Methods and devices for analyzing, diagnosing and optimizing the power generation capacity of wind turbine units

Abstract

This invention discloses a method and apparatus for analyzing, diagnosing, and optimizing the power generation capacity of wind turbine generators, relating to the field of wind turbine generator capacity monitoring. The method includes: acquiring multiple sets of data to be processed, including wind turbine operation data, meteorological data, and wind farm operation data; preprocessing the data and inputting it into multiple analysis models, including a power curve fitting analysis model, a wind deviation analysis model, a temperature-power analysis model, an operating time analysis model, a power loss analysis model, a pitch angle analysis model, a micro-situation analysis model, and / or an AGC curve deviation analysis model, to obtain multiple sets of analytical data; inputting the multiple sets of analytical data into a diagnostic model based on machine learning to obtain diagnostic results; and inputting the diagnostic results into an expert optimization suggestion model to obtain optimized suggestion schemes. This invention can improve the power generation efficiency of wind farms, enhance system efficiency and accuracy, reduce operating costs, and facilitate the digital transformation of the wind power industry.

Description

Technical Field

[0001] This invention belongs to the technical field of wind turbine power generation capacity monitoring, and specifically relates to a method and device for analyzing, diagnosing and optimizing wind turbine power generation capacity. Background Technology

[0002] Wind power is an important clean energy source with enormous development potential. However, due to factors such as wind resources, turbine performance, and operation and maintenance, the power generation capacity of wind turbines varies significantly, making it difficult to fully realize their power generation potential. Existing wind farm management systems mainly rely on experience-based judgment and manual analysis. However, manual analysis involves large amounts of data and is time-consuming, making it difficult to promptly identify and resolve potential problems affecting wind turbine power generation capacity. Experience-based judgment is highly subjective, lacks scientific basis, and is difficult to accurately assess wind turbine power generation capacity. Moreover, existing management systems cannot monitor the operating status of wind turbines in real time, making it difficult to promptly detect and handle abnormal situations. Therefore, existing wind farm management systems result in low wind farm power generation efficiency, high operating costs, and difficulties in achieving sustainable development. Summary of the Invention

[0003] To address the aforementioned issues, this invention proposes a method and apparatus for analyzing, diagnosing, and optimizing the power generation capacity of wind turbine generators. This method can improve the power generation efficiency of wind farms, enhance system efficiency and accuracy, reduce operating costs, and facilitate the digital transformation of the wind power industry.

[0004] In a first aspect, the present invention provides a method for analyzing, diagnosing, and optimizing the power generation capacity of wind turbine generators, comprising:

[0005] Acquire multiple sets of data to be processed, including wind turbine operation data, meteorological data, and / or wind farm operation data;

[0006] The multiple sets of data to be processed are preprocessed to obtain multiple sets of preprocessed data;

[0007] The multiple sets of preprocessed data are input into the corresponding multiple analysis models to obtain multiple sets of analysis data; the multiple analysis models include power curve fitting analysis model, wind deviation analysis model, temperature-power analysis model, operating time analysis model, power loss analysis model, pitch angle analysis model, micro-situation analysis model and / or AGC curve deviation analysis model;

[0008] The multiple sets of analytical data are input into the trained diagnostic model to obtain diagnostic results; the diagnostic model is built based on a machine learning model.

[0009] The diagnostic results are input into a pre-built expert optimization suggestion model to obtain an optimization suggestion scheme.

[0010] In an optional implementation, the wind turbine operating data includes one or more combinations of wind turbine power, wind turbine wind speed, wind turbine wind direction, wind turbine temperature, wind turbine pitch angle, wind turbine windward angle, wind turbine speed, and wind turbine status code; the meteorological data includes one or more combinations of ambient wind speed, ambient wind direction, ambient temperature, ambient air pressure, and ambient humidity; and the wind farm operating data includes one or more combinations of wind turbine grid connection status, wind turbine power curtailment status, and maintenance records.

[0011] In an optional implementation, the power curve fitting analysis model includes:

[0012] The wind turbine power prediction module is used to predict the wind turbine power under different environmental wind speeds based on the preprocessed data and the pre-built power curve sub-model.

[0013] The deviation rate prediction module is used to determine the power curve deviation rate based on the actual power under different environmental wind speeds, the rated power of the wind turbine, and the theoretical output power of the wind turbine.

[0014] The update module is used to update the power curve sub-model based on the power curve deviation rate;

[0015] The power curve sub-model is constructed using the following method:

[0016] Multiple machine learning algorithms are used to perform curve fitting on the historical data of the wind turbine to obtain multiple primary models of the wind turbine's power curves; the machine learning algorithms include support vector machines, neural networks, and / or random forests.

[0017] The fitting effect of the multiple primary power curve models is evaluated using evaluation metrics. Based on the fitting effect, the primary power curve model with the best fitting effect is determined from the multiple primary power curve models, and this model is taken as the power curve sub-model. The evaluation metrics include the R-squared value and / or the root mean square error.

[0018] In an optional implementation, the wind deviation analysis model includes:

[0019] The yaw error data module is used to obtain static yaw error data based on the wind turbine wind speed preprocessing data, wind turbine wind angle preprocessing data and the pre-built wind deviation sub-model in the preprocessed data, wherein the wind deviation sub-model is constructed based on the support vector regression algorithm.

[0020] An error correction module is used to adjust the yaw control strategy of the wind turbine based on the static yaw error data in order to optimize wind energy utilization efficiency.

[0021] The yaw error compensation module is used to compensate the yaw error data into the power curve sub-model to improve the accuracy of the power curve sub-model.

[0022] In an optional implementation, the temperature-power analysis model includes:

[0023] The power generation capacity prediction module is used to determine the influence coefficient of temperature on power based on the environmental temperature preprocessing data, wind turbine power preprocessing data and the pre-built temperature-power relationship sub-model in the preprocessing data, and to predict the wind turbine power at different temperatures based on the influence coefficient of temperature on power.

[0024] A temperature compensation module is used to adjust the power curve sub-model according to the influence coefficient of temperature on power, so as to improve the accuracy of the power curve sub-model;

[0025] The temperature-power relationship sub-model is determined using linear regression, multinomial regression, or neural network models based on the maximum power variation of wind turbines under different ambient temperatures.

[0026] In an optional implementation, the power loss analysis model includes:

[0027] The loss type module is used to determine the loss amount corresponding to various loss types based on the preprocessed data and the pre-built loss amount submodule; wherein, the various loss types include one or more of maintenance, fault, power restriction, performance, and damage.

[0028] The loss ratio module is used to determine the power loss ratio relative to each loss type based on the power loss.

[0029] The loss cause module is used to determine the loss cause corresponding to each type of loss;

[0030] The power loss submodule is constructed using the following method:

[0031] The wind turbine power generation should be determined for multiple time periods based on the wind turbine operation data and the meteorological data;

[0032] The expected power generation of wind turbines in multiple time periods is compared with the actual power generation in the corresponding time periods to determine the power loss in each time period.

[0033] Determine the type of loss corresponding to the lost electricity in each time period.

[0034] In an optional implementation, the micro-site selection analysis model is used to acquire meteorological preprocessed data from multiple sets of preprocessed data at the target location, input the meteorological preprocessed data of the target location into a pre-constructed wind field model, and output wind resource distribution data; wherein, the wind field model is constructed based on computational fluid dynamics software and combined with the terrain data of the target location.

[0035] In an optional implementation, the AGC curve deviation analysis model includes:

[0036] The deviation assessment module is used to determine the AGC curve deviation rate of the wind turbine based on the pre-processed wind turbine power and the pre-built AGC model; wherein the AGC model is constructed based on time series analysis methods.

[0037] The control optimization module is used to adjust the wind turbine operation strategy according to the AGC curve deviation rate of the wind turbine. The wind turbine operation strategy includes a shutdown strategy, a startup strategy, and a power control strategy.

[0038] The AGC model correction module is used to correct the AGC model based on the deviation rate of the AGC curve of the wind turbine.

[0039] In an optional implementation, the optimization suggestions include adjusting the wind turbine control parameters, optimizing the wind turbine operation strategy, and developing a maintenance plan;

[0040] The adjustment of wind turbine control parameters includes adjusting yaw angle, pitch angle and / or speed;

[0041] Optimizing wind turbine operation strategies includes shutdown strategies and / or startup strategies;

[0042] Developing a maintenance plan includes regular inspections and / or troubleshooting.

[0043] Secondly, the present invention provides a wind turbine generator power generation capacity analysis, diagnosis, and optimization device, comprising:

[0044] The acquisition module is used to acquire multiple sets of data to be processed, including wind turbine operation data, meteorological data and / or wind farm operation data.

[0045] The preprocessing module is used to preprocess the multiple sets of data to be processed to obtain multiple sets of preprocessed data;

[0046] The analysis module is used to input the multiple sets of preprocessed data into the corresponding multiple analysis models to obtain multiple sets of analysis data; the multiple analysis models include a power curve fitting analysis model, a wind deviation analysis model, a temperature-power analysis model, an operating time analysis model, a power loss analysis model, a pitch angle analysis model, a micro-situation analysis model, and / or an AGC curve deviation analysis model;

[0047] The diagnostic module is used to input the multiple sets of analytical data into a trained diagnostic model to obtain diagnostic results; the diagnostic model is built based on a machine learning model.

[0048] The optimization suggestion module is used to input the diagnostic results into a pre-built expert optimization suggestion model to obtain an optimization suggestion scheme.

[0049] The beneficial effects of the technical solution provided by the embodiments of the present invention are as follows: The present invention can fuse and analyze multi-source data, including wind turbine operation data, meteorological data, and wind farm operation data, and perform fault diagnosis by combining a diagnostic model constructed with a machine learning model. The diagnostic results are then input into an expert optimization suggestion model to obtain optimized suggestion solutions. The present invention can improve the power generation efficiency of wind turbines and increase annual power generation; it can reduce operating costs and improve economic benefits; it utilizes machine learning algorithms to improve the level of intelligence and intelligent diagnosis and optimization suggestions, thereby improving the level of intelligence of wind farms; and it promotes the digital transformation of the wind power industry and provides technical support for the sustainable development of the new energy industry. Attached Figure Description

[0050] Figure 1 A flowchart illustrating a method for analyzing, diagnosing, and optimizing the power generation capacity of a wind turbine generator set, provided in an embodiment of the present invention;

[0051] Figure 2 The present invention provides a system schematic diagram of a wind turbine power generation capacity analysis, diagnosis and optimization device.

[0052] In the diagram: 100 - Acquisition module; 200 - Preprocessing module; 300 - Analysis module; 400 - Diagnostic module; 500 - Optimization suggestion module. Detailed Implementation

[0053] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0054] See Figure 1 This embodiment provides a method for analyzing, diagnosing, and optimizing the power generation capacity of wind turbine generators, including the following steps S100 to S500.

[0055] Step S100: Obtain multiple sets of data to be processed, including wind turbine operating data, meteorological data, and / or wind farm operating data. Specifically, wind turbine operating data includes turbine power, wind speed, wind direction, temperature, pitch angle, speed, and status code; meteorological data includes ambient wind speed, wind direction, temperature, air pressure, and humidity; wind farm operating data includes wind turbine grid connection status, power curtailment status, and maintenance records. It should be noted that, as needed, the data to be processed may also include other data.

[0056] Devices that acquire data to be processed include one or more of the following: data acquisition devices, SCADA (Supervisory Control and Data Acquisition) systems, and cloud platforms. For example, wind turbine operation data is acquired using data acquisition devices, while some meteorological and wind farm operation data is acquired using SCADA (wind turbine operation data can also be acquired using this system). Typically, data acquired by data acquisition devices is sent to the SCADA system, while cloud platforms are usually used to integrate data from multiple sources, which may include data from the SCADA system, additional meteorological data (such as forecast data from meteorological services and satellite or radar meteorological data), more detailed wind farm operation data (such as complete maintenance records and historical performance data), power grid data (such as regional electricity demand and electricity price information), and other external data sources (such as environmental monitoring data and bird activity data).

[0057] It's important to note that wind turbine operating data differs from meteorological data. Wind turbine operating data is typically obtained directly from sensors installed on the turbine, offering high measurement accuracy and frequency (e.g., per second or per minute). It reflects the local conditions at a single turbine location and is primarily used for monitoring and analyzing the performance of individual turbines. Meteorological data, on the other hand, usually comes from weather stations or meteorological towers, which may be located within or near the wind farm. Its accuracy may be lower, and the measurement intervals may be longer (e.g., every 10 minutes or hour). It can represent average meteorological conditions over a larger area and is commonly used for wind resource assessment and power generation forecasting for the entire wind farm. Status codes are identifiers for the operating status of wind turbines, indicating their current operating state or fault condition. They are typically represented by numbers or codes; for example, "0" indicates normal operation, "1" indicates standby, "2" indicates startup, "3" indicates shutdown, "4" indicates emergency shutdown, "5" indicates scheduled maintenance, and "10" to "99" represent various specific faults (such as yaw fault, pitch fault, overspeed fault, etc.). The grid-connected status of wind turbines is either grid-connected or not grid-connected. Grid-connected refers to the connection of the wind turbine to the power grid. Power curtailment refers to situations where the power output is restricted due to reasons such as grid load limitations, wind farm output limitations, wind farm safety controls, and policy-related power curtailment.

[0058] This embodiment integrates and analyzes multi-source data such as wind turbine operation data, meteorological data, and wind farm operation data, thereby improving the accuracy and reliability of the analysis results.

[0059] Step S200: Preprocess the multiple sets of data to be processed to obtain multiple sets of preprocessed data.

[0060] The preprocessing methods in this embodiment include data cleaning (removing erroneous, missing, and abnormal data), data transformation (unifying different types of data into a processable format), data normalization (scaling data to a uniform range to improve data consistency and comparability, facilitating model training and analysis), and data interpolation (using interpolation algorithms to fill in missing data and ensure data integrity).

[0061] Step S300: Input multiple sets of preprocessed data into the corresponding multiple analysis models to obtain multiple sets of analysis data; the multiple analysis models include power curve fitting analysis model, wind deviation analysis model, temperature-power analysis model, running time analysis model, power loss analysis model, pitch angle analysis model, micro-site selection analysis model and AGC curve deviation analysis model, as detailed below (1)-(8).

[0062] Power curve fitting analysis model

[0063] The power curve fitting analysis model includes a wind turbine power prediction module, a wind turbine power prediction module, and an update module.

[0064] The wind turbine power prediction module predicts the corresponding wind turbine power under different environmental wind speeds based on preprocessed data and a pre-built power curve sub-model. The preprocessed data input to the power curve sub-model can be wind speed preprocessed data, temperature preprocessed data, or multiple sets of preprocessed data such as wind speed, temperature, wind deviation, and blade pitch angle can be integrated into a single power curve sub-model to establish a more comprehensive power model. The deviation rate prediction module determines the power curve deviation rate based on the actual wind turbine power, the rated power of the wind turbine, and the theoretical output power under different environmental wind speeds. The update module dynamically updates the power curve sub-model based on the power curve deviation rate to reflect changes in wind turbine performance in a timely manner.

[0065] Here, the deviation rate is determined by the following formula (1).

[0066]

[0067] In equation (1), P hi P represents the actual power of the wind turbine at time i. ri This indicates that the wind speed at time i corresponds to the power value on the standard power curve, P. dThe value represents the design power of the wind turbine, and n represents the number of sampling points.

[0068] The power curve sub-model is constructed through the following steps (1-1) to (1-2).

[0069] Step (1-1) utilizes various machine learning algorithms to perform curve fitting on historical data of wind turbines, obtaining multiple primary power curve models for the wind turbines. These machine learning algorithms include Support Vector Machines (SVMs), Neural Networks (NNs), and / or Random Forests. SVMs excel at handling high-dimensional data and have good robustness to small sample data, making them suitable for classification and regression problems. In this embodiment, SVMs can be used to fit the power data of wind turbines under different wind speeds. Neural Networks can learn complex nonlinear relationships and are suitable for handling large amounts of data, applicable to classification, regression, and prediction problems. In this embodiment, NNs can be used to fit the power data of wind turbines under different wind speeds, temperatures, and wind deviations to establish a more refined power model. Random Forests are an ensemble learning algorithm that improves the prediction accuracy of the model by combining multiple decision trees, making them suitable for classification and regression problems. In this embodiment, Random Forests can be used to fit the power data of wind turbines under different wind speeds, temperatures, and wind deviations to establish a more stable power model.

[0070] Step (1-2) uses evaluation metrics to assess the fitting effect of multiple primary power curve models. Based on the fitting effect, the primary power curve model with the best fitting effect is determined from the multiple primary power curve models, and the primary power curve model with the best fitting effect is taken as the power curve sub-model. The evaluation metrics include the R-squared value and the root mean square error.

[0071] It should be noted that in practical applications, the required operating data should be selected to build the power curve sub-model as needed. For example, only wind speed and temperature data can be used to build the power curve sub-model, or different power curve sub-models can be built according to different application scenarios.

[0072] The power curve fitting analysis model in this embodiment uses machine learning algorithms to fit historical data, establishing a power curve sub-model for the wind turbine. Evaluation metrics such as R-squared and root mean square error are used to assess the fitting effect, selecting the best-fit model. Based on this best-fit model, the power curve deviation rate is calculated, and the deviation rate is used to update the power curve sub-model. An update module is also included to dynamically update the power curve sub-model, facilitating timely reflection of changes in wind turbine performance and improving the real-time performance and accuracy of the analysis results.

[0073] (1) Wind deviation analysis model

[0074] The wind deviation analysis model includes a yaw error data module, an error correction module, and a yaw error compensation module. The yaw error data module is used to obtain static yaw error data based on the wind turbine direction preprocessing data, wind angle preprocessing data, and the pre-built wind deviation sub-model in the preprocessed data. The wind deviation sub-model is constructed based on the support vector regression algorithm.

[0075] Specifically, to maximize wind energy utilization, wind turbines need to maintain an optimal windward angle at all times, meaning the turbine blades are parallel to the wind direction. The yaw system, which adjusts the turbine's direction, has static yaw error, including errors in the wind direction sensor, controller errors, actuator errors, and installation errors. Static yaw error affects power generation efficiency, causes blade fatigue, and increases the risk of equipment failure. The windward deviation sub-model calculates the windward deviation using the turbine's wind direction and windward angle, obtaining power output data under different windward deviation angles. The yaw error data module identifies the static yaw error based on the power output data output by the windward deviation sub-model. It should be noted that the power output data refers to the average power output of the wind turbine over a specific time period.

[0076] The error correction module is used to adjust the yaw control strategy of the wind turbine based on static yaw error data to optimize wind energy utilization efficiency. The yaw control strategy adjusts the yaw angle of the wind turbine to ensure it maintains optimal wind alignment, thereby improving power generation efficiency and reducing blade fatigue risk. Yaw control strategies include fixed-point yaw control, wind direction tracking control, optimal power point tracking control, and a comprehensive control strategy.

[0077] Fixed-point yaw control strategy refers to adjusting the wind turbine to a preset yaw angle and maintaining stability. This strategy is suitable when wind direction changes are minor or the turbine's installation location is relatively fixed. For example, if the turbine is installed in an area where the wind direction is predominantly northeast, the yaw angle can be set to northeast, and the control system will keep the turbine pointing northeast at all times. Wind direction tracking control strategy involves monitoring wind direction changes in real time using wind direction sensors and adjusting the turbine's yaw angle accordingly to keep the turbine parallel to the wind direction. This strategy is used when wind direction changes are significant or more precise control is required. For example, if the wind direction gradually changes from northeast to southeast, the wind direction tracking control system will monitor the wind direction change in real time and adjust the turbine's yaw angle to always point towards the current wind direction, ensuring the turbine maintains optimal wind alignment. Optimal power point tracking (OPT) control involves monitoring the wind turbine's power output and adjusting its yaw angle accordingly to ensure optimal power output. OPT is used when wind speed varies significantly or when maximizing power generation efficiency is crucial. For example, if a wind turbine generates the most power at 8 m / s, the OPT control system will monitor wind speed changes in real time and adjust the yaw angle to the optimal windward angle to ensure the turbine operates at 8 m / s, thus achieving maximum power output. A comprehensive control strategy combines multiple control strategies, such as wind direction tracking and OPT, selecting the appropriate strategy based on different operating conditions to achieve optimal control. For complex wind turbine operating environments, combining multiple control strategies is necessary for efficient and stable control. For instance, wind direction tracking can be used at lower wind speeds to ensure the turbine maintains the optimal windward angle, while OPT can be used at higher wind speeds to achieve maximum power output.

[0078] The yaw error compensation module is used to compensate for yaw error data in the power curve sub-model to improve the accuracy of the power curve sub-model.

[0079] This embodiment

[0080] (2) Temperature-Power Analysis Model

[0081] The temperature-power analysis model includes a power generation capacity prediction module and a temperature compensation module. The power generation capacity prediction module determines the temperature-power influence coefficient based on preprocessed ambient temperature data, wind turbine power data, and a pre-built temperature-power relationship sub-model. It then predicts the wind turbine power at different temperatures based on this coefficient. The temperature compensation module adjusts the power curve sub-model based on the temperature-power influence coefficient to improve its accuracy. The temperature-power relationship sub-model is determined using linear regression, multinomial regression, or a neural network model, based on the maximum power variation of the wind turbine under different ambient temperatures.

[0082] Specifically, temperature and power typically exhibit a non-linear relationship, which polynomial regression can capture. Therefore, this embodiment uses polynomial regression to construct a temperature-power relationship sub-model. This embodiment utilizes the temperature-power relationship sub-model to predict the wind turbine's power generation capacity at different temperatures, optimize wind turbine operation strategies, increase power generation, and provide a reference for wind turbine operation control.

[0083] (3) Running time analysis model

[0084] The runtime analysis model includes a problem diagnosis module and a maintenance strategy optimization module. The problem diagnosis module determines the operating and non-operating times of the wind turbine in different wind speed ranges based on pre-processed wind speed data, runtime pre-processed data, and a pre-built wind speed-runtime sub-model. It then diagnoses potential causes of wind turbine shutdown based on these operating and non-operating times, such as faults, maintenance, and power outages. Additionally, power outages are determined based on measurement point data, such as 1 representing maintenance, 2 representing operation, and 3 representing overhaul. In some cases, power outages are determined by combining multiple measurement points, such as the AGC active power setpoint and output reduction.

[0085] The maintenance strategy optimization module is used to formulate corresponding maintenance strategies based on diagnostic results, reducing downtime and improving wind turbine utilization. The wind speed-operation time sub-model analyzes the operating and non-operation times of wind turbines in different wind speed ranges to identify potential problems affecting power generation capacity.

[0086] The wind turbine shutdown cause analysis, based on the wind speed-run time relationship model, identifies short running times within certain wind speed ranges as potentially due to faults. The analysis and diagnosis are performed using the model's output, combined with alarm and maintenance records, and then employing a process of elimination and logical reasoning to deduce the shutdown cause. If necessary, experts in relevant fields can be consulted for professional advice. In the maintenance strategy optimization module, corresponding measures are taken based on the identified shutdown cause, such as repairing faults, adjusting control parameters, and optimizing operating strategies. Preventive measures can also be developed based on the shutdown cause to avoid recurrence. This embodiment analyzes running and non-running times under different wind speed ranges and formulates corresponding maintenance strategies based on the analysis results to improve wind turbine utilization.

[0087] (4) Power Loss Analysis Model

[0088] The power loss analysis model includes a loss type module, a loss ratio module, and a loss cause module. The loss type module determines the corresponding power loss for various loss types based on pre-processed data and pre-built power loss sub-modules. These loss types include one or more of maintenance, fault, power curtailment, performance issues, and impact. The loss ratio module determines the power loss ratio for each loss type, providing a reference for optimization strategies. The loss cause module identifies the corresponding causes for each loss type, enabling the development of targeted improvement measures. These improvement measures could include: if the downtime due to a fault is long, enhancing fault repair training and spare parts reserves; and strengthening the performance evaluation and management of maintenance personnel to improve fault repair efficiency and reduce downtime; if the power curtailment is long, actively communicating with the power grid to minimize power curtailment.

[0089] The power loss submodule is constructed through the following steps (5-1) to (5-3).

[0090] Step (5-1): Determine the power generation of the wind turbine for multiple time periods based on the wind turbine operation data and meteorological data.

[0091] Step (5-2) compares the expected power generation of the wind turbines in multiple time periods with the actual power generation in the corresponding time periods to determine the power loss in each time period.

[0092] Step (5-3) determines the type of loss corresponding to the power loss in each time period.

[0093] This embodiment analyzes the power loss caused by various factors through a power loss analysis model, and takes measures to reduce power loss based on the loss analysis results, thereby improving the overall power generation efficiency of the wind farm.

[0094] (5) Pitch Angle Analysis Model

[0095] The pitch angle analysis model includes a pitch angle error correction submodule and a pitch angle compensation submodule. The pitch angle error correction submodule determines the pitch angle error based on the pre-processed pitch angle data and a pre-built pitch angle identification submodel. It then adjusts the pitch angle control strategy according to the error to improve wind energy utilization efficiency. The pitch angle compensation submodule supplements the power curve submodel based on the pitch angle error to improve its accuracy. The pitch angle identification submodel is determined by combining the actual and theoretical pitch angles with a machine learning model. Specifically, machine learning algorithms such as random forests, support vector machines, or neural networks may be used to establish a relationship model between the pitch angle and wind turbine performance (e.g., power output, power generation efficiency).

[0096] Adjusting the pitch angle control strategy involves proposing practical solutions based on the analysis results and addressing the root causes of the problem.

[0097] Root cause-based improvement strategies identify the root causes of problems through in-depth analysis and develop solutions accordingly. This strategy is applicable to various issues, such as power curve deviation, wind misalignment, operating time analysis, and power loss analysis. For example, if a wind turbine has a high power curve deviation rate, analysis reveals it's due to dust accumulation on the blade surface. The improvement strategy involves developing a blade cleaning and maintenance plan, regularly cleaning the blades, improving blade efficiency, and reducing the power curve deviation rate.

[0098] Data-driven improvement refers to utilizing wind turbine operating data and historical fault records, through data analysis and machine learning, to identify potential problems and improvement directions, and then formulate corresponding measures. This strategy is applicable to scenarios with large amounts of data and complex information, such as intelligent diagnostics and optimization suggestions. For example, data analysis might reveal that a certain wind turbine operates for a short time within a specific wind speed range, resulting in low power generation efficiency. The improvement strategy would be to adjust the wind turbine control strategy based on the data analysis results, such as optimizing the start-up and shutdown strategies, to increase the turbine's operating time within the specific wind speed range and improve power generation efficiency.

[0099] Expert-based improvement refers to combining the experience and expertise of wind power specialists to propose improvement solutions for specific problems. This strategy is suitable for problems that are difficult to solve through data analysis or machine learning, such as complex fault diagnosis and handling of special situations. For example, if a wind turbine malfunctions but the analysis results cannot determine the specific cause, the improvement measure is to consult wind power experts, use their experience to investigate and diagnose the fault, and develop a corresponding maintenance plan.

[0100] Cost-effective improvement refers to evaluating the costs and benefits of improvement measures and selecting the most cost-effective solution for implementation. This strategy is suitable for scenarios where costs and benefits need to be weighed, such as replacing equipment or upgrading systems. For example, if a wind turbine has a high power curve deviation rate, it can be resolved by replacing the blades or upgrading the control system, but the costs and benefits of the two solutions are different. The improvement measure is to compare the costs, benefits, and implementation difficulty of the two solutions and select the most cost-effective solution for implementation. For example, if replacing the blades is less costly and has a significant effect, then blade replacement can be chosen; if upgrading the control system is more costly but can improve overall operating efficiency, then upgrading the control system can be chosen.

[0101] The following example further illustrates the cost-benefit-based improvement strategy. First, a threshold (e.g., 10%) is set for the deviation rate to determine whether the deviation is significant. If the deviation rate exceeds the threshold, the cost-benefit analysis phase begins, including options for replacing blades and upgrading the control system.

[0102] a) Blade replacement solution.

[0103] Cost calculation: Consider the price of new blades, installation costs, downtime losses, etc.

[0104] Benefit estimation: Based on the characteristics of the new blades, a power model is used to predict the increase in power generation after the improvement.

[0105] b) Upgrade the control system solution.

[0106] Cost calculation: Consider the price of the new control system, installation and commissioning costs, downtime losses, etc.

[0107] Benefit estimation: Based on the improved control strategy, the potential increase in power generation is predicted using a power model.

[0108] Calculate the payback period and long-term benefits for each option, compare the cost-benefit ratio of two options using decision trees or other algorithms, and select the option with the highest cost-benefit ratio as the final recommendation. Conduct sensitivity analysis, for example, considering uncertainties such as electricity price fluctuations and wind resource changes; use methods such as Monte Carlo simulation to evaluate the feasibility of the options under different scenarios. In addition, periodically re-evaluate the power curve and deviation rate, and continuously optimize the cost-benefit analysis model based on actual implementation results.

[0109] This approach combines machine learning models with traditional cost-benefit analysis, considering multiple influencing factors and improving decision-making accuracy. Sensitivity analysis enhances the robustness of decisions, automating and intelligentizing the decision-making process. This data-driven intelligent decision-making method helps wind farm managers make more scientific and efficient equipment modification decisions. This embodiment identifies pitch angle errors and improves wind energy utilization efficiency by correcting the pitch angle control strategy.

[0110] (6) Micro-site selection analysis model

[0111] The micro-site selection analysis model is used to acquire meteorological preprocessed data from multiple sets of preprocessed data for the target location. This meteorological preprocessed data is then input into a pre-constructed wind field model, which outputs wind resource distribution data. The wind field model is constructed using computational fluid dynamics software, incorporating topographic data from the target location. The computational fluid dynamics software can accurately simulate wind field distribution under complex terrain conditions. The meteorological and processing data includes both topographic preprocessed data and meteorological preprocessed data.

[0112] Specifically, the micro-site selection analysis model includes a wind resource assessment module, a layout optimization module, and a site selection recommendation module. The wind resource assessment module uses a pre-built wind farm model to assess wind resources. This model displays information such as wind turbine locations, wind rose diagrams, and energy rose diagrams on a map to evaluate wind resource utilization efficiency. The layout optimization module optimizes the wind turbine deployment scheme based on the analysis results of the wind resource assessment module to improve wind energy utilization efficiency. Similarly, the site selection recommendation module uses the analysis results of the wind resource assessment module to provide reference suggestions for future wind farm site selection. The site selection recommendation module can provide site selection suggestions based on various factors such as existing wind turbine layout, grid facilities, topography, and environmental factors.

[0113] The micro-site selection analysis model is further illustrated below through examples.

[0114] First, preliminary data collection is necessary, including gathering data on the location, operation, meteorological, grid, and topographical information of wind turbines in operational wind farms. The current operating status of the turbines should be assessed, including power generation efficiency, failure rate, and blade fatigue. Interactions between turbines should be analyzed, such as wind shear and eddy current effects. The capacity and transmission capabilities of grid facilities, such as grid connection points, substations, and transmission lines, should also be analyzed. Next, data analysis is performed. Wind energy resource assessment software is used to evaluate the wind energy utilization efficiency of the existing turbine layout and analyze whether there is room for improvement. CFD simulation software is used to simulate the interactions between turbines, analyze wind shear and eddy current effects, and propose optimized layout schemes. Finally, the carrying capacity of grid facilities is analyzed, and the feasibility of wind farm expansion is assessed.

[0115] For the layout optimization module, the goals are to improve wind energy utilization efficiency, reduce mutual interference between wind turbines, reduce operating costs, and ensure grid security. For wind farms that have already been put into operation, layout optimization needs to take into account factors such as the operating status of existing wind turbines, the limitations of grid facilities, and topography, and formulate targeted optimization plans.

[0116] Common optimization solutions include adjusting turbine spacing, adjusting turbine orientation, adding turbines, and replacing existing turbines. Adjusting turbine spacing involves adjusting the spacing based on factors such as turbine type, wind speed, and wind direction to avoid mutual interference and improve wind energy utilization efficiency. Adjusting turbine orientation uses wind rose and energy rose diagrams to adjust turbine orientation, making it more adaptable to wind energy distribution and improving utilization efficiency. Adding turbines involves adding turbines in suitable areas based on the grid's carrying capacity and wind energy resource assessment results to increase the wind farm's power generation capacity. Replacing existing turbines involves replacing older or lower-capacity turbines based on their operating status to improve the overall power generation efficiency of the wind farm.

[0117] For the site selection recommendation module, based on the analysis results of the wind farms already in operation, it can provide reference suggestions for future expansion, such as determining suitable expansion areas, wind turbine types, and layout schemes; it can also apply the experience of the wind farms already in operation to the site selection of new wind farms, avoid similar mistakes, and improve the design efficiency and operation efficiency of new wind farms.

[0118] The principle of micro-site selection mainly considers factors such as wind energy resources, topography, power grid facilities, and environmental impact to select the optimal layout scheme so that wind turbine generators can efficiently utilize wind energy resources and minimize their impact on the environment.

[0119] Wind energy resource assessment utilizes data such as wind speed, wind direction, and wind energy density to evaluate the wind energy potential of a wind farm area. In mountainous or coastal areas, wind energy resources are typically abundant due to topographical features, making them suitable for wind farm construction. For example, a wind rose diagram of a certain area might show a high proportion of northeasterly winds and a high wind energy density. Based on this analysis, a wind farm could be selected for construction in this area, with wind turbines arranged according to wind direction to maximize the utilization of wind energy resources.

[0120] For topographical analysis, the impact of topography on wind energy resources should be considered, selecting areas conducive to wind energy collection and avoiding areas unfavorable to wind energy collection. Hillsides, coastlines, and other topographical features are conducive to wind energy collection due to the wind acceleration effect, making them suitable for constructing wind power plants. For example, placing wind turbines on hillsides can utilize the wind acceleration effect to improve wind energy utilization efficiency.

[0121] Matching with power grid infrastructure is crucial; placing wind turbines in locations compatible with the power grid ensures successful grid connection and power transmission for wind farms. Wind turbines need to be close to grid facilities, such as substations and transmission lines, to achieve successful grid connection and power generation. For example, choosing an area near a substation for wind farm construction can reduce transmission line losses and improve power generation efficiency.

[0122] Environmental impact assessments evaluate the environmental impacts of wind turbine construction, such as effects on birds, bats, and other wildlife, as well as on the landscape, and propose mitigation measures. The construction of wind turbines needs to consider the impact on the surrounding environment, such as the impact on bird migration routes and the landscape. For example, wind farms should be built in areas far from bird migration routes, and environmentally friendly wind turbines should be used to reduce the impact on birds.

[0123] This embodiment evaluates the efficiency of wind resource utilization and optimizes the wind turbine layout scheme based on the analysis results to improve the efficiency of wind energy utilization.

[0124] (7) AGC curve deviation analysis model

[0125] The AGC curve deviation analysis model includes a deviation assessment module, a control optimization module, and an AGC model correction module. The deviation assessment module determines the AGC curve deviation rate of the wind turbine based on the pre-processed turbine power and a pre-built AGC model; the AGC model is constructed based on time series analysis methods (such as ARIMA). The control optimization module adjusts the wind turbine operation strategy according to the AGC curve deviation rate, including shutdown, startup, and power control strategies. The AGC model correction module corrects the AGC model based on the AGC curve deviation rate.

[0126] Specifically, time series analysis methods are suitable for processing time series data with periodicity and trends. In this embodiment, time series analysis methods are used to predict power fluctuation characteristics and obtain power output results. The deviation assessment module compares the power output results with the actual power output of the wind farm to evaluate the deviation rate of the AGC curve of the wind farm.

[0127] The control optimization module's role is to propose improvements to AGC control strategies, such as adjusting control parameters and optimizing response algorithms; it suggests optimizing wind turbine group control schemes to improve overall AGC response capabilities; it proposes improvements to wind speed prediction models to enhance the foresight of AGC control; and if the problem involves hardware limitations, it may suggest upgrading or modifying relevant equipment. The control optimization module mainly includes model correction, parameter adjustment, and control strategy improvement. Model correction primarily uses AGC deviation analysis results to modify the AGC model and improve AGC control accuracy. Parameter adjustment primarily uses AGC deviation analysis results to adjust AGC control parameters, such as control cycle and power adjustment range, to optimize AGC control performance. Control strategy improvement primarily uses AGC deviation analysis results to improve AGC control strategies, such as introducing prediction algorithms and improving control system response speed, to increase AGC control efficiency.

[0128] For the AGC model correction module, the main causes of deviation include wind speed fluctuations, wind direction changes, wind turbine failures, grid load changes, and control system errors. For wind speed fluctuations, the power control strategy can be adjusted to adjust the wind turbine's power output in a timely manner according to wind speed changes, thus reducing AGC deviation. For wind direction changes, the yaw control strategy can be adjusted to ensure the wind turbine maintains the optimal windward angle, reducing AGC deviation. For wind turbine failures, the faulty turbine needs to be repaired promptly, and the turbine's operating strategy adjusted according to the failure situation to avoid affecting AGC control. For grid load changes, coordination with the grid operator is needed to adjust the wind farm's power output in a timely manner according to grid load changes, reducing AGC deviation. For control system errors, the AGC control system needs to be optimized to improve control accuracy and response speed, thereby reducing AGC deviation.

[0129] Assuming that AGC deviation analysis reveals significant deviations in wind farms when wind speed fluctuations are large, indicating that the AGC model does not adequately account for the impact of wind speed fluctuations, resulting in insufficient control accuracy. To address this, the AGC model is improved by incorporating wind speed fluctuations into the model and predicting the wind farm's power generation based on wind speed changes, thereby enhancing AGC control accuracy.

[0130] In some embodiments, the output of the AGC curve deviation analysis model includes AGC deviation rate, AGC response time, AGC control accuracy, power fluctuation characteristics, and possible influencing factors. Here, AGC deviation rate refers to the percentage difference between the actual power output and the required AGC output; AGC response time refers to the time required for the wind farm to adjust its output to meet AGC requirements; AGC control accuracy is an assessment of the wind farm's ability to maintain its output within the required AGC range; power fluctuation characteristics refer to fluctuation frequency, amplitude, etc.; and possible influencing factors include wind speed changes, equipment response delay, etc.

[0131] The following examples further illustrate the application scenarios of the AGC curve deviation analysis model. The AGC curve deviation analysis model can be applied to the optimization of grid-connected operation of large wind farms. For example, in order to optimize the AGC response capability of wind farms and improve the quality of grid-connected operation, a large wind farm operating under the requirements of grid dispatch needs to accurately respond to AGC commands to ensure the stability of the grid.

[0132] First, operational data is collected, including real-time power output data from the wind farm acquired through the SCADA system, AGC command data obtained from the power grid dispatch center, and meteorological data such as wind speed and direction. Then, data preprocessing is performed, including time series alignment to ensure consistency between the timestamps of AGC commands and actual output data; data smoothing is also performed, using methods such as moving averages to handle instantaneous fluctuations. Next, AGC curve deviation analysis is conducted, using the Dynamic Time Warping (DTW) algorithm to compare the AGC command curve and the actual output curve, calculating indicators such as deviation rate and response time, and using Fourier transform to analyze power fluctuation characteristics. Intelligent diagnosis is then performed, using a decision tree algorithm to analyze the causes of deviations. Inputs include AGC deviation indicators, wind speed data, and equipment status; the output is the main factors causing AGC deviations (such as wind speed prediction errors, inappropriate control strategies, etc.). Finally, based on historical AGC response data and diagnostic results, the AGC control strategy is optimized using a reinforcement learning algorithm, resulting in optimized AGC control parameters and wind turbine group control strategy adjustment suggestions. The output can be a detailed AGC response capability analysis report, specific AGC control strategy optimization suggestions, and a prediction of the improved AGC response effect.

[0133] Through this application scenario, we can see how AGC curve deviation analysis can work in conjunction with the intelligent diagnosis and optimization suggestion module to improve the grid connection quality and grid friendliness of wind farms, which is of great significance for large-scale wind power grid connection and stable grid operation.

[0134] Furthermore, regarding the selection of algorithm models, Support Vector Machines (SVMs) can be used to analyze wind turbine operating data and identify potential faults. Neural Networks can be used to analyze historical wind speed data to predict future wind speeds, providing a reference for wind farm operation. Neural Networks can also be used to analyze wind turbine operating data to predict the future operating status of wind turbines and prevent faults in advance. Random Forests can be used to analyze different types of power loss and identify the main causes of loss. Random Forests can also be used to analyze wind turbine operating data and maintenance records to optimize maintenance strategies and reduce downtime.

[0135] Step S400: Input multiple sets of analysis data into the trained diagnostic model to obtain diagnostic results; the diagnostic model is built based on a machine learning model.

[0136] Based on the analysis results of the aforementioned step S300, the following are included: power curve deviation rate reflecting the degree of deviation between the wind turbine power output and the theoretical model; yaw error reflecting the degree of deviation of the wind turbine from the optimal wind angle; temperature-power relationship reflecting the relationship between wind turbine power and temperature; operating time analysis reflecting the operating time and downtime under different wind speed ranges; power loss analysis reflecting the power loss caused by various factors; power loss reflecting the power loss caused by various factors; pitch angle error reflecting the deviation between the actual pitch angle and the theoretical pitch angle; and micro-site selection analysis results reflecting the efficiency of wind energy resource utilization and the rationality of wind turbine layout.

[0137] The deviation rate of the AGC curve, which reflects the difference between the actual power output of the wind farm and the expected power output of the AGC curve, is used as a feature and input into a pre-trained diagnostic model to identify the key factors affecting the power generation capacity of the wind turbine.

[0138] For diagnostic models, if it is necessary to predict future operating states, supervised learning algorithms such as random forests and gradient boosting trees can be used; if it is necessary to discover hidden patterns in the data, unsupervised learning algorithms such as clustering and dimensionality reduction can be used; if it is necessary to process complex data, deep learning algorithms such as convolutional neural networks and recurrent neural networks can be used.

[0139] The output of the diagnostic model can identify key factors affecting the power generation capacity of wind turbines, predict potential future failures, and assess the operating status of wind turbines.

[0140] In step S500, the diagnostic results are input into a pre-built expert optimization suggestion model to obtain optimization suggestion solutions. Based on the diagnostic results of step S400, targeted optimization suggestions are provided, such as adjusting wind turbine control parameters, optimizing operating strategies, and developing maintenance plans to improve the power generation capacity and operational efficiency of the wind turbine units. The optimization suggestion solutions include adjusting wind turbine control parameters, optimizing wind turbine operating strategies, and developing maintenance plans. Adjusting wind turbine control parameters includes adjusting yaw angle, pitch angle, and / or speed; optimizing wind turbine operating strategies includes shutdown strategies and / or startup strategies; and developing maintenance plans includes regular inspections and / or fault repairs.

[0141] Experts recommend using neural network models such as Multilayer Perceptron (MLP), Convolutional Neural Network (CNN), and Recurrent Neural Network (RNN) depending on the specific situation. Alternatively, tree models such as Random Forest and Gradient Boosting Tree, or other machine learning models, can be used. The model structure needs to be designed based on the data characteristics and analysis objectives, and model performance should be optimized through training and evaluation.

[0142] This embodiment utilizes diagnostic models and expert optimization suggestion models built using algorithms such as machine learning to provide targeted optimization suggestions, thereby improving the power generation capacity and operational efficiency of wind turbine units.

[0143] The principle of the method in this embodiment will be further explained below through Example 1 and Example 2.

[0144] Example 1: Optimization of wind turbine power curve.

[0145] A wind farm operator discovered that the actual power generation of a certain wind turbine was lower than expected. Therefore, it is necessary to optimize the turbine's power curve and improve its power generation efficiency. This includes the following steps (11) to (15).

[0146] Step (11) involves collecting data from multiple sources.

[0147] Wind turbine operating data: Power output, speed, pitch angle, yaw angle, etc. are collected through the SCADA system.

[0148] Meteorological data: collected from weather stations, including wind speed, wind direction, temperature, and air pressure.

[0149] Wind farm operation data: Grid connection status, power curtailment status, etc. are collected through the SCADA system.

[0150] Step (12) involves data preprocessing.

[0151] Data cleaning, normalization, and time alignment are performed. Data cleaning removes outliers, such as data exceeding physically possible ranges. Data with different dimensions is standardized to the [0, 1] interval for easier model processing. Time alignment is performed to ensure consistency in timestamps across different data sources.

[0152] Step (13) involves model selection and processing.

[0153] Three analytical models were selected: a power curve fitting analysis model, a temperature-power analysis model, and a wind deviation analysis model. For the power curve fitting analysis model, a random forest algorithm was chosen. This is because random forests can handle nonlinear relationships, have strong noise resistance, and are less prone to overfitting. Input data included wind speed, temperature, and air pressure. The model outputs predicted power. Temperature and power typically have a nonlinear relationship; multinomial regression can capture this relationship, so it was used to establish a temperature-power analysis model. Temperature data was used as model data, and the model outputs the influence coefficient of temperature on power. Support vector regression also has good generalization ability for small samples and is suitable for handling fine-grained relationships such as yaw error; therefore, it was used to establish a wind deviation analysis model. Wind direction and yaw angle data were used as model inputs, and the model outputs an estimate of yaw error.

[0154] Step (14) Perform intelligent diagnosis.

[0155] Ensemble learning can combine the advantages of multiple models to improve diagnostic accuracy. Therefore, ensemble learning methods (such as gradient boosting decision trees) are used to construct a diagnostic module, which comprehensively analyzes the outputs of each module. In this embodiment, the diagnostic module outputs the main causes of power curve deviation (such as temperature effects, yaw error, etc.).

[0156] Step (15) provides optimization suggestions.

[0157] Expert systems can encode domain knowledge into rules and generate interpretable recommendations. This embodiment uses a rule-based expert system to generate optimization recommendations.

[0158] The expert optimization suggestion model outputs specific optimization recommendations, such as adjusting the yaw control strategy and optimizing the cooling system. It also generates an optimized fan power curve, including specific parameter adjustment suggestions and expected improvement effects.

[0159] Example 2: Optimization of micro-site selection for wind farms.

[0160] To optimize the site selection of new wind turbines and maximize wind energy utilization efficiency, wind farm developers plan to expand existing wind farms. This includes the following steps (21) to (25).

[0161] Step (21) involves collecting data from multiple sources.

[0162] Wind turbine operating data: SCADA data of existing wind turbines.

[0163] Meteorological data: On-site meteorological tower and numerical weather prediction data.

[0164] Topographic data: high-resolution topographic maps and satellite imagery.

[0165] Step (22) involves data preprocessing.

[0166] Perform data interpolation and time series processing; interpolate terrain data to generate continuous terrain models; perform time series analysis on meteorological data to extract long-term trends.

[0167] Step (23) involves model selection and processing.

[0168] We selected a micro-site selection analysis model, a power loss analysis model, and an AGC curve deviation analysis model.

[0169] For the micro-site selection analysis model, a computational fluid dynamics (CFD) model is used to simulate the wind field, taking topographic and meteorological data as input; the output is a high-precision wind resource distribution map. For the power loss analysis model, neural networks can capture complex nonlinear relationships and are suitable for simulating wake effects; therefore, an artificial neural network is used to predict possible wake losses. The model takes wind turbine location, wind speed, and wind direction data as input. The model outputs the predicted wake losses. For the AGC curve deviation analysis model, time series analysis methods (such as ARIMA) are used to predict power fluctuations; the model takes historical power output data as input and outputs the predicted power fluctuation characteristics.

[0170] Step (24) is to perform intelligent diagnosis.

[0171] Genetic algorithms can find a balance among multiple objectives, making them suitable for complex site selection problems. Therefore, the diagnostic module in this embodiment employs a multi-objective optimization algorithm (such as a genetic algorithm). The inputs to the diagnostic module include wind resource distribution, wake loss prediction, and grid constraints. The output of the diagnostic module is a number of possible wind turbine layout schemes.

[0172] Step (25) provides optimization suggestions.

[0173] Decision support systems can consider multiple factors and assist in complex decision-making. Therefore, the expert optimization suggestion model in this embodiment uses a decision support system to evaluate various options. The input of the expert optimization suggestion model is the layout scheme generated by intelligent diagnosis; the output is the optimal wind turbine layout suggestion, including the expected power generation and economic benefit analysis for each location. The results of the expert optimization suggestion model can provide detailed wind farm expansion plans, including the optimal location of new wind turbines, expected power generation, and economic benefit assessment.

[0174] In summary, this embodiment can effectively improve the power generation efficiency of wind turbine units and increase annual power generation through analysis, diagnosis, and optimization; reduce operating costs and improve economic benefits by optimizing wind turbine operation and maintenance; improve the intelligence level of wind farms by utilizing machine learning algorithms to achieve intelligent diagnosis and optimization suggestions; and promote the digital transformation of the wind power industry, providing technical support for the sustainable development of the new energy industry.

[0175] See Figure 2This embodiment provides a wind turbine power generation capacity analysis, diagnosis, and optimization device, including: an acquisition module 100, a preprocessing module 200, an analysis module 300, a diagnosis module 400, and an optimization suggestion module 500. The acquisition module 100 acquires multiple sets of data to be processed, including wind turbine operation data, meteorological data, and / or wind farm operation data. The preprocessing module 200 preprocesses the multiple sets of data to obtain multiple sets of preprocessed data. The analysis module 300 inputs the multiple sets of preprocessed data into corresponding multiple analysis models to obtain multiple sets of analysis data; the multiple analysis models include a power curve fitting analysis model, a wind deviation analysis model, a temperature-power analysis model, an operating time analysis model, a power loss analysis model, a pitch angle analysis model, a micro-situation analysis model, and / or an AGC curve deviation analysis model. The diagnosis module 400 inputs the multiple sets of analysis data into a trained diagnostic model to obtain diagnostic results; the diagnostic model is constructed based on a machine learning model. The optimization suggestion module 500 is used to input the diagnostic results into the pre-built expert optimization suggestion model to obtain optimization suggestion solutions.

[0176] In optional embodiments, wind turbine operating data includes one or more combinations of wind turbine power, wind turbine wind speed, wind turbine wind direction, wind turbine temperature, wind turbine pitch angle, wind turbine windward angle, wind turbine speed, and wind turbine status code; meteorological data includes one or more combinations of ambient wind speed, ambient wind direction, ambient temperature, ambient air pressure, and ambient humidity; and wind farm operating data includes one or more combinations of wind turbine grid connection status, wind turbine power curtailment status, and maintenance records.

[0177] In an optional embodiment, the power curve fitting analysis model includes a wind turbine power prediction module, a deviation rate prediction module, and an update module. The wind turbine power prediction module predicts the wind turbine power corresponding to different environmental wind speeds based on preprocessed data and a pre-built power curve sub-model. The deviation rate prediction module determines the power curve deviation rate based on the actual power, rated power, and theoretical output power of the wind turbine at different environmental wind speeds. The update module updates the power curve sub-model based on the power curve deviation rate; wherein, the power curve sub-model includes a curve fitting module and a sub-model module. The curve fitting module uses various machine learning algorithms to perform curve fitting on historical data of the wind turbine, obtaining multiple primary power curve models for the wind turbine; the machine learning algorithms include support vector machines, neural networks, and / or random forests; the sub-model module evaluates the fitting effect of multiple primary power curve models using evaluation metrics, determines the primary power curve model with the best fitting effect from among the multiple primary power curve models based on the fitting effect, and uses the primary power curve model with the best fitting effect as the power curve sub-model; wherein, the evaluation metrics include the R-squared value and / or the root mean square error.

[0178] In an optional embodiment, the wind deviation analysis model includes a yaw error data module, an error correction module, and a yaw error compensation module. The yaw error data module is used to obtain static yaw error data based on preprocessed wind turbine wind speed data, preprocessed wind turbine angle data, and a pre-built wind deviation sub-model. The wind deviation sub-model is constructed based on a support vector regression algorithm. The error correction module is used to adjust the wind turbine's yaw control strategy based on the static yaw error data to optimize wind energy utilization efficiency. The yaw error compensation module is used to compensate for the yaw error data in the power curve sub-model to improve its accuracy.

[0179] In an optional embodiment, the temperature-power analysis model includes a power generation capacity prediction module and a temperature compensation module. The power generation capacity prediction module determines the temperature-power influence coefficient based on preprocessed ambient temperature data, preprocessed wind turbine power data, and a pre-built temperature-power relationship sub-model. It then predicts the wind turbine power at different temperatures based on this influence coefficient. The temperature compensation module adjusts the power curve sub-model based on the temperature-power influence coefficient to improve its accuracy. The temperature-power relationship sub-model is determined using linear regression, multinomial regression, or a neural network model based on the maximum power variation of the wind turbine under different ambient temperatures.

[0180] In an optional embodiment, the power loss analysis model includes a loss type module, a loss ratio module, and a loss cause module. The loss type module determines the power loss corresponding to various loss types based on preprocessed data and a pre-built power loss sub-module; wherein the various loss types include one or more of maintenance, fault, power curtailment, performance issues, and damage. The loss ratio module determines the power loss ratio relative to each loss type based on the power loss. The loss cause module determines the corresponding loss cause for each loss type. The power loss sub-module includes a power generation module, a power generation module, and a loss type module. The power generation module determines the expected power generation of the wind turbines for multiple time periods based on wind turbine operating data and meteorological data. The power loss module compares the expected power generation of the wind turbines for multiple time periods with the actual power generation for the corresponding time periods to determine the power loss for each time period. The loss type module determines the loss type corresponding to the power loss for each time period.

[0181] In an optional embodiment, the micro-site selection analysis model is used to acquire meteorological preprocessed data from multiple sets of preprocessed data at the target location, input the meteorological preprocessed data of the target location into a pre-built wind field model, and output wind resource distribution data; wherein, the wind field model is constructed based on computational fluid dynamics software and combined with the terrain data of the target location.

[0182] In optional embodiments, the optimization suggestions include adjusting the wind turbine control parameters, optimizing the wind turbine operation strategy, and developing a maintenance plan;

[0183] Adjusting wind turbine control parameters includes adjusting yaw angle, pitch angle, and / or speed;

[0184] Optimizing wind turbine operation strategies includes shutdown strategies and / or startup strategies;

[0185] Developing a maintenance plan includes regular inspections and / or troubleshooting.

[0186] In optional embodiments, the optimization suggestions include adjusting the wind turbine control parameters, optimizing the wind turbine operation strategy, and developing a maintenance plan;

[0187] Adjusting wind turbine control parameters includes adjusting yaw angle, pitch angle, and / or speed;

[0188] Optimizing wind turbine operation strategies includes shutdown strategies and / or startup strategies;

[0189] Developing a maintenance plan includes regular inspections and / or troubleshooting.

[0190] The apparatus provided in the embodiments of this application has the same inventive concept as the method provided in the embodiments of this application. As long as the method can solve the technical problem, the apparatus can also solve the technical problem. This will not be elaborated here.

[0191] As is known from common technical knowledge, this invention can be implemented through other embodiments that do not depart from its spirit or essential characteristics. Therefore, the disclosed embodiments described above are merely illustrative in all respects and are not the only ones. All modifications within the scope of this invention or its equivalents are included in this invention.

Claims

1. A method for analyzing, diagnosing, and optimizing the power generation capacity of wind turbine generators, characterized in that, include: Acquire multiple sets of data to be processed, including wind turbine operation data, meteorological data, and / or wind farm operation data; The multiple sets of data to be processed are preprocessed to obtain multiple sets of preprocessed data; The multiple sets of preprocessed data are input into the corresponding multiple analysis models to obtain multiple sets of analysis data; the multiple analysis models include power curve fitting analysis model, wind deviation analysis model, temperature-power analysis model, operating time analysis model, power loss analysis model, pitch angle analysis model, micro-situation analysis model and / or AGC curve deviation analysis model; The multiple sets of analytical data are input into the trained diagnostic model to obtain diagnostic results; the diagnostic model is built based on a machine learning model. The diagnostic results are input into a pre-built expert optimization suggestion model to obtain an optimization suggestion scheme; The power curve fitting analysis model includes an update module, which is used to update the power curve sub-model according to the power curve deviation rate. The wind deviation analysis model includes: The yaw error data module is used to obtain static yaw error data based on the wind turbine wind speed preprocessing data, wind turbine wind angle preprocessing data and the pre-built wind deviation sub-model in the preprocessed data, wherein the wind deviation sub-model is constructed based on the support vector regression algorithm. A yaw error compensation module is used to compensate the yaw error data into the power curve sub-model to improve the accuracy of the power curve sub-model. The temperature-power analysis model includes: The power generation capacity prediction module is used to determine the influence coefficient of temperature on power based on the environmental temperature preprocessing data, wind turbine power preprocessing data and the pre-built temperature-power relationship sub-model in the preprocessing data, and to predict the wind turbine power at different temperatures based on the influence coefficient of temperature on power. A temperature compensation module is used to adjust the power curve sub-model according to the influence coefficient of temperature on power, so as to improve the accuracy of the power curve sub-model.

2. The method for analyzing, diagnosing, and optimizing the power generation capacity of wind turbine units according to claim 1, characterized in that, The wind turbine operating data includes one or more combinations of wind turbine power, wind turbine wind speed, wind turbine wind direction, wind turbine temperature, wind turbine pitch angle, wind turbine windward angle, wind turbine speed, and wind turbine status code; the meteorological data includes one or more combinations of ambient wind speed, ambient wind direction, ambient temperature, ambient air pressure, and ambient humidity; the wind farm operating data includes one or more combinations of wind turbine grid connection status, wind turbine power curtailment status, and maintenance records.

3. The method for analyzing, diagnosing, and optimizing the power generation capacity of wind turbine units according to claim 1, characterized in that, The power curve fitting analysis model also includes: The wind turbine power prediction module is used to predict the wind turbine power under different environmental wind speeds based on the preprocessed data and the pre-built power curve sub-model. The deviation rate prediction module is used to determine the power curve deviation rate based on the actual power under different environmental wind speeds, the rated power of the wind turbine, and the theoretical output power of the wind turbine. The power curve sub-model is constructed using the following method: Multiple machine learning algorithms are used to perform curve fitting on the historical data of the wind turbine to obtain multiple primary models of the wind turbine's power curves; the machine learning algorithms include support vector machines, neural networks, and / or random forests. The fitting effect of the multiple primary power curve models is evaluated using evaluation metrics. Based on the fitting effect, the primary power curve model with the best fitting effect is determined from the multiple primary power curve models, and the primary power curve model with the best fitting effect is used as the power curve sub-model. The evaluation metrics include the R-squared value and / or the root mean square error.

4. The method for analyzing, diagnosing, and optimizing the power generation capacity of wind turbine units according to claim 3, characterized in that, The wind deviation analysis model also includes: The error correction module is used to adjust the yaw control strategy of the wind turbine based on the static yaw error data in order to optimize wind energy utilization efficiency.

5. The method for analyzing, diagnosing, and optimizing the power generation capacity of wind turbine units according to claim 3, characterized in that, The temperature-power relationship sub-model is determined using linear regression, multinomial regression, or neural network models based on the maximum power variation of wind turbines under different ambient temperatures.

6. The method for analyzing, diagnosing, and optimizing the power generation capacity of wind turbine units according to claim 1, characterized in that, The power loss analysis model includes: The loss type module is used to determine the loss amount corresponding to various loss types based on the preprocessed data and the pre-built loss amount submodule; wherein, the various loss types include one or more of maintenance, fault, power restriction, performance, and damage. The loss ratio module is used to determine the power loss ratio relative to each loss type based on the power loss. The loss cause module is used to determine the loss cause corresponding to each type of loss; The power loss submodule is constructed using the following method: The wind turbine power generation should be determined for multiple time periods based on the wind turbine operation data and the meteorological data; The expected power generation of wind turbines in multiple time periods is compared with the actual power generation in the corresponding time periods to determine the power loss in each time period. Determine the type of loss corresponding to the lost electricity in each time period.

7. The method for analyzing, diagnosing, and optimizing the power generation capacity of wind turbine units according to claim 1, characterized in that, The micro-site selection analysis model is used to acquire meteorological preprocessed data from multiple sets of preprocessed data at the target location, input the meteorological preprocessed data of the target location into a pre-constructed wind field model, and output wind resource distribution data; wherein, the wind field model is constructed based on computational fluid dynamics software and combined with the terrain data of the target location.

8. The method for analyzing, diagnosing, and optimizing the power generation capacity of wind turbine units according to claim 2, characterized in that, The AGC curve deviation analysis model includes: The deviation assessment module is used to determine the AGC curve deviation rate of the wind turbine based on the pre-processed wind turbine power and the pre-built AGC model; wherein the AGC model is constructed based on time series analysis methods. The control optimization module is used to adjust the wind turbine operation strategy according to the AGC curve deviation rate of the wind turbine. The wind turbine operation strategy includes a shutdown strategy, a startup strategy, and a power control strategy. The AGC model correction module is used to correct the AGC model based on the deviation rate of the AGC curve of the wind turbine.

9. The method for analyzing, diagnosing, and optimizing the power generation capacity of wind turbine units according to claim 1, characterized in that, The proposed optimization scheme includes adjusting the wind turbine control parameters, optimizing the wind turbine operation strategy, and developing a maintenance plan. The adjustment of wind turbine control parameters includes adjusting yaw angle, pitch angle and / or speed; Optimizing wind turbine operation strategies includes shutdown strategies and / or startup strategies; Developing a maintenance plan includes regular inspections and / or troubleshooting.

10. A device for analyzing, diagnosing, and optimizing the power generation capacity of a wind turbine generator set, characterized in that, include: The acquisition module is used to acquire multiple sets of data to be processed, including wind turbine operation data, meteorological data and / or wind farm operation data. The preprocessing module is used to preprocess the multiple sets of data to be processed to obtain multiple sets of preprocessed data; The analysis module is used to input the multiple sets of preprocessed data into the corresponding multiple analysis models to obtain multiple sets of analysis data; the multiple analysis models include a power curve fitting analysis model, a wind deviation analysis model, a temperature-power analysis model, an operating time analysis model, a power loss analysis model, a pitch angle analysis model, a micro-situation analysis model, and / or an AGC curve deviation analysis model; The diagnostic module is used to input the multiple sets of analytical data into a trained diagnostic model to obtain diagnostic results; the diagnostic model is built based on a machine learning model. The optimization suggestion module is used to input the diagnostic results into a pre-built expert optimization suggestion model to obtain an optimization suggestion scheme; The power curve fitting analysis model includes an update module, which is used to update the power curve sub-model according to the power curve deviation rate. The wind deviation analysis model includes: The yaw error data module is used to obtain static yaw error data based on the wind turbine wind speed preprocessing data, wind turbine wind angle preprocessing data and the pre-built wind deviation sub-model in the preprocessed data, wherein the wind deviation sub-model is constructed based on the support vector regression algorithm. A yaw error compensation module is used to compensate the yaw error data into the power curve sub-model to improve the accuracy of the power curve sub-model. The temperature-power analysis model includes: The power generation capacity prediction module is used to determine the influence coefficient of temperature on power based on the environmental temperature preprocessing data, wind turbine power preprocessing data and the pre-built temperature-power relationship sub-model in the preprocessing data, and to predict the wind turbine power at different temperatures based on the influence coefficient of temperature on power. A temperature compensation module is used to adjust the power curve sub-model according to the influence coefficient of temperature on power, so as to improve the accuracy of the power curve sub-model.

Images