Wind power plant environmental parameter identification and control method, system, equipment, medium and program

Abstract

The invention provides a wind power plant environmental parameter identification and control method, system, device, medium and program, and belongs to the technical field of wind power plants. The method comprises the steps of collecting a plurality of types of data in a wind power plant to form multi-source heterogeneous data; preprocessing the multi-source heterogeneous data to obtain the processed multi-source heterogeneous data; constructing an environmental parameter identification model by using a deep neural network, and performing environmental parameter identification processing on the processed multi-source heterogeneous data to obtain control parameters of the wind turbine generator; optimizing the obtained control parameters of the wind turbine generator by using a particle swarm optimization algorithm to obtain optimized control parameters; and adjusting the running state of the wind turbine generator according to the optimized control parameters. According to the method, the multi-source heterogeneous data in the wind power plant environment is identified, so that the identification degree of environmental parameters is improved, and a control strategy can adapt to various conditions.

Description

Technical Field

[0001] This invention relates to the field of wind farm technology, specifically to a method, system, equipment, medium, and program for identifying and controlling environmental parameters in wind farms. Background Technology

[0002] Wind power has become an important component of clean energy, especially in the construction of large-scale wind farms, where the operational efficiency and equipment safety of wind turbines are crucial to their power generation benefits. The power generation efficiency and equipment safety of wind farms highly depend on the accurate monitoring and control of environmental parameters, primarily including wind speed, wind direction, temperature, and air pressure. These environmental parameters directly affect the operating status and power generation capacity of wind turbines. Therefore, obtaining accurate environmental data and adjusting wind turbine operating strategies based on this data is key to improving power generation efficiency and ensuring equipment safety. However, with the gradual expansion of wind farm scale, the types and sources of environmental data are increasing. How to effectively integrate multi-source heterogeneous data and improve the accuracy of environmental parameter monitoring and identification has become an urgent problem to be solved in wind farm operation.

[0003] Existing methods for monitoring and controlling environmental parameters in wind farms have several limitations. First, traditional monitoring methods based on a single data source, such as relying on sensors on the wind turbine itself or data from weather stations, suffer from limited data volume and are easily affected by equipment failures or localized environmental factors. This method cannot comprehensively reflect the complex environmental conditions of wind farms and is limited by factors such as sensor performance and installation location, resulting in low accuracy and reliability of the monitoring data. Therefore, environmental monitoring schemes relying solely on a single data source cannot adequately guarantee the safe and efficient operation of wind farms.

[0004] Secondly, many traditional environmental parameter identification methods rely on simple linear models, such as linear regression, to describe the relationship between environmental parameters and power generation. While these methods can achieve some success in certain simple scenarios, the nonlinear, time-varying, and complex dynamic characteristics of wind farm environments make it difficult for a single linear model to accurately capture their changing patterns. The relationships between environmental parameters such as wind speed and temperature often exhibit nonlinear characteristics and are subject to interference and influence from various factors. Simple linear models cannot fully reflect the complex behavior of wind farms, resulting in low accuracy in environmental parameter identification, which in turn affects the accuracy and effectiveness of control strategies.

[0005] Finally, wind farm control strategies are typically designed based on experience or pre-set rules, such as setting fixed parameters like pitch angle and yaw angle to adjust the operation of wind turbines. This fixed control strategy fails to consider real-time changes in environmental parameters and struggles to adapt to dynamic changes in environmental conditions such as wind speed and direction. As wind farms expand and environmental conditions continuously change, traditional control strategies often cannot adjust operating states in real time, leading to equipment operating outside of optimal conditions and even the risk of equipment damage. The lack of flexibility and adaptability in fixed control strategies can result in wind turbines exhibiting low power generation efficiency or excessive energy consumption in specific environments. Summary of the Invention

[0006] To address the problems of low accuracy in identifying environmental parameters and a lack of flexibility and adaptability in control strategies in existing wind farm environmental parameter monitoring and control methods, which are mainly based on a single data source or a simple linear model, this invention provides a wind farm environmental parameter identification and control method. By identifying multi-source heterogeneous data in the wind farm environment, the method improves the identification accuracy of environmental parameters, enabling control strategies to adapt to various situations.

[0007] To achieve the above objectives, the present invention provides the following technical solution.

[0008] In a first aspect, the present invention provides a method for identifying and controlling environmental parameters of a wind farm, comprising:

[0009] Collect several types of data from wind farms to form multi-source heterogeneous data; Preprocessing of multi-source heterogeneous data yields processed multi-source heterogeneous data. A deep neural network is used to build an environmental parameter identification model, and the processed multi-source heterogeneous data is processed to identify environmental parameters and obtain the control parameters of the wind turbine. The control parameters of the wind turbine were optimized using the particle swarm optimization algorithm to obtain the optimized control parameters. Adjust the operating status of the wind turbine based on the optimized control parameters.

[0010] As a further improvement of the present invention, the collection of several types of data from the wind farm to form multi-source heterogeneous data includes: Data from various types of equipment, sensors, meteorological systems, and monitoring technologies in wind farms are collected to obtain meteorological data, internal data of wind turbine units, ground environment and geographic data, equipment health monitoring data, operational data, external environment data, remote sensing and satellite data, historical and forecast data, and external scheduling and market data, forming multi-source heterogeneous data.

[0011] As a further improvement of the present invention, the preprocessing of multi-source heterogeneous data to obtain processed multi-source heterogeneous data includes: Errors and outliers in multi-source heterogeneous data are removed to obtain the removed multi-source heterogeneous data. The removed multi-source heterogeneous data is filled in to obtain the completed multi-source heterogeneous data. The completed multi-source heterogeneous data is normalized to obtain the processed multi-source heterogeneous data.

[0012] As a further improvement of the present invention, the method of using a deep neural network to construct an environmental parameter identification model, and performing environmental parameter identification processing on the processed multi-source heterogeneous data to obtain the control parameters of the wind turbine, including: An environmental parameter identification model is constructed based on a deep neural network using an LSTM network. In the deep neural network-based environmental parameter identification model, the input layer feeds the processed multi-source heterogeneous data to the hidden layer step by step. The LSTM unit in the hidden layer processes the multi-source heterogeneous data at each time step in sequence and passes the final processed sequence context representation to the output layer to generate the control parameters of the wind turbine.

[0013] As a further improvement of the present invention, the step of using a particle swarm optimization algorithm to optimize the control parameters of the wind turbine to obtain optimized control parameters includes: In the control parameters of wind turbine generators, the particle swarm optimization algorithm is used to find the target control parameters of the wind turbine generators to optimize. The fitness of the control parameter is evaluated based on the objective function, and the optimized control parameter is found. The control objective function

[0014] Where P is the power generation capacity of the wind farm. It is the failure risk indicator of the i-th device. It is the first weighting coefficient; It is the second weighting coefficient.

[0015] As a further improvement of the present invention, the step of adjusting the operating state of the wind turbine according to the optimized control parameters includes: Based on the optimized control parameters, a smooth transition trajectory from the current operating parameters to the target optimized parameters is generated, and the real-time setpoint after smoothing is obtained. The operating status of the wind turbine is changed based on the real-time set value after smoothing.

[0016] Secondly, the present invention provides a wind farm environmental parameter identification and control system, comprising: Data acquisition module: used to collect several types of data in the wind farm to form multi-source heterogeneous data; Data processing module: used to preprocess multi-source heterogeneous data to obtain processed multi-source heterogeneous data; Control parameter module: Used to build an environmental parameter identification model using a deep neural network, perform environmental parameter identification processing on the processed multi-source heterogeneous data, and obtain the control parameters of the wind turbine. The optimization parameter module is used to optimize the control parameters of the wind turbine using the particle swarm optimization algorithm to obtain the optimized control parameters. Operation Status Module: Used to adjust the operation status of the wind turbine based on the optimized control parameters.

[0017] Thirdly, the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the wind farm environmental parameter identification and control method.

[0018] Fourthly, the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the wind farm environmental parameter identification and control method.

[0019] Fifthly, this aspect provides a computer program product, including computer instructions that, when executed by a processor, implement the steps of the wind farm environmental parameter identification and control method.

[0020] Compared with the prior art, the present invention has the following beneficial effects: This invention collects various types of data from wind farms to form multi-source heterogeneous data. After preprocessing, it uses an environmental parameter identification model built with a deep neural network for in-depth analysis. This accurately captures the subtle features and complex correlations of various environmental parameters from the multi-source heterogeneous data, improving the identification accuracy of environmental parameters. Secondly, it introduces a particle swarm optimization algorithm to optimize the wind turbine control parameters identified based on the deep neural network. According to the real-time environmental conditions of the wind farm, it quickly finds the optimal combination of control parameters in a vast parameter space. This makes the control parameters no longer static, but dynamically adjustable according to different environmental conditions, resulting in a highly flexible and adaptable control strategy that can quickly and appropriately respond to any complex and changing wind farm environment. Attached Figure Description

[0021] The accompanying drawings described herein are for illustrative purposes only and are not intended to limit the scope of the invention in any way. In the drawings: Figure 1 This is a flowchart illustrating a wind farm environmental parameter identification and control method according to the present invention. Figure 2 This is a schematic diagram of the structure of a wind farm environmental parameter identification and control system according to the present invention; Figure 3 This is a schematic diagram of an electronic device in an embodiment of the present invention. Detailed Implementation

[0022] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. The described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.

[0023] 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 invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0024] Definitions: Multi-source heterogeneous data refers to data from different data sources (such as weather stations, satellite remote sensing, wind turbine sensors, etc.) with different data formats, types, accuracy, and spatiotemporal resolutions.

[0025] Wind farm environmental parameters include meteorological and environmental factors that affect the power generation efficiency and equipment safety of wind farms, such as wind speed, wind direction, temperature, air pressure, humidity, and turbulence intensity.

[0026] Intelligent identification: Utilizing advanced algorithms and models, useful information is automatically extracted from multi-source heterogeneous data to accurately identify the true state and variation patterns of wind farm environmental parameters.

[0027] Adaptive control: The system can automatically adjust the control strategy and parameters based on the real-time changes and identification results of the wind farm's environmental parameters in order to achieve efficient and stable operation of the wind farm.

[0028] Deep Neural Network (DNN): A neural network model containing multiple hidden layers that can automatically learn complex features and patterns in data.

[0029] Long Short-Term Memory (LSTM) network: a special type of recurrent neural network that can effectively handle long-term dependencies in sequential data.

[0030] Particle Swarm Optimization (PSO) is an optimization algorithm based on swarm intelligence that finds the optimal solution by simulating the collective behavior of flocks of birds or schools of fish.

[0031] To address the problems in existing wind farm environmental parameter monitoring and control methods, which are mainly based on single data sources or simple linear models and cannot fully utilize the value of multi-source heterogeneous data, resulting in low accuracy in environmental parameter identification and a lack of flexibility and adaptability in control strategies, this invention provides a wind farm environmental parameter identification and control method, such as... Figure 1 As shown, the method includes: S100: Collects several types of data from wind farms to form multi-source heterogeneous data; S200: Preprocess the multi-source heterogeneous data to obtain the processed multi-source heterogeneous data; S300: Uses a deep neural network to build an environmental parameter identification model, performs environmental parameter identification processing on the processed multi-source heterogeneous data, and obtains the control parameters of the wind turbine. S400: The control parameters of the wind turbine are optimized using the particle swarm optimization algorithm to obtain the optimized control parameters; S500: Adjusts the operating status of the wind turbine unit based on the optimized control parameters.

[0032] This method improves the identification of environmental parameters by identifying multi-source heterogeneous data in the wind farm environment, enabling the control strategy to adapt to various situations.

[0033] The present invention will be further explained and described below with reference to specific solutions.

[0034] This invention provides a method for identifying and controlling environmental parameters in wind farms, specifically including the following steps: S1: Data Acquisition In a wind farm environment, data is collected from different types of equipment, sensors, meteorological systems, and monitoring technologies, resulting in multi-source heterogeneous data. Therefore, multi-source heterogeneous data comes from various sources and has different formats, structures, and characteristics.

[0035] Specifically, multi-source heterogeneous data includes meteorological data, wind turbine internal data, ground environment and geographic data, equipment health monitoring data, operational data, external environment data, remote sensing and satellite data, historical and forecast data, and external scheduling and market data.

[0036] The meteorological data includes: wind speed and direction, temperature, air pressure, humidity, precipitation, and radiation. The internal data of the wind turbine includes: turbine speed, pitch angle, yaw angle, power generation, vibration and temperature sensor data, as well as current, voltage and power factor; Ground environment and geographic data include: topographic data, soil moisture, and geographic location data; Equipment health monitoring data includes: mechanical health monitoring data, as well as fault and alarm data; Operational data includes: wind farm operation data and maintenance and service records; External environmental data includes: power grid data and environmental pollution data; Remote sensing and satellite data include: satellite remote sensing data and UAV data; Satellite remote sensing data: Information such as wind field distribution, wind speed changes, and cloud cover over a wide area is obtained through satellite imagery, providing support for the overall layout and optimization of wind farms; Drone data: Drones can provide real-time images of wind farm areas, wind conditions, terrain changes and other data, which can help detect the installation location or maintenance area of ​​wind turbines; In summary, the multi-source heterogeneous data from wind farms includes not only basic information from meteorology, equipment, and geography, but also multi-dimensional health monitoring, external environmental impacts, and operational and grid dispatch data. The collection, fusion, and analysis of this data are crucial for optimizing wind farm operating efficiency, ensuring equipment safety, and improving energy output.

[0037] S2: Multi-source heterogeneous data preprocessing Data from different data sources is examined to remove errors, missing values, and outliers. Missing values ​​can be handled using methods such as linear interpolation and mean imputation. Data of different types and ranges are normalized to the same scale. Weighted averaging and principal component analysis (PCA) are used to fuse data from different data sources to obtain more comprehensive and accurate environmental information.

[0038] S21: Error Value Detection and Removal Errors in the data are usually caused by faults in the data acquisition process or input errors. Data validation rules are used to detect these errors. For example: Range check: Set reasonable maximum and minimum values; data outside the range can be considered incorrect.

[0039] Consistency checks: For example, the order of timestamps and data should be consistent. If the data order is abnormal, it may need to be corrected.

[0040] Rule checks: Some data may have logical relationships, such as a variable that cannot be negative, or certain variables that have specific mathematical relationships. Data that violates these rules should be removed.

[0041] S22: Missing Value Handling Missing values ​​are a common problem in data analysis. Simply ignoring missing values ​​can lead to incomplete data or biased analysis results; therefore, it is necessary to handle missing values. Common methods for handling missing values ​​include: Deletion method: When the proportion of missing values ​​is small and does not affect the analysis, rows or columns containing missing values ​​can be directly deleted.

[0042] Interpolation methods: For missing numerical data, linear interpolation can be used (linear interpolation assumes that the missing value is located between two adjacent known data points) to fill in the gaps. In addition to linear interpolation, polynomial interpolation, spline interpolation, and other methods can also be used.

[0043] Mean / Median Fill: For missing values, you can fill them with the mean or median of the column. This is especially suitable when the data is relatively evenly distributed and there are few missing data points.

[0044] Model-based imputation: For example, using the K-Nearest Neighbors (KNN) algorithm, regression models, etc., to predict missing values ​​and impute them.

[0045] S23: Outlier Detection and Handling Outliers can significantly impact data analysis and modeling results, therefore they need to be detected and handled. Common outlier detection methods include: Statistical methods: For example, detecting outliers based on boxplots. Data points outside the box are usually considered outliers.

[0046] Based on Z-value (standard deviation): If the Z-value of a data point exceeds a certain threshold (such as 3), the point is considered an outlier.

[0047] Based on IQR (interquartile range): Outliers are defined as values ​​below Q1-1.5×IQR or above Q3+1.5×IQR by calculating the IQR of the data.

[0048] Detected outliers can be handled in the following ways: Deletion: Extreme outliers can be deleted directly.

[0049] Replacement: Replace outliers with the mean or median, or fill them with interpolation.

[0050] Transformation: For example, performing logarithmic transformations or other changes on the data to mitigate the impact of outliers.

[0051] S24. Normalization and Standardization Since different data sources have different scales and units, normalization and standardization are ways to unify the scale of data.

[0052] S25: Data Fusion Data fusion is the process of combining data from different data sources to obtain more comprehensive and accurate information. The following methods can be used when performing data fusion: Weighted average method: Based on the importance of different data sources, different weights are assigned to each data source, and the values ​​of multiple data sources are averaged according to their weights.

[0053] Principal Component Analysis (PCA): PCA is a dimensionality reduction technique that transforms high-dimensional data into lower-dimensional features while retaining most of the information. PCA maps data from different data sources to a new feature space, reducing redundant information and improving data processing efficiency.

[0054] Kalman filtering and Bayesian methods: These methods can fuse multi-source data in dynamic environments and are suitable for time-series data or noisy sensor data. They can automatically adjust the weights of data sources based on the correlation and uncertainty of the data to achieve optimal data fusion.

[0055] By following the steps above, we can ensure that the data obtained from different data sources meets the quality standards and provides more accurate and comprehensive environmental information during the fusion process.

[0056] S3: Intelligent identification of environmental parameters Use an LSTM network to build a deep neural network to construct an environmental parameter identification model.

[0057] The input layer of an LSTM network receives processed heterogeneous data from multiple sources, the hidden layer processes sequence data through LSTM units, and the output layer outputs the identified environmental parameter values.

[0058] The input layer feeds the processed sequence data to the hidden layer step by step. The LSTM units in the hidden layer process each time step in turn, and cyclically pass their internal state, hidden state and cell state to remember historical information. Finally, the processed sequence context representation is passed to the output layer to be mapped to specific environmental parameter values.

[0059] The LSTM model is trained using historical data, with mean squared error (MSE) as the loss function. The model's weights and biases are updated using backpropagation. During training, stochastic gradient descent (SGD) or its variants (such as the Adam optimizer) are used to optimize the model parameters.

[0060] Specifically: The input layer receives preprocessed multi-source heterogeneous data. Typically, the input to an LSTM network is a three-dimensional tensor with the shape (number of samples, time step, number of features). For each time step, the network receives the corresponding feature data.

[0061] The hidden layer uses LSTM units to process time series data. LSTM networks can capture long-term dependencies in time series data, which helps to identify dynamic changes in environmental parameters.

[0062] Multiple LSTM layers can be stacked (stacked LSTM) to enhance the model's learning ability.

[0063] The fully connected output layer outputs the identified environmental parameter values ​​(such as temperature and humidity).

[0064] Using mean squared error (MSE) as the loss function can quantify the difference between the model's predicted values ​​and the true values.

[0065] The gradient is calculated using the backpropagation algorithm, the model parameters are updated, and the value of the loss function is gradually reduced.

[0066] The training process is as follows: The model is trained using the training dataset. During each training iteration, the prediction results are calculated via forward propagation, the loss is calculated, and the model parameters are adjusted via backpropagation.

[0067] During training, a validation set is used to evaluate the model's performance and adjust model hyperparameters, such as the number of LSTM units and the learning rate, to prevent overfitting.

[0068] The trained model is evaluated using a test set, the model's accuracy is calculated, and the difference between the predicted results and the true values ​​is analyzed.

[0069] Adjust the hyperparameters of the LSTM network (such as the number of LSTM layers, the number of units per layer, and the learning rate) based on the model's performance on the validation set.

[0070] Use methods such as Dropout or L2 regularization to prevent overfitting and ensure the model's generalization ability.

[0071] To prevent overfitting caused by excessively long training periods, an early stopping mechanism is implemented during the training process.

[0072] Use the trained model for inference: predict new input data and identify environmental parameters. Calculate the model's prediction error and evaluate its performance in real-world applications.

[0073] S4: Use the Particle Swarm Optimization (PSO) algorithm to optimize the control strategy and obtain the optimized control parameters.

[0074] The control parameters of the wind turbines to be optimized are defined as the positions of particles in the search space, with defined range limits. A swarm of particles is then initialized, and each particle is randomly assigned an initial position and velocity. During iteration, for each particle's position (i.e., a set of control parameters), a wind farm model is run to calculate its corresponding power generation and equipment failure risk, and its fitness is evaluated based on a pre-defined weighted objective function. Each particle records its historical best position, and the entire swarm tracks the global best position. In each iteration, particles update their velocity and position by integrating individual cognition and collective experience, thus exploring new solutions in the parameter space. This process is repeated until the maximum number of iterations is reached or the quality of the solution converges to a satisfactory level. Finally, the global best position is output as the optimized set of control parameters, thereby maximizing the wind farm's power generation efficiency while ensuring equipment safety.

[0075] With the goals of maximizing the power generation efficiency and ensuring equipment safety of the wind farm, the control objective function is defined as follows:

[0076] Where P is the power generation capacity of the wind farm. It is the failure risk indicator of the i-th device. It is the first weighting coefficient; It is the second weighting coefficient.

[0077] Particle swarm optimization (PSO) algorithm is used to optimize the control parameters of wind turbines, such as pitch angle and yaw angle.

[0078] S5: Adjust the operating status of the wind turbine based on the optimized control parameters. Specifically, adjusting the operating status of wind turbine units includes: blade angle (yaw angle) adjustment, pitch control, speed control, power output adjustment, fault warning and adaptive adjustment, load balancing and wind turbine array scheduling.

[0079] S51: Parameter Reception and Verification The optimized control parameter set, such as the optimal pitch angle, yaw angle, and generator torque setpoint, is received from the PSO optimization algorithm output. First, the validity of these parameters is verified in the control system to ensure they are within the allowable range for safe operation of the wind turbine.

[0080] S52: Smooth transition of setpoint To avoid mechanical stress shocks to the equipment caused by sudden changes in control commands, new parameters are not applied to the unit instantaneously. Instead, the control system generates a smooth transition trajectory from the current operating parameters to the target optimized parameters, such as a ramp change, to achieve a smooth and gradual switching of operating states.

[0081] S53: Actuator Drive The smoothed real-time setpoints are then sent to the corresponding actuators. For example, the pitch angle setpoint is sent to the pitch control system, the yaw angle setpoint is sent to the yaw drive system, and the torque setpoint is sent to the frequency converter, thereby physically changing the operating state of the wind turbine.

[0082] S54: Real-time Status Monitoring and Security Verification During the adjustment process and after reaching a new steady state, key performance and safety indicators such as power generation, component load, vibration, and temperature are continuously monitored. This real-time data is compared with the fault risk model in the optimization target for online safety verification. If any indicator exceeds the safety threshold, protection logic is immediately triggered, the optimization strategy is terminated, and the system reverts to a safe mode.

[0083] S55: Performance Evaluation and Feedback After operating in the new steady state for a period of time, the actual effect of this parameter adjustment will be evaluated. The core task is to calculate the actual improvement in power generation efficiency and comprehensively weigh it with safety indicators such as equipment fatigue loss, thereby verifying the achievement of the PSO optimization objectives in actual operation.

[0084] S56: Closed-loop adaptive optimization The performance data and corresponding control parameters collected during actual operation are stored as new empirical data in the historical database. This data can be used to periodically restart the PSO optimization process, providing the optimization algorithm with training data that is closer to actual operating conditions. This enables the control system to have self-learning and adaptive capabilities, track changes in wind resources and equipment performance, and achieve continuous optimization.

[0085] By adjusting the operating status of wind turbines through optimized control parameters, the power generation efficiency of wind turbines can be greatly improved, operating costs reduced, equipment lifespan extended, and the stable operation of wind power systems under different environmental conditions can be ensured.

[0086] In summary, this method mainly includes three steps: multi-source heterogeneous data preprocessing, intelligent identification of environmental parameters, and adaptive control. By cleaning, fusing, and extracting features from the multi-source heterogeneous data, deep neural networks are used for intelligent identification of environmental parameters. Finally, based on the identification results, particle swarm optimization algorithm is used to adjust the adaptive control strategy.

[0087] The second objective of this invention is to provide a wind farm environmental parameter identification and control system, such as... Figure 2 As shown, it includes: Data Acquisition Module 100: Used to collect several types of data in the wind farm to form multi-source heterogeneous data; Data processing module 200: used to preprocess multi-source heterogeneous data to obtain processed multi-source heterogeneous data; Control parameter module 300: Used to build an environmental parameter identification model using a deep neural network, perform environmental parameter identification processing on the processed multi-source heterogeneous data, and obtain the control parameters of the wind turbine. Optimization parameter module 400: Used to optimize the control parameters of the wind turbine using the particle swarm optimization algorithm to obtain optimized control parameters; Operation Status Module 500: Used to adjust the operation status of the wind turbine based on the optimized control parameters.

[0088] like Figure 3 As shown, a third objective of this invention is to provide an electronic device comprising a processor 601, a memory 602, and a display screen 603. The memory 602 and the display screen 603 are both connected to the processor 601, such as via a bus 604. Optionally, the electronic device may further include a transceiver 605. It should be noted that in practical applications, the transceiver 605 is not limited to one type, and the structure of this electronic device does not constitute a limitation on the embodiments of this application.

[0089] Processor 601 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 601 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0090] Bus 604 may include a pathway for transmitting information between the aforementioned components. Bus 604 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. Bus 604 can be divided into address bus, data bus, control bus, etc.

[0091] The memory 602 may be a ROM (Read Only Memory) or other type of static storage device capable of storing static information and instructions, RAM (Random Access Memory) or other type of dynamic storage device capable of storing information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), CD-ROM (Compact Disc Read Only Memory) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto.

[0092] The memory 602 is used to store application code that executes the solution of this application, and its execution is controlled by the processor 601. The processor 601 is used to execute the application code stored in the memory 602 to implement the content shown in the foregoing method embodiments.

[0093] Figure 3 The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0094] A fourth objective of this invention is to provide a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, performs the aforementioned functions. Figure 1 The illustrated method embodiments include various processes. For example, a memory may include instructions that can be executed by a processor of an electronic device to perform the described method.

[0095] A computer-readable storage medium can be a tangible device that holds and stores instructions used by an instruction execution device. A computer-readable storage medium can be, but is not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any combination thereof. Specifically, a computer-readable storage medium can be a portable computer disk, a hard disk, a USB flash drive, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), staging random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory stick, floppy disk, optical disk, magnetic disk, mechanical encoding device, or any combination thereof.

[0096] A fifth objective of this invention is to provide a computer program product comprising computer instructions that, when executed by a processor, implement the above-described... Figure 1 The various processes of the method embodiments shown can achieve the same technical effect, and will not be described again here to avoid repetition.

[0097] Many embodiments and applications beyond the examples provided will be apparent to those skilled in the art upon reading the foregoing description. Therefore, the scope of this teaching should not be determined by reference to the foregoing description, but rather by reference to the foregoing claims and the full scope of their equivalents. For purposes of completeness, all articles and references, including patent applications and publications, are incorporated herein by reference. The omission of any aspect of the subject matter disclosed herein in the foregoing claims is not intended as a waiver of that subject matter, nor should it be construed as an indication that the applicant has not considered that subject matter as part of the disclosed inventive subject matter.

[0098] The above content provides a further detailed description of the present invention. It should not be construed that the specific embodiments of the present invention are limited to this. For those skilled in the art, several simple deductions or substitutions can be made without departing from the concept of the present invention, and all such deductions or substitutions should be considered to fall within the scope of protection of the present invention as defined by the submitted claims.

Claims

1. A wind farm environment parameter identification and control method, characterized in that, The method comprises the following steps: Collecting various types of data in a wind farm to form multi-source heterogeneous data; Preprocessing the multi-source heterogeneous data to obtain processed multi-source heterogeneous data; Using a deep neural network to build an environmental parameter recognition model to perform environmental parameter recognition processing on the processed multi-source heterogeneous data to obtain control parameters of a wind turbine; Using a particle swarm optimization algorithm to optimize the obtained control parameters of the wind turbine to obtain optimized control parameters; Adjusting the operating state of the wind turbine according to the optimized control parameters.

2. A wind farm environmental parameter identification and control method according to claim 1, characterized in that, The collection of various types of data in the wind farm to form multi-source heterogeneous data comprises: Collecting data of various types of equipment, sensors, meteorological systems and monitoring technologies in the wind farm to obtain meteorological data, wind turbine internal data, ground environment and geographical data, equipment health monitoring data, operation data, external environment data, remote sensing and satellite data, historical and predicted data, and external scheduling and market data, thereby forming multi-source heterogeneous data.

3. The method according to claim 1, wherein, The preprocessing of the multi-source heterogeneous data to obtain processed multi-source heterogeneous data comprises: Removing error data and outliers in the multi-source heterogeneous data to obtain removed multi-source heterogeneous data; Filling and completing the removed multi-source heterogeneous data to obtain completed multi-source heterogeneous data; Normalizing the completed multi-source heterogeneous data to obtain processed multi-source heterogeneous data.

4. The wind farm environmental parameter identification and control method of claim 1, wherein, The use of a deep neural network to build an environmental parameter recognition model to perform environmental parameter recognition processing on the processed multi-source heterogeneous data to obtain control parameters of a wind turbine comprises: Building a deep neural network to build an environmental parameter recognition model based on an LSTM network; The input layer of the deep neural network to build an environmental parameter recognition model feeds the processed multi-source heterogeneous data to the hidden layer in time steps, and the LSTM unit of the hidden layer processes each time step of the multi-source heterogeneous data in turn and transmits the final processed sequence context representation to the output layer to generate control parameters of a wind turbine.

5. The method of claim 1, wherein, The use of a particle swarm optimization algorithm to optimize the obtained control parameters of the wind turbine to obtain optimized control parameters comprises: Finding the optimized target wind turbine control parameters in the control parameters of the wind turbine through a particle swarm optimization algorithm; Evaluating the fitness according to the control target function to find the optimized control parameters. The control target function wherein P is the power generated by the wind farm, is the failure risk indicator of the i-th device, is the first weight coefficient; is the second weight coefficient.

6. The wind farm environmental parameter identification and control method of claim 1, wherein, The adjustment of the operating state of the wind turbine according to the optimized control parameters comprises: Generating a smooth transition trajectory from the current operating parameters to the target optimized parameters according to the optimized control parameters to obtain smoothed real-time set values; Changing the operating state of the wind turbine according to the smoothed real-time set values.

7. A wind farm environment parameter identification and control system, characterized in that, The method comprises the following steps: Data collection module: for collecting various types of data in a wind farm to form multi-source heterogeneous data; Data processing module: for preprocessing the multi-source heterogeneous data to obtain processed multi-source heterogeneous data; Control parameter module: for using a deep neural network to build an environmental parameter recognition model to perform environmental parameter recognition processing on the processed multi-source heterogeneous data to obtain control parameters of a wind turbine; The optimization parameter module is configured to optimize the obtained control parameters of the wind turbine generator using a particle swarm optimization algorithm to obtain optimized control parameters. The running state module is configured to adjust the running state of the wind turbine generator according to the optimized control parameters.

8. An electronic device, comprising: The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the steps of the wind farm environment parameter identification and control method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the steps of the wind farm environment parameter identification and control method according to any one of claims 1-6.

10. A computer program product, characterised in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the steps of the wind farm environment parameter identification and control method according to any one of claims 1-6.

Images