Method and device for detecting state of wind turbine generator system and storage medium

Abstract

Embodiments of the present application disclose a wind turbine generator set state detection method and device and a storage medium. The method comprises: obtaining actual wind turbine operation data collected by a target wind turbine generator set; obtaining actual power distribution and actual confidence of the actual wind turbine operation data in each reference power distribution interval according to M reference power distribution intervals determined based on a target power curve learning model, the target power curve learning model being a model selected from N candidate power curve models and achieving a preset prediction index in a prediction process of an output power prediction result, and the M reference power distribution intervals corresponding to M wind speed values respectively; and determining a current operation state of the wind turbine generator set according to a comparison result. The technical problem of inaccurate detection results caused by the fact that the current wind turbine generator set detection method cannot dynamically adapt to the operation environment is solved.

Description

Technical Field

[0001] This application relates to the field of wind power generation control, and more specifically, to a method, apparatus, and storage medium for detecting the status of a wind turbine generator set. Background Technology

[0002] To achieve low-carbon and environmentally friendly practices, more and more industries are applying renewable energy to production and daily life. Among them, wind energy, a clean and renewable energy source, has received widespread attention in recent years. However, due to the complex structure and harsh operating environment of wind turbine generators, their failure rate is relatively high, especially for critical components such as gearboxes, generators, and blades. These failures not only incur high repair costs but also lead to prolonged downtime. Therefore, condition monitoring of wind turbine generators has become crucial for maintaining their efficient operation.

[0003] Currently, common methods for detecting the condition of wind turbine generators include signal-based methods, physical model methods, and data-driven methods. Among these, signal-based methods are costly to implement, physical model methods require the involvement of experts with extensive knowledge, and data-driven methods rely on Supervisory Control and Data Acquisition (SCADA) systems for wind turbine generators. However, existing data-driven models not only suffer from complex data preprocessing and poor model generalization ability, but also gradually become ineffective due to changes in wind turbine generator operating conditions (such as aging, sensor recalibration, and replacement of key components), necessitating frequent model updates and retraining.

[0004] In other words, the condition monitoring methods for wind turbine generators provided by relevant technologies cannot dynamically adapt to the constantly changing operating environment of wind turbine generators, resulting in inaccurate condition monitoring results.

[0005] There is currently no effective solution to the above problems. Summary of the Invention

[0006] This application provides a method, apparatus, and storage medium for detecting the condition of a wind turbine generator set, in order to at least solve the technical problem that current wind turbine generator set detection methods cannot dynamically adapt to the operating environment, resulting in inaccurate detection results.

[0007] According to one aspect of the embodiments of this application, a method for detecting the state of a wind turbine generator set is provided, comprising: acquiring actual wind turbine operation data collected from wind turbine generator sets within a target wind farm; acquiring the actual power distribution and actual confidence level of the actual wind turbine operation data in each of the M reference power distribution intervals determined based on a target power curve learning model, wherein the target power curve learning model is a model whose output power prediction result prediction process reaches a preset prediction index selected from N candidate power curve models, and the M reference power distribution intervals correspond to M wind speed values, where M is a natural number greater than 1; traversing the M reference power distribution intervals and comparing the reference power distribution and actual power distribution in each reference power distribution interval; and determining the current operating state of the wind turbine generator set based on the comparison results.

[0008] In one optional implementation, the above-mentioned traversal of M reference power distribution intervals and comparison of the reference power distribution and the actual power distribution in each reference power distribution interval includes: traversing the M reference power distribution intervals and performing the following operations respectively: obtaining the reference power distribution and reference confidence level in the i-th reference power distribution interval; comparing the reference power distribution and the actual power distribution in the i-th reference power distribution interval to obtain the i-th power distribution difference, and comparing the reference confidence level and the actual confidence level in the i-th reference power distribution interval to obtain the i-th confidence level difference; obtaining M power distribution differences and M confidence level differences.

[0009] In one optional implementation, determining the current operating state of the wind turbine generator set based on the comparison results includes: if the difference in the i-th power distribution indicates a power difference between the reference power distribution and the actual power distribution within the i-th reference power distribution interval, and the power difference in the i-th confidence level indicates a power difference greater than a preset threshold, then determining the i-th reference power distribution interval as a candidate fault power distribution interval; counting the number of candidate fault power distribution intervals in the M comparison results; if the number of candidate fault power distribution intervals is greater than a target number threshold, then determining the current operating state of the wind turbine generator set as a fault state; if the number of candidate fault power distribution intervals is less than or equal to the target number threshold, then determining the current operating state of the wind turbine generator set as a non-fault state.

[0010] In an optional implementation, before acquiring the actual wind turbine operation data collected from the wind turbine generators in the target wind farm, the method further includes: acquiring the processed result data after filtering the wind turbine operation data collected from the wind turbine generators in the target wind farm; determining the target power curve learning model from N candidate power curve learning models based on the processed result data, where N is a natural number greater than or equal to 2; and acquiring the M reference power distribution intervals predicted by the target power curve learning model and the reference confidence level corresponding to each reference power distribution interval.

[0011] In one optional implementation, the above-mentioned acquisition of the processed result data after filtering the wind turbine operation data collected from the wind turbine generator sets in the target wind farm includes: calculating the quantiles of the wind turbine operation data to obtain specific quantile values ​​of power output in each wind speed range; filtering the wind turbine operation data based on the specific quantile values ​​to obtain the processed result data, wherein the processed result data is the wind turbine operation data after filtering out abnormal data.

[0012] In one optional implementation, the above-mentioned determination of the target power curve learning model from N candidate power curve learning models based on the processing result data includes: obtaining the learning process evaluation parameters obtained by each candidate power curve learning model after processing the processing result data, wherein the learning process evaluation parameters include: calculation process evaluation parameters and parameter tuning complexity; comparing the learning process evaluation parameters corresponding to each candidate power curve learning model; and determining the target power curve learning model based on the comparison results.

[0013] In an alternative implementation, after determining the current operating status of the wind turbine generator set based on the comparison results, the method further includes: outputting a fault report if the current operating status indicates that the wind turbine generator set is in a fault state.

[0014] According to another aspect of the embodiments of this application, a state detection device for a wind turbine generator set is also provided, comprising: a first acquisition unit, configured to acquire actual wind turbine operation data collected from wind turbine generator sets within a target wind farm; a second acquisition unit, configured to acquire the actual power distribution and actual confidence level of the actual wind turbine operation data within each of the M reference power distribution intervals determined based on a target power curve learning model, wherein the target power curve learning model is a model whose prediction process of output power prediction results selected from N candidate power curve models reaches a preset prediction index, and the M reference power distribution intervals correspond to M wind speed values, where M is a natural number greater than 1; a comparison unit, configured to traverse the M reference power distribution intervals and compare the reference power distribution and actual power distribution within each reference power distribution interval; and a detection unit, configured to determine the current operating state of the wind turbine generator set based on the comparison results.

[0015] In one optional implementation, the comparison unit includes: a processing module, configured to traverse M reference power distribution intervals and perform the following operations respectively: obtain the reference power distribution and reference confidence level in the i-th reference power distribution interval; compare the reference power distribution and actual power distribution in the i-th reference power distribution interval to obtain the i-th power distribution difference, and compare the reference confidence level and actual confidence level in the i-th reference power distribution interval to obtain the i-th confidence level difference; and obtain M power distribution differences and M confidence level differences.

[0016] In one optional implementation, the detection unit includes: a first determining module, configured to determine the i-th reference power distribution interval as a candidate fault power distribution interval when the difference in the i-th power distribution indicates a power difference between the reference power distribution and the actual power distribution within the i-th reference power distribution interval, and the power difference in the i-th confidence difference indicates a power difference greater than a preset threshold; a statistics module, configured to count the number of candidate fault power distribution intervals in the results of M comparisons; and a second determining module, configured to determine the current operating state of the wind turbine generator set as a fault state when the number of candidate fault power distribution intervals is greater than a target number threshold, and to determine the current operating state of the wind turbine generator set as a non-fault state when the number of candidate fault power distribution intervals is less than or equal to the target number threshold.

[0017] In an optional implementation, the above apparatus further includes: a third acquisition unit, configured to acquire filtered wind turbine operation data collected from wind turbine generators in the target wind farm before acquiring actual wind turbine operation data collected from wind turbine generators in the target wind farm; a first determination unit, configured to determine a target power curve learning model from N candidate power curve learning models based on the processed result data, wherein N is a natural number greater than or equal to 2; and a fourth acquisition unit, configured to acquire M reference power distribution intervals predicted by the target power curve learning model and the reference confidence level corresponding to each reference power distribution interval.

[0018] In one optional implementation, the third acquisition unit includes: a quantile calculation module, used to perform quantile calculation on the wind turbine operating data to obtain a specific quantile value of the power output in each wind speed range; and a filtering module, used to filter the wind turbine operating data based on the specific quantile value to obtain processed result data, wherein the processed result data is the wind turbine operating data after filtering out abnormal data.

[0019] In one optional implementation, the first determining unit includes: a first acquiring module, used to acquire learning process evaluation parameters obtained by each candidate power curve learning model after processing the processing result data, wherein the learning process evaluation parameters include: calculation process evaluation parameters and parameter tuning complexity; a comparison module, used to compare the learning process evaluation parameters corresponding to each candidate power curve learning model; and a determining module, used to determine the target power curve learning model based on the comparison results.

[0020] In an alternative embodiment, the above-mentioned device further includes an output unit, configured to output a fault report when the current operating state indicates that the wind turbine generator set is in a fault state, after determining the current operating state of the wind turbine generator set based on the comparison results.

[0021] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored program, wherein the program is executed by a processor to implement the above-described wind turbine generator state detection method.

[0022] According to another aspect of the embodiments of this application, a computer program product is also provided, including a computer program / instructions that, when executed by a processor, implement the above-described wind turbine generator state detection method.

[0023] In this embodiment, after acquiring the actual wind turbine operation data collected from the wind turbine generators within the target wind farm, the actual power distribution and actual confidence level of the actual wind turbine operation data within each of the M reference power distribution intervals determined by the target power curve learning model are obtained. Here, the target power curve learning model is a model whose output power prediction process, selected from N candidate power curve models, achieves a preset prediction index. The M reference power distribution intervals correspond to M wind speed values. Then, the M reference power distribution intervals are traversed, and the reference power distribution and actual power distribution within each interval are compared separately. Based on the comparison results, the current operating state of the wind turbine generator is determined. By selecting a target power curve from N candidate power curve models, learning a model, obtaining a reference power distribution range based on the target power curve learning model, and then comparing the reference power distribution within the reference power distribution range with the actual power to determine the current operating status of the wind turbine generator, the method essentially completes the state detection process of the wind turbine generator in the wind farm by selecting a suitable power curve model. This dynamically adapts to the constantly changing operating environment of the wind turbine generator, thereby improving the accuracy of the dynamic detection results of the wind turbine generator and solving the technical problem of inaccurate detection results caused by the inability of current wind turbine generator detection methods to dynamically adapt to the operating environment. Attached Figure Description

[0024] The accompanying drawings, which are included to provide a further understanding of the embodiments of this application and constitute a part of the embodiments of this application, illustrate exemplary embodiments of this application and, together with their descriptions, serve to explain the embodiments of this application and do not constitute an improper limitation of the embodiments of this application. In the drawings:

[0025] Figure 1 This is a flowchart of an optional wind turbine generator state detection method according to an embodiment of this application;

[0026] Figure 2 This is a flowchart of another optional wind turbine generator state detection method according to an embodiment of this application;

[0027] Figure 3 This is a flowchart of another optional wind turbine generator state detection method according to an embodiment of this application;

[0028] Figure 4 This is a flowchart of another optional wind turbine generator state detection method according to an embodiment of this application;

[0029] Figure 5 This is a schematic diagram of the structure of an optional wind turbine generator condition detection device according to an embodiment of this application;

[0030] Figure 6 This is a schematic diagram of the structure of an optional computer system for implementing the above-described wind turbine generator state detection method according to an embodiment of this application. Detailed Implementation

[0031] To enable those skilled in the art to better understand the embodiments of this application, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the protection scope of the embodiments of this application.

[0032] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of the embodiments of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the present application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0033] It should be noted that all information and data involved in this disclosure are information and data authorized by the user or fully authorized by all parties. The acquisition, storage, and processing of data involved in the embodiments of this application comply with relevant regulations. The information collected is information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of related data all comply with the relevant laws, regulations, and standards of the relevant regions, necessary confidentiality measures have been taken, and they do not violate public order and good morals. Corresponding operation entry points are provided for users to choose to authorize or refuse.

[0034] The present invention will now be described in conjunction with preferred implementation steps. Figure 1 This is a flowchart of a wind turbine generator state detection method according to an embodiment of this application, as follows: Figure 1 As shown, the method includes the following steps:

[0035] S102, acquire actual wind turbine operation data collected from wind turbine generators within the target wind farm;

[0036] Optionally, in this embodiment, the wind turbine generator set is a device that generates electricity using wind energy, and may be composed of, but is not limited to, components such as a wind turbine generator, tower, blades, control system, and converter. When the wind turbine generator set is subjected to wind force, the blades begin to rotate, driving the generator to rotate and generate electrical energy. After being processed by the converter, the electrical energy is fed into the power grid for supply.

[0037] Optionally, in this embodiment, the actual wind turbine operating data is data collected during the actual operation of the wind turbine generator set, which can be exported from a software system for monitoring and controlling industrial processes (Supervisory Control and Data Acquisition, or SCADA for short).

[0038] Optionally, the operational data in this embodiment is used to reflect parameters reflecting the operating status of the wind turbine generator set. The operational data includes internal and external factors. Internal factors reflect the wind turbine generator set's own structure and movement, while external factors reflect the impact of the environment on the wind turbine generator set's operation. External factors include, but are not limited to, wind speed, air density, and ambient temperature. Internal factors include, but are not limited to, blade pitch angle. Wind speed can be obtained by correcting wind speed data obtained from a weather station or a weather station data source; the data format before and after correction is consistent.

[0039] Optionally, for correcting the wind speed data source, it is necessary to collect the original wind speed from a weather station or observatory, as well as the original height at which the original wind speed was measured. The corrected height to be converted should be determined based on actual needs.

[0040] S104. According to the M reference power distribution intervals determined by the target power curve learning model, obtain the actual power distribution and actual confidence level of the actual wind turbine operation data in each reference power distribution interval. The target power curve learning model is a model that achieves the preset prediction index by predicting the output power prediction result selected from N candidate power curve models. The M reference power distribution intervals correspond to M wind speed values, and M is a natural number greater than 1.

[0041] It should be noted that the actual power distribution in this implementation can be obtained from SCADA, but is not limited to.

[0042] Optionally, the target power curve learning model described above in this embodiment is a model for learning and predicting the power curve of a wind turbine generator. In wind power generation, there is a certain relationship between wind speed and power; that is, different wind speeds correspond to different output power. The target power curve learning model establishes a relationship model between wind speed and output power by learning from historical data, and is used to predict the expected output power at different wind speeds.

[0043] Optionally, the confidence level mentioned above can be used to represent the degree of confidence in the current operating status of the wind turbine generator set, or it can be used to represent the degree of confidence in the power distribution range. In wind turbine generator set condition monitoring, the confidence level can help determine the current operating status of the wind turbine generator set so that appropriate measures can be taken in a timely manner.

[0044] Optionally, the collected actual wind turbine operating data can be input into the target power curve learning model. Within a set wind speed range, the model can be divided into several wind speed intervals based on a set starting point, ending point, and step size. M reference power distribution intervals can then be derived from these wind speed intervals. For example, using a wind speed of 10 m / s as the starting point, 5 m / s as the step size, and 30 m / s as the ending point, wind speed intervals of 10 m / s to 15 m / s, 15 m / s to 20 m / s, 20 m / s to 25 m / s, and 25 m / s to 30 m / s can be obtained, corresponding to a value of 4 for M.

[0045] In an optional implementation, the target power curve learning model is selected from N candidate power curve models, and the prediction process of the output power prediction result reaches the preset prediction index. The candidate power curve models are different prediction models or learning models used to predict the output power of wind turbine generators.

[0046] S106, Traverse the M reference power distribution intervals and compare the reference power distribution and the actual power distribution in each reference power distribution interval separately;

[0047] Optionally, in this embodiment, the reference power distribution interval is used to compare the actual power distribution and confidence level of the wind turbine operating data. The reference power distribution interval corresponds to different wind speed values. By comparing the reference power distribution and the actual power distribution within each reference power distribution interval, the current operating state of the wind turbine generator can be determined. These reference power distribution intervals are determined based on a preset wind speed range and the corresponding output power curve model. The reference power distribution interval is a standard power range for measuring the normal operating state of the wind turbine.

[0048] S108 determines the current operating status of the wind turbine generator set based on the comparison results.

[0049] Optionally, in this embodiment, when the actual power distribution is less than the reference power distribution, the actual output power of the wind turbine is less than the theoretical power that the wind turbine should output. Conversely, when the actual output power of the wind turbine is greater than or equal to the theoretical power that the wind turbine should output, the operating status of the wind turbine is determined by comparing the difference between the actual power distribution and the reference power distribution with a set threshold.

[0050] Through the embodiments described above, dynamic adaptation is achieved based on the constantly changing operating environment of wind turbine generator sets, ensuring the accuracy of detection results. By acquiring the actual power distribution and actual confidence level of actual wind turbine operating data within each reference power distribution interval, and comparing the reference power distribution and the actual power distribution separately, a comprehensive analysis of the wind turbine generator set's status can be performed from multiple dimensions. By statistically analyzing the number of candidate fault power distribution intervals in the comparison results, it can be determined whether the current operating state of the wind turbine generator set is a fault state, enabling timely fault detection and output of fault reports. The power curve learning model in the method can predict based on actual wind turbine operating data, more accurately reflecting the actual operating state of the wind turbine generator set and improving the accuracy and efficiency of detection. The flexible selection of N candidate power curve models and M reference power distribution intervals can adapt to the detection needs of wind turbine generator sets in different wind farms.

[0051] As an optional approach, step S106 above, which involves traversing M reference power distribution intervals and comparing the reference power distribution with the actual power distribution within each interval, includes:

[0052] Traverse the M reference power distribution intervals and perform the following operations respectively:

[0053] S1061, obtain the reference power distribution and reference confidence level within the i-th reference power distribution interval;

[0054] S1062, compare the reference power distribution and the actual power distribution within the i-th reference power distribution interval to obtain the i-th power distribution difference, and compare the reference confidence level and the actual confidence level within the i-th reference power distribution interval to obtain the i-th confidence level difference;

[0055] Optionally, in this embodiment, the confidence level is obtained:

[0056] The distance between two points in space is determined as the reference confidence level R1. The straight-line distance between the two points is calculated using the coordinates of the two points and used as the reference confidence level R1.

[0057] Optionally, in this embodiment, R1 is determined based on the following formula:

[0058]

[0059] Where X1 and X2 represent the horizontal axis of the set power, and Y1 and Y2 represent the vertical axis of the set power.

[0060] Treat the actual power distribution and the reference power distribution as two points on a two-dimensional coordinate system, calculate the distance between the two points as the actual confidence level R2, convert the actual power distribution and the reference power distribution into coordinate representation, and obtain the straight-line distance between the two coordinate points as the actual confidence level R2.

[0061] Alternatively, in this embodiment, R2 can be determined based on the following formula:

[0062]

[0063] Where X represents the horizontal axis, Y represents the vertical axis, P2 represents the actual power, and P1 represents the reference power.

[0064] The confidence differences are obtained by comparing R1 and R2.

[0065] S1063, obtain M power distribution differences and M confidence differences.

[0066] Through the embodiments described above, the actual operating data of the wind turbine generator set is dynamically compared with the reference power distribution within a reference power distribution range, thereby achieving dynamic adaptation to the operating environment and making the detection results more accurate. Furthermore, by comparing the reference power distribution and the actual power distribution and obtaining the power distribution differences and confidence level differences, the current operating status of the wind turbine generator set can be determined more accurately, especially in fault diagnosis where it exhibits high accuracy. Additionally, by obtaining reference power distributions and confidence levels within multiple reference power distribution ranges and comparing them with the actual power distribution, the reliability of status detection can be improved, reducing the possibility of misjudgment. Finally, by determining the reference power distribution range based on actual wind turbine operating data and a power curve learning model, and by fully utilizing a data-driven approach to detect the status of the wind turbine generator set, it is possible to better adapt to different operating environments and wind farm conditions.

[0067] As an optional solution, step S108 above, determining the current operating status of the wind turbine generator based on the comparison results, includes:

[0068] S1081, if the difference in the i-th power distribution indicates that there is a power difference between the reference power distribution and the actual power distribution in the i-th reference power distribution interval, and the confidence difference in the i-th power distribution interval indicates that the power difference is greater than a preset threshold, then the i-th reference power distribution interval is determined as a candidate fault power distribution interval.

[0069] S1082, count the number of candidate fault power distribution intervals in the results of M comparisons;

[0070] S1083, if the number of candidate fault power distribution intervals is greater than the target number threshold, the current operating state of the wind turbine generator set is determined to be a fault state; if the number of candidate fault power distribution intervals is less than or equal to the target number threshold, the current operating state of the wind turbine generator set is determined to be a non-fault state.

[0071] Optionally, in this embodiment, the number of candidate fault power distribution intervals can be calculated by, but is not limited to, by setting the effect size and setting the confidence interval.

[0072] Optionally, in this embodiment, the candidate fault power confidence interval width is determined, where the confidence interval width is the allowable error range; the effect size is determined, where the effect size is the degree of difference between the number of candidate fault power distribution intervals and the target number threshold distribution. The number of candidate fault power distribution intervals is calculated using the candidate fault power confidence interval width and the effect size, as shown in the following formula:

[0073]

[0074] Where Z is the quantile of the standard normal distribution of the number of candidate fault power distribution intervals, P is the effect size, and E is the width of the confidence interval.

[0075] Through the embodiments described above, based on the comparison results between the reference power distribution range and the actual power distribution, power distribution ranges that may have faults can be specifically identified, thereby more accurately judging the operating status of wind turbine generators. Furthermore, by statistically analyzing the number of candidate fault power distribution ranges in the comparison results, the current operating status of wind turbine generators can be automatically determined, improving efficiency and reducing manual intervention. In addition, by setting preset thresholds and target quantity thresholds, false judgments and false alarms can be avoided to a certain extent, improving the reliability of the judgment results. Moreover, this method can perform status judgment based on real-time collected wind turbine operating data, enabling timely detection of potential faults and the implementation of corresponding measures to ensure the safe operation of wind turbine generators. Finally, this method can be flexibly applied to the status detection of different wind turbine generators, exhibiting strong versatility and applicability.

[0076] As an alternative solution, such as Figure 2 As shown, before acquiring the actual wind turbine operating data collected from the wind turbine generators within the target wind farm, the following steps are also included:

[0077] S202, Obtain the processed data after filtering the wind turbine operation data collected from the wind turbine generators in the target wind farm;

[0078] Optionally, in this embodiment, the comparison result is obtained by comparing the collected wind turbine operating data with the quantiles of the wind turbine operating data. Anomaly detection is performed on the wind turbine operating data based on the comparison result, filtering out abnormal wind turbine operating data and normal wind turbine operating data, and using the normal wind turbine operating data as the filtered processing result data. The quantiles of the wind turbine operating data can be obtained through methods such as enumeration or cross-validation. The method for obtaining the quantiles in this embodiment is merely an example, and its selection is not limited.

[0079] S204, Based on the processing result data, determine the target power curve learning model from N candidate power curve learning models, where N is a natural number greater than or equal to 2;

[0080] In applying this embodiment, three models can be selected: Multilayer Perceptron (MLP), Radial Basis Function Network (RBFNetwork), and Radial Basis Function Network (LightGBM), with N set to 3. MLP has strong fitting capabilities and is suitable for various types of data, thus it can be used to handle complex power characteristic relationships and wind turbine operating data. RBF performs well in nonlinear problems, thus it can effectively reflect the power curve. LightGBM has efficient training speed and excellent performance, making it suitable for the large number of features and the need for efficient training processes required in this embodiment. The actual wind turbine operating data is input into these three models, and the model whose output power prediction results meet the preset prediction indicators is the target power curve learning model. In this embodiment, the wind speed range and candidate power curve models are merely examples, and their selection is not limited.

[0081] S206, obtain the M reference power distribution intervals predicted by the target power curve learning model and the reference confidence level corresponding to each reference power distribution interval.

[0082] Optionally, in this embodiment, the above-mentioned processed result data after filtering the wind turbine operation data collected from the wind turbine generators in the target wind farm includes:

[0083] S2021, perform quantile calculations on the wind turbine operating data to obtain specific quantile values ​​of power output within each wind speed range;

[0084] S2022, the wind turbine operation data is filtered based on specific quantile values ​​to obtain the processed result data, which is the wind turbine operation data after filtering out abnormal data.

[0085] Optionally, in this embodiment, the above quantile calculation may, but is not limited to, adopt a quantile filtering processing method. That is, the quantile of the power output within each wind speed interval is obtained by performing quantile calculation on the wind turbine operation data through the quantile filtering method.

[0086] Optionally, in this embodiment, the above specific quantile value can be used to measure the power output in different wind speed intervals. By calculating the specific quantile value, the power output under different wind speeds can be better understood, and then the operating status of the wind turbine generator set can be evaluated and monitored.

[0087] Optionally, in this embodiment, a specific quantile q is obtained through the above quantile filtering method. The wind turbine operation data is set as x. The wind turbine operation data x falling outside the specific quantile q is set as normal data, and the wind turbine operation data x falling within the specific quantile q is set as abnormal data. The abnormal data in the wind turbine operation data is filtered, and the obtained normal data is the processed result data. Among them, the probability that x falls within the quantile q is determined by P[X<x]≤k / q, and the probability that x falls outside the quantile q is determined by P[X<x]≥1 - k / q, where q is the specific quantile, x is the k-th quantile of the variable X, and k is the quantile order.

[0088] Through the above embodiments of the present application, the quantile calculation method has strong robustness to outliers, can effectively handle outliers, and improve the accuracy and reliability of data. The quantile calculation method is applicable to various types of data distributions, without assuming that the data follows a specific distribution, and has strong adaptability. By calculating the specific quantile value, the power output under different wind speeds can be intuitively understood, which is beneficial for engineers and operation and maintenance personnel to evaluate and monitor the operating status of the wind turbine generator set. The quantile calculation method has a fast calculation speed and can process a large amount of data in real time, which is beneficial for timely monitoring and analysis of the wind turbine operation data.

[0089] As an optional solution, based on the processed result data, determining the target power curve learning model from N candidate power curve learning models includes:

[0090] S1, obtaining the learning process evaluation parameters obtained by each candidate power curve learning model after processing the processed result data, where the learning process evaluation parameters include: calculation process evaluation parameters, tuning parameter complexity;

[0091] S2, comparing the learning process evaluation parameters corresponding to each candidate power curve learning model;

[0092] S3, determining the target power curve learning model according to the comparison result.

[0093] Optionally, in this embodiment, the learning process evaluation parameters are obtained after the learning model of each candidate power curve processes the processing result data. The learning process evaluation parameters include: calculation process evaluation parameters and parameter tuning complexity.

[0094] Optionally, in this embodiment, the computational evaluation parameters are used to assess and compare the performance of different models, including but not limited to evaluating the computational speed and accuracy of the candidate power curve learning model. The hyperparameter tuning complexity is used to evaluate the complexity of hyperparameter tuning in the target power curve learning model.

[0095] Optionally, in this embodiment, obtaining the learning process evaluation parameters after each candidate power curve learning model processes the processing result data includes: calculating the predicted average value parameter of the result of each candidate power curve learning model. By comparing the predicted average value with the predicted critical value, the accuracy of the calculation results of the candidate power curve learning model is determined using the obtained learning process evaluation parameters.

[0096] For example, specific combinations such as Figure 3 The following process will be used to illustrate:

[0097] S301, define the null hypothesis and the alternative hypothesis.

[0098] Optionally, in this embodiment, the null hypothesis H0 is defined as assuming that all candidate power curve learning models have the same predictive ability; the alternative hypothesis H1 is defined as assuming that at least one candidate power curve learning model has a different predictive ability from the other models.

[0099] S302, Calculate the predicted average value;

[0100] Optionally, in this embodiment, the calculation process may include, but is not limited to, two parts: first, the calculation of the prediction error sequence {fit}, which calculates the prediction results of n models at time t. With truth value y t The difference. Second, define the loss function used in the model as g. For example, the formula can be:

[0101] Optionally, in this embodiment, the prediction results of the learning model for each candidate power curve are calculated to obtain a loss error sequence; a loss difference sequence is calculated based on the prediction loss error sequence; and a prediction average parameter is calculated based on the loss difference sequence.

[0102] Optionally, in this embodiment, the prediction loss error sequence f of the model can be learned through candidate power curves, but is not limited to... it Calculate the loss difference sequence d it The formula is as follows:

[0103] {dit}={g(f it )-g(f i+1，t )}, i=1,2,3,...,n (5)

[0104] Where d is the loss difference sequence; g is the loss function; f is the prediction loss error sequence; i is the candidate power curve learning model order; and t is the running time of the candidate power curve learning model.

[0105] Obtaining the loss error sequence helps analyze the prediction bias of each model under different wind speeds, aiding in the discovery of the model's predictive ability under specific conditions. Evaluation based on the loss difference sequence and the predicted average parameter allows for an objective comparison of differences between models, avoiding the influence of subjective factors.

[0106] Furthermore, in this embodiment, the predicted average value d can be obtained by learning the loss difference sequence of the model from the candidate power curve at time t, but is not limited to. s The formula is as follows:

[0107]

[0108] Where, d s d represents the predicted average; d represents the loss difference sequence; i represents the candidate power curve learning model order; and t represents the running time of the candidate power curve learning model.

[0109] By comparing the predicted average and the predicted critical value, the accuracy of the candidate model's calculation results can be evaluated more intuitively, thus allowing for the selection of the optimal model. Combining the predicted average and the predicted critical value allows for a comprehensive assessment of the model's predictive power and its ability to handle uncertainty, making model selection more comprehensive and objective. Evaluating the predictive power curve learning model using the predicted average parameter enables better selection of the optimal model suitable for the specific scenario in practical applications, improving the practicality and applicability of the technical solution.

[0110] Optionally, in this embodiment, the predicted power result of the candidate power curve learning model is obtained as follows: Obtain the actual power data y from the actual wind turbine operating data, and predict the power result. The prediction loss sequence {f} of the learning model at time t is calculated sequentially using the actual power data y to determine the candidate power curves. it The following formula is used to obtain the loss error sequence:

[0111]

[0112] Where f is the prediction loss error sequence; t represents the predicted power result of the wind turbine state prediction; i represents the order of the candidate power curve learning models; t represents the running time of the candidate power curve learning models.

[0113] Through the embodiments described above, the prediction loss error sequence can objectively reflect the model's prediction accuracy without being affected by other factors. This helps to more accurately evaluate the model's performance. By calculating the prediction loss error sequence at different times, the model's performance under different conditions can be compared. This helps to identify the model's strengths and weaknesses under different conditions, and thus select the most suitable model.

[0114] S303, Determine the prediction critical value;

[0115] Optionally, in this embodiment, the prediction critical value is obtained based on the following method: setting the significance level ∝ value; calculating the prediction critical value based on the significance level and the cumulative probability point 1-∝.

[0116] By setting different significance levels α values ​​through the above embodiments of this application, the strictness of hypothesis testing can be flexibly controlled, thereby meeting the needs of different practical scenarios.

[0117] S305, Judgment of predictive ability of the same model.

[0118] Optionally, in this embodiment, the accuracy of the calculation results of the candidate power curve learning model is determined by comparing the predicted average value with the predicted critical value. This includes: if the predicted average value is less than or equal to the predicted critical value, it is the null hypothesis, and the candidate power curve learning model with the shortest prediction time parameter is selected as the target power curve learning model; if the predicted average value is greater than the predicted critical value, it is the alternative hypothesis, and the target power curve learning model is selected based on the coefficient of determination. The prediction time parameter refers to the computation time of the candidate power curve learning model from the input of actual wind turbine operating data to the calculation of the learning process evaluation parameters. The coefficient of determination is used to determine the degree of fit between the candidate power curve learning model and the actual data, i.e., to determine the accuracy of the prediction by the candidate power curve learning model, thereby determining whether there is a fault in the wind turbine generator set.

[0119] Through the embodiments described above in this application, the predicted average value and the predicted critical value are compared. Combined with the prediction time parameter and the coefficient of determination, the prediction accuracy, running time, and fit to actual data of the model are comprehensively considered, thus enabling a more comprehensive selection of the target power curve learning model. By comprehensively considering multiple parameters, including the prediction time parameter and the coefficient of determination, the process of selecting the target power curve learning model becomes more flexible and adaptable to different situations and requirements.

[0120] S306-1, when the model prediction capabilities are equal, directly select the target power curve to learn the model;

[0121] S306-2, When models have unequal predictive capabilities, the target power curve is selected to learn the model based on the first weight (determination coefficient) and the second weight (prediction time).

[0122] Optionally, in this embodiment, if the predicted average value is greater than the predicted critical value, the target power curve learning model is selected based on the coefficient of determination. This includes: sorting the candidate power curve learning models in descending order of their coefficients of determination; if the difference in the coefficients of determination between the first-ranked candidate power curve learning model and the second-ranked candidate power curve learning model is greater than or equal to 2%, then the first-ranked candidate power curve learning model is selected as the target power curve learning model. If the difference in the coefficients of determination between the first-ranked candidate power curve learning model and the second-ranked candidate power curve learning model is less than 2%, then a weighted calculation is performed based on the runtime to obtain evaluation parameters for the calculation process, and the candidate power curve learning model with the maximum value of the evaluation parameters for the calculation process is selected as the target power curve learning model.

[0123] The method for selecting the optimal model by comparing the coefficient of determination and runtime, as described in the above embodiments of this application, is highly efficient. This embodiment first sorts the models based on their coefficients of determination, and then determines the optimal model based on whether the difference is greater than or equal to 2%. This allows for rapid selection of the optimal model and improves the efficiency of the entire process.

[0124] Optionally, in this embodiment, the determination coefficients based on the candidate power curve learning model... The prediction time parameters of the candidate power curve learning model are weighted to obtain the evaluation parameters of the calculation process, as shown in the following formula:

[0125]

[0126] Among them: Score wi Parameters for evaluating the calculation process; is the determination coefficient of candidate power curve learning model i; W is the weighting coefficient of runtime; T is the runtime of candidate power curve learning model i.

[0127] As an optional approach, after determining the current operating status of the wind turbine generator based on the comparison results, the following steps are also included:

[0128] If the current operating status indicates that the wind turbine generator is in a fault state, output a fault report.

[0129] The embodiments described above enable real-time monitoring of the operating status of wind turbine generators and timely detection of faults. Once a fault is detected, the system automatically outputs a fault report without manual intervention, saving human resource costs.

[0130] The solution in this embodiment will be explained below with reference to a specific application scenario. Assuming that the target wind farm includes three wind power stations, such as A1, A2, and A3, each containing a set of wind turbine generators, the status detection of the wind turbine generators within the wind farm can be achieved through the following process: Figure 4 As shown:

[0131] S401, Start: Begin checking the status of the wind turbine generator set;

[0132] S402, Obtain actual wind turbine operating data and calculate the minimum sample size required for data verification;

[0133] Actual wind turbine operating data for A1, A2, and A3 were acquired using the SCADA system. Data verification was performed on each wind power station to determine the required sample size. Actual data, including actual power distribution and actual confidence levels, was then obtained based on the minimum sample size.

[0134] S403, filter out abnormal data;

[0135] The actual wind turbine operating data is processed using quantile filtering to calculate quantiles, obtaining specific quantile values ​​for power output within each wind speed range. Anomalies are then detected in the actual wind turbine operating data using quantile filtering, filtering out abnormal data and retaining normal data as the processing result.

[0136] S404, Select the power curve learning model;

[0137] MLP, RBF Network, and LightGBM were selected as candidate power curve learning models.

[0138] S405, Evaluate the model's predictive performance;

[0139] The processed data is input into each of the three candidate power curve learning models, and the output of each model is obtained. Based on the output, the candidate power curve learning models are sorted from largest to smallest by their coefficient of determination. If the difference in the coefficient of determination between the first and second ranked candidate power curve learning models is greater than or equal to 2%, the first ranked candidate power curve learning model is selected as the target power curve learning model. If the difference in the coefficient of determination between the first and second ranked candidate power curve learning models is less than 2%, a weighted calculation is performed based on runtime, using the formula... The evaluation parameters of the calculation process are obtained, and the candidate power curve learning model with the maximum value of the evaluation parameters of the calculation process is selected as the target power curve learning model.

[0140] S406-1, Obtain the reference power distribution range;

[0141] S406-2, Obtain the reference confidence level;

[0142] Four wind speed ranges are set: 10–15 m / s, 15–20 m / s, 20–25 m / s, and 25–30 m / s. The wind speed range values ​​are input into the target power curve learning model to obtain the reference power distribution range and reference confidence level.

[0143] S407, Inspection Data;

[0144] Compare the power distribution and confidence level of the faulty data: Compare the actual power distribution with the reference power distribution within the reference power distribution interval. If the power difference between the actual power distribution and the reference power distribution interval is greater than a preset threshold, and the difference between the actual confidence level and the reference confidence level is greater than a preset threshold, the actual power is determined to be abnormal and marked as a candidate faulty power distribution interval. Count the number of candidate faulty power distribution intervals in the results of a set number of comparisons. If the number of candidate faulty power distribution intervals is greater than a target number threshold, the current operating state of the wind turbine generator is determined to be a faulty state. If the number of candidate faulty power distribution intervals is less than or equal to the target number threshold, the current operating state of the wind turbine generator is determined to be a non-faulty state.

[0145] S408 outputs faulty data;

[0146] If the current operating status indicates that the wind turbine generator is in a fault state, output a fault report.

[0147] S409, End.

[0148] According to another aspect of the embodiments of this application, a condition detection device for a wind turbine generator set for implementing the above-described condition detection method for wind turbine generator sets is also provided. For example... Figure 5 As shown, the device includes:

[0149] The first acquisition unit 502 is used to acquire actual wind turbine operation data collected from wind turbine generator sets in the target wind farm.

[0150] The second acquisition unit 504 is used to acquire the actual power distribution and actual confidence level of the actual wind turbine operation data in each reference power distribution interval according to the M reference power distribution intervals determined by the target power curve learning model. The target power curve learning model is a model that achieves the preset prediction index by predicting the output power prediction result selected from N candidate power curve models. The M reference power distribution intervals correspond to M wind speed values, and M is a natural number greater than 1.

[0151] The comparison unit 506 is used to traverse M reference power distribution intervals and compare the reference power distribution and the actual power distribution in each reference power distribution interval separately.

[0152] The detection unit 508 is used to determine the current operating status of the wind turbine generator set based on the comparison results.

[0153] By acquiring actual wind turbine operating data and actual power distribution through the embodiments described above, it is possible to understand the current operating status of the wind turbine generator set in a timely manner, which is beneficial for timely identification and resolution of problems. By comparing the reference power distribution with the actual power distribution, the current operating status of the wind turbine generator set can be accurately determined, improving the accuracy and reliability of data analysis. Automatic data acquisition and comparative analysis reduce manual intervention and improve work efficiency. Furthermore, the reference power distribution range can be flexibly set according to actual conditions to adapt to different wind speed values, improving applicability and flexibility. Finally, by using a model with preset prediction indicators, actual needs can be better met, improving the accuracy and reliability of predictions.

[0154] Optionally, in this embodiment, the comparison unit 506 includes:

[0155] The processing module is used to traverse the M reference power distribution intervals and perform the following operations respectively:

[0156] Obtain the reference power distribution and reference confidence level within the i-th reference power distribution interval; compare the reference power distribution and the actual power distribution within the i-th reference power distribution interval to obtain the i-th power distribution difference, and compare the reference confidence level and the actual confidence level within the i-th reference power distribution interval to obtain the i-th confidence level difference; obtain M power distribution differences and M confidence level differences.

[0157] Through the embodiments described above, real-time acquisition of actual wind turbine operating data enables timely understanding of the wind turbine's operating status, facilitating timely detection and problem-solving. Furthermore, by comparing the reference power distribution range with the actual power distribution, the wind turbine's operating status can be determined more accurately, improving the accuracy and reliability of detection. Additionally, determining the reference power distribution range through a target power curve learning model allows for more effective analysis and evaluation of wind turbine performance, as well as condition identification and assessment.

[0158] Optionally, in this embodiment, the detection unit 508 includes:

[0159] The first determining module is used to determine the i-th reference power distribution interval as a candidate fault power distribution interval when the power distribution difference indicates that there is a power difference between the reference power distribution and the actual power distribution in the i-th reference power distribution interval, and the power difference indicates that the confidence difference is greater than a preset threshold.

[0160] The statistics module is used to count the number of candidate fault power distribution intervals in the results of M comparisons;

[0161] The second determining module is used to determine the current operating state of the wind turbine generator set as a fault state when the number of candidate fault power distribution intervals is greater than the target number threshold; and to determine the current operating state of the wind turbine generator set as a non-fault state when the number of candidate fault power distribution intervals is less than or equal to the target number threshold.

[0162] Optionally, in this embodiment, the above-mentioned device further includes:

[0163] The third acquisition unit is used to acquire the filtered wind turbine operation data collected from the wind turbine generators in the target wind farm before acquiring the actual wind turbine operation data collected from the wind turbine generators in the target wind farm.

[0164] The first determining unit is used to determine the target power curve learning model from N candidate power curve learning models based on the processing result data, where N is a natural number greater than or equal to 2.

[0165] The fourth acquisition unit is used to acquire the M reference power distribution intervals predicted by the target power curve learning model and the reference confidence level corresponding to each reference power distribution interval.

[0166] Through the above embodiments of this application, a modular approach to fault diagnosis can automatically determine the operating status of wind turbine generator sets, reducing the need for manual intervention and improving diagnostic efficiency and accuracy. By setting preset thresholds and target quantity thresholds, adjustments can be made flexibly according to actual conditions, making it applicable to different types and scales of wind turbine generator sets, demonstrating a certain degree of versatility and applicability. Furthermore, it can quickly make judgments in fault conditions, enabling timely implementation of corresponding maintenance and repair measures, effectively reducing the impact of faults on the generator set and improving power generation efficiency and reliability.

[0167] Optionally, in this embodiment, the third acquisition unit includes:

[0168] The quantile calculation module is used to perform quantile calculations on the wind turbine operating data to obtain specific quantile values ​​of power output within each wind speed range.

[0169] The filtering module filters the fan operation data based on specific quantile values ​​to obtain the processed result data, which is the fan operation data after filtering out abnormal data.

[0170] Through the embodiments described above in this application, quantile calculation and filtering can effectively eliminate abnormal data and improve the accuracy of wind turbine operation data. Furthermore, the filtered processing results are cleaner and more reliable, improving the efficiency and accuracy of data analysis.

[0171] Optionally, in this embodiment, the first determining unit includes:

[0172] The first acquisition module is used to acquire the learning process evaluation parameters obtained by the learning model of each candidate power curve after processing the processing result data. The learning process evaluation parameters include: calculation process evaluation parameters and parameter tuning complexity.

[0173] The comparison module is used to compare the evaluation parameters of the learning process corresponding to each candidate power curve learning model;

[0174] The determination module is used to determine the learning model of the target power curve based on the comparison results.

[0175] Through the above embodiments of this application, by comparing the evaluation parameters of the learning process, the most suitable learning model for the target power curve can be objectively determined, avoiding the influence of subjective factors on model selection. Accuracy: By comparing the evaluation parameters of the learning process, the most suitable learning model for processing the result data can be selected, thereby improving the accuracy of the model's prediction and analysis.

[0176] Optionally, in this embodiment, the above-mentioned device further includes: an output unit, which is used to output a fault report when the current operating state indicates that the wind turbine generator set is in a fault state after determining the current operating state of the wind turbine generator set based on the comparison result.

[0177] Through the above embodiments of this application, the output unit can automatically determine the current operating status of the wind turbine generator set based on the comparison results and output a fault report, thereby reducing manual intervention and improving efficiency.

[0178] In this embodiment, the embodiment of the above-mentioned wind turbine generator condition detection device can refer to the above-mentioned method embodiment, and will not be repeated here.

[0179] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided for implementing the above-described wind turbine generator state detection method.

[0180] Optionally, in this embodiment, the computer-readable storage medium may be configured to store a computer program for performing the following steps:

[0181] S1, acquire actual wind turbine operation data collected from wind turbine generators within the target wind farm;

[0182] S2, according to the M reference power distribution intervals determined by the target power curve learning model, obtain the actual power distribution and actual confidence level of the actual wind turbine operation data in each reference power distribution interval. The target power curve learning model is a model that achieves the preset prediction index by predicting the output power prediction result selected from N candidate power curve models. The M reference power distribution intervals correspond to M wind speed values, and M is a natural number greater than 1.

[0183] S3, traverse the M reference power distribution intervals, and compare the reference power distribution and the actual power distribution in each reference power distribution interval separately;

[0184] S4 determines the current operating status of the wind turbine generator set based on the comparison results.

[0185] Optionally, in embodiments of this application, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0186] Optionally, in this embodiment, those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0187] If the integrated units in the above embodiments are implemented as software functional units and sold or used as independent products, they can be stored in the aforementioned computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solutions, can be embodied in the form of software products. These computer software products are stored in storage media and include several instructions to cause one or more computer devices (which may be personal computers, servers, or network devices, etc.) to execute all or part of the steps of the methods in the various embodiments of the embodiments of this application.

[0188] According to one aspect of an embodiment of this application, a computer program product is provided, the computer program product including a computer program / instructions containing program code for performing the methods shown in the flowchart. In such an embodiment, the computer program can be used as... Figure 6 The communication component 609 in the illustrated computer system is downloaded and installed from a network, and / or installed from a removable medium 611. When the computer program is executed by the central processing unit 601, it performs various functions provided in the embodiments of this application.

[0189] Among them, such as Figure 6 As shown, the computer system described above includes a central processing unit (CPU) 601, which performs various appropriate actions and processes based on programs stored in read-only memory (ROM) 602 or programs loaded from storage section 608 into random access memory (RAM). The RAM 603 also stores various programs and data required for system operation. The CPU 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output interface 605 (I / O interface) is also connected to the bus 604.

[0190] The following components are connected to the input / output interface 605: an input section 606 including a keyboard, mouse, etc.; an output section 607 including a cathode ray tube (CRT), liquid crystal display (LCD), and speakers, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card, such as a local area network card or modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the input / output interface 605 as needed. A removable medium 611, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 610 as needed so that computer programs read from it can be installed into the storage section 608 as needed.

[0191] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0192] In the several embodiments provided in this application, it should be understood that the disclosed client can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, indirect coupling or communication connection between units or modules, and may be electrical or other forms.

[0193] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0194] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0195] The above description is only a preferred embodiment of the present application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present application embodiments, and these improvements and modifications should also be considered as the protection scope of the present application embodiments.

Claims

1. A method of detecting a state of a wind turbine, characterized by, The method comprises the following steps: obtaining processed result data filtered from wind turbine operation data collected by a wind turbine generator set in a target wind farm; determining a target power curve learning model from N candidate power curve learning models based on the processed result data, wherein N is a natural number greater than or equal to 2; obtaining M reference power distribution intervals predicted by the target power curve learning model and reference confidence degrees corresponding to each of the reference power distribution intervals; the reference confidence degree is obtained by the following formula: R1 = ； wherein R1 is used to indicate the reference confidence degree, X1 is used to indicate a first wind speed corresponding to a first set power, X2 is used to indicate a second wind speed corresponding to a second set power, Y1 is used to indicate a power value corresponding to the first set power, and Y2 is used to indicate a power value corresponding to the second set power; obtaining actual wind turbine operation data collected by the wind turbine generator set in the target wind farm; obtaining actual power distribution and actual confidence degree of the actual wind turbine operation data in each of the M reference power distribution intervals determined based on the target power curve learning model, wherein the target power curve learning model is a model selected from N candidate power curve models, and a prediction process of an output power prediction result of the model reaches a preset prediction index, and the M reference power distribution intervals correspond to M wind speed values respectively, and M is a natural number greater than 1; the actual confidence degree is obtained by the following formula: R2= ； Wherein, R2 is used for indicating the actual confidence, P2 is used for indicating the actual power, P1 is used for indicating the reference power, the actual confidence is calculated according to the actual power and the reference power, and the actual power is calculated according to the actual confidence and the reference power. A third wind speed corresponding to the actual power is indicated. A fourth wind speed corresponding to the reference power is indicated. A power value corresponding to the actual power is indicated. A power value corresponding to the reference power is indicated. traversing the M reference power distribution intervals, and performing the following operations respectively: obtaining reference power distribution and reference confidence degree in the i-th reference power distribution interval; comparing the reference power distribution in the i-th reference power distribution interval with the actual power distribution to obtain an i-th power distribution difference, and comparing the reference confidence degree in the i-th reference power distribution interval with the actual confidence degree to obtain an i-th confidence degree difference; obtaining M power distribution differences and M confidence degree differences; determining a current operation state of the wind turbine generator set according to the comparison results.

2. The method of claim 1, wherein, The determination of the current operation state of the wind turbine generator set according to the comparison results comprises: in a case where the i-th power distribution difference indicates that there is a power difference between the reference power distribution in the i-th reference power distribution interval and the actual power distribution, and the i-th confidence degree difference indicates that a difference between the reference confidence degree in the i-th reference power distribution interval and the actual confidence degree is greater than a preset threshold, determining that the i-th reference power distribution interval is a candidate fault power distribution interval; counting a number of the candidate fault power distribution intervals in the M comparison results; in a case where the number of the candidate fault power distribution intervals is greater than a target number threshold, determining that the current operation state of the wind turbine generator set is a fault state; and in a case where the number of the candidate fault power distribution intervals is less than or equal to the target number threshold, determining that the current operation state of the wind turbine generator set is a non-fault state.

3. The method of claim 1, wherein, The obtaining of the processed result data filtered from the wind turbine operation data collected by the wind turbine generator set in the target wind farm comprises: The fan operation data is quantile calculated to obtain a specific quantile value of power output in each wind speed interval; The fan operation data is filtered based on the specific quantile value to obtain the processing result data, wherein the processing result data is fan operation data after filtering out abnormal data.

4. The method of claim 1, wherein, The target power curve learning model is determined from the N candidate power curve learning models based on the processing result data, including: Obtain a learning process evaluation parameter obtained after each candidate power curve learning model processes the processing result data, wherein the learning process evaluation parameter includes a calculation process evaluation parameter and a parameter adjustment complexity; The learning process evaluation parameters corresponding to each candidate power curve learning model are compared; The target power curve learning model is determined according to the comparison result.

5. The method according to any one of claims 1 to 4, characterized in that, After determining the current operating state of the wind turbine generator set according to the comparison result, the method further includes: In the case that the current operating state indicates that the wind turbine generator set is in a fault state, output a fault report.

6. A condition detection device for a wind power plant, characterized in that It includes: The first acquisition unit is configured to acquire actual fan operation data collected by wind turbine generators in a target wind farm; The second acquisition unit is configured to acquire actual power distribution and actual confidence in each reference power distribution interval of the actual fan operation data according to M reference power distribution intervals determined based on a target power curve learning model, wherein the target power curve learning model is a model selected from N candidate power curve models, and the prediction process of the output power prediction result reaches a preset prediction index, and the M reference power distribution intervals correspond to M wind speed values respectively, and M is a natural number greater than 1; the actual confidence is obtained by the following formula: R2= ； Wherein, R2 is used for indicating the actual confidence, P2 is used for indicating the actual power, P1 is used for indicating the reference power, the actual confidence is calculated according to the actual power and the reference power, and the actual power is calculated according to the actual confidence and the reference power. A third wind speed corresponding to the actual power is indicated. A fourth wind speed corresponding to the reference power is indicated. A power value corresponding to the actual power is indicated. A power value corresponding to the reference power is indicated. The comparison unit is configured to traverse M reference power distribution intervals and perform the following operations respectively: acquiring reference power distribution and reference confidence in the i-th reference power distribution interval; comparing the reference power distribution in the i-th reference power distribution interval with the actual power distribution to obtain the i-th power distribution difference, and comparing the reference confidence in the i-th reference power distribution interval with the actual confidence to obtain the i-th confidence difference; obtaining M power distribution differences and M confidence differences; The detection unit is configured to determine the current operating state of the wind turbine generator set according to the comparison result; The device is also used for: Obtain the processing result data after filtering the fan operation data collected by the wind turbine generators in the target wind farm; determine the target power curve learning model from N candidate power curve learning models based on the processing result data, wherein N is a natural number greater than or equal to 2; obtain the M reference power distribution intervals and the reference confidence corresponding to each reference power distribution interval predicted by the target power curve learning model; the reference confidence is obtained by the following formula: R1 = ； Wherein, R1 is used to indicate the reference confidence, X1 is used to indicate a first wind speed corresponding to a first set power, X2 is used to indicate a second wind speed corresponding to a second set power, Y1 is used to indicate a power value corresponding to the first set power, and Y2 is used to indicate a power value corresponding to the second set power.

7. A computer readable storage medium, characterized in that, The computer readable storage medium comprises a stored program, wherein the program, when executed by a processor, performs the method of any one of claims 1 to 5.

8. A computer program product comprising computer programs / instructions, characterized in that, The computer program / instruction, when executed by a processor, implements the steps of the method of any one of claims 1 to 5.

Images