract features from the feature data in different channels of the input sample, generating vibration frequency features and temperature change features. Then, through a pooling layer, the vibration frequency features and temperature change features are integrated to generate equipment operation features for each time period.

3. The method for predicting offshore wind power operation risks based on multi-source data as described in claim 2, characterized in that, The environmental data includes: wind data, temperature data, and marine environmental data; The step of extracting features from the equipment operation data, the environmental data, and the electrical data to generate equipment operation features, environmental features, and electrical features for each time period also includes: The wind data, temperature data, and marine environment data are input into a preset environmental feature extraction model, so that the environmental feature extraction model can extract the local fluctuation features of the wind data, temperature data, and marine environment data respectively. Then, a bidirectional gated loop unit is used to capture the correlation features of the wind data, temperature data, and marine environment data in each time period. The correlation features and the local fluctuation features are integrated to generate the environmental features in each time period.

4. The method for predicting offshore wind power operation risks based on multi-source data as described in claim 3, characterized in that, The electrical data includes: power generation capacity; The step of extracting features from the equipment operation data, the environmental data, and the electrical data to generate equipment operation features, environmental features, and electrical features for each time period also includes: A two-dimensional feature map is constructed based on the power generation and wind power data within a preset time window. The two-dimensional feature map is then input into a preset electrical feature extraction model. The electrical feature extraction model extracts local correlation features of the power generation and wind power data in the short-term time series based on the two-dimensional feature map. According to the preset wind turbine operation rules that characterize the correlation pattern between power generation and wind power, the local correlation features are aggregated and compressed to generate electrical features that characterize whether the power generation in each time period conforms to the wind turbine operation rules.

5. The method for predicting offshore wind power operation risks based on multi-source data as described in claim 4, characterized in that, The equipment operating characteristics, environmental characteristics, and electrical characteristics are sequentially aligned in time sequence, transformed to a unified dimension, and standardized to generate risk characteristic sequences for each time period, including: Based on a preset time dimension, the equipment operation characteristics, environmental characteristics and electrical characteristics under the same time period are time-aligned, and the time-aligned equipment operation characteristics, environmental characteristics and electrical characteristics are converted into one-dimensional vectors to generate corresponding one-dimensional equipment operation characteristics, one-dimensional environmental characteristics and one-dimensional electrical characteristics. The standard fraction method is used to standardize the one-dimensional equipment operation characteristics, one-dimensional environmental characteristics, and one-dimensional electrical characteristics to generate standard equipment operation characteristics, standard environmental characteristics, and standard electrical characteristics. By splicing together the standard equipment operation characteristics, standard environmental characteristics, and standard electrical characteristics under the same time period, a risk characteristic sequence for each time period is generated.

6. The method for predicting offshore wind power operation risks based on multi-source data as described in claim 5, characterized in that, The step of inputting the risk feature sequence into a preset risk prediction model includes: Construct a high-dimensional feature matrix based on the risk characteristic sequences of each time period; Principal component analysis is used to reduce the dimensionality of the high-dimensional feature matrix to generate a low-dimensional feature matrix, which is then input into the risk prediction model.

7. A device for predicting the operational risks of offshore wind power based on multi-source data, characterized in that, include: The data acquisition module is used to acquire environmental data of offshore wind farms over several time periods, as well as equipment operation data and electrical data of each wind turbine unit in the offshore wind farm. The feature extraction module is used to extract features from the equipment operation data, the environmental data, and the electrical data respectively, and generate equipment operation features, environmental features, and electrical features for each time period. The feature integration module is used to sequentially align, unify, and standardize the device operation features, environmental features, and electrical features to generate risk feature sequences for each time period. The risk prediction module is used to input the risk feature sequence into a preset risk prediction model, so that the risk prediction model can analyze the operation risk of the wind turbine at various future times based on the risk feature sequence, from equipment factors, environmental factors and electrical factors, and output the operation risk prediction result based on the operation risk.

8. The offshore wind power operation risk prediction device based on multi-source data as described in claim 7, characterized in that, The equipment operating data includes: equipment vibration frequency and equipment operating temperature; The feature extraction module extracts features from the equipment operation data, the environmental data, and the electrical data respectively, generating equipment operation features, environmental features, and electrical features for each time period, including: The vibration frequency and operating temperature of the equipment are used as feature data for different channels to construct input samples for a multi-channel tensor. The input sample is input into a preset equipment feature extraction model, so that the equipment feature extraction model uses different convolutional layers to extract features from the feature data in different channels of the input sample, generating vibration frequency features and temperature change features. Then, through a pooling layer, the vibration frequency features and temperature change features are integrated to generate equipment operation features for each time period.

9. A terminal device, characterized in that, include: One or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the offshore wind power operation risk prediction method based on multi-source data as described in any one of claims 1-6.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements a method for predicting the operational risks of offshore wind power based on multi-source data as described in any one of claims 1-6.

Images