Wind power plant blade icing prediction method, device, equipment and medium

Abstract

The invention discloses a wind power plant blade icing prediction method and device, equipment and a medium, and relates to the field of wind power generation. The method comprises the following steps: acquiring predicted meteorological data of a wind power plant in a future preset time period, wherein the predicted meteorological data comprises temperature and humidity; and inputting the predicted meteorological data into a pre-trained blade icing prediction model, processing the predicted meteorological data by the blade icing prediction model, and outputting a blade icing prediction result of the wind power plant, the blade icing prediction result representing a blade icing condition in a future preset time period. The blade icing prediction model is obtained through training according to historical sample data of the wind power plant, the historical sample data comprises meteorological data in historical icing time and meteorological data in historical non-icing time, the meteorological data comprises temperature and humidity, and the historical sample data has labels used for representing blade icing conditions corresponding to the meteorological data. According to the embodiment of the invention, the safety risk of the wind power plant can be reduced.

Description

Technical Field

[0001] The present application relates to the field of wind power generation, and in particular to a method, device, equipment and medium for predicting icing on blades in a wind farm. Background Art

[0002] In winter and spring, when wind turbines are operating in low temperature environments, water droplets in the natural environment may form thin ice on the blades of wind turbines, i.e., blade ice. After the blades are covered with ice, the ice layer will gradually grow and thicken, causing the aerodynamic characteristics of the blades to change, and the dynamic load of the wind turbine to increase, which will bring safety risks to the wind turbine.

[0003] Blade icing is caused by natural phenomena and is irresistible. In order to ensure the safety of wind turbines, it is necessary to monitor the status of blade icing. At present, it mostly depends on manual observation of blade status to determine whether blade icing has occurred. However, real-time observation of blade icing has a lag. When manual observation of blade icing occurs, the wind turbines in the wind farm have been running for a long time with blade icing, making it difficult to detect blade icing in time, which makes the wind farm have a greater safety risk. Summary of the invention

[0004] The embodiments of the present application provide a method, device, equipment and medium for predicting blade icing in a wind farm, which can detect blade icing in a timely and accurate manner and reduce the safety risks of the wind farm.

[0005] In a first aspect, an embodiment of the present application provides a method for predicting blade icing in a wind farm, comprising: obtaining predicted meteorological data of the wind farm within a preset time period in the future, the predicted meteorological data including temperature and humidity; inputting the predicted meteorological data into a pre-trained blade icing prediction model, the blade icing prediction model processes the predicted meteorological data, and outputs a blade icing prediction result of the wind farm, the blade icing prediction result is used to characterize the blade icing condition within the preset time period in the future, wherein the blade icing prediction model is trained based on historical sample data of the wind farm, the historical sample data including meteorological data within the historical icing time and meteorological data within the historical non-icing time, the meteorological data including temperature and humidity, and the historical sample data has a label, and the label is used to characterize the blade icing condition corresponding to the meteorological data.

[0006] In some possible embodiments, before the predicted meteorological data is input into a pre-trained blade icing prediction model, it also includes: obtaining historical sample data and an icing shutdown ratio corresponding to the meteorological data in the historical sample data, the icing shutdown ratio including the ratio of the number of wind turbines shut down due to blade icing in the wind farm to the number of wind turbines in the wind farm; generating a label according to the icing shutdown ratio; establishing a corresponding relationship between the label and the historical sample data; and using at least a portion of the historical sample data with the label as a training data set to train a machine learning model to obtain a blade icing prediction model.

[0007] In some possible embodiments, the historical sample data further includes meteorological data within a first preset time period before at least one data collection moment in the historical ice coverage time.

[0008] In some possible embodiments, the meteorological data during the historical icing period used to train the blade icing prediction model includes first-category icing meteorological data, second-category icing meteorological data, and third-category icing meteorological data.

[0009] Among them, the first type of ice-covered meteorological data includes meteorological data with corresponding labels characterizing blade ice covering, temperature lower than the preset ice-covered temperature and humidity higher than the preset ice-covered humidity; the second type of ice-covered meteorological data includes meteorological data with corresponding labels characterizing blade ice covering, continuous appearance of labels characterizing blade ice covering within the previous second preset time period, temperature lower than the preset ice-covered temperature and humidity lower than or equal to the preset ice-covered humidity; the third type of ice-covered meteorological data includes meteorological data with corresponding labels characterizing blade ice covering, continuous appearance of labels characterizing blade ice covering within the previous third preset time period and temperature higher than or equal to the preset ice-covered temperature.

[0010] In some possible embodiments, the blade icing prediction model includes a blade icing classification prediction model and / or a blade icing ratio prediction model; when the blade icing prediction model includes a blade icing classification prediction model, the blade icing prediction result includes blade icing or blade non-icing, and the label characterizes blade icing or blade non-icing; when the blade icing prediction model includes a blade icing ratio prediction model, the blade icing prediction result includes an icing shutdown ratio, the label characterizes the icing shutdown ratio, and the icing shutdown ratio includes the ratio of the number of wind turbines in the wind farm that have been shut down due to blade icing to the number of wind turbines in the wind farm.

[0011] In some possible embodiments, the blade icing prediction result includes an icing shutdown ratio, and the icing shutdown ratio includes a ratio of the number of wind turbine generator sets that are shut down due to blade icing in the wind farm to the number of wind turbine generator sets in the wind farm.

[0012] After the predicted meteorological data is input into a pre-trained blade icing prediction model, the predicted meteorological data is processed by the blade icing prediction model, and the blade icing prediction result of the wind farm is output, the method further includes: calculating the number of wind turbines in the wind farm predicted to be shut down due to icing in the preset area according to the blade icing prediction result of the wind farm in the preset area and the number of wind turbines in the wind farm in the preset area; determining the number of wind turbines in the preset area that are shut down due to icing according to the number of wind turbines in the wind farm predicted to be shut down due to icing in the preset area; and executing a unit protection strategy corresponding to the number of wind turbines shut down due to icing in the preset area.

[0013] In some possible embodiments, before using at least part of the historical sample data with labels as a training data set for model training, the method further includes: selecting one wind farm from more than two wind farms as a standard wind farm; normalizing the historical sample data of the standard wind farm and obtaining a normalized proportional coefficient; and normalizing the historical sample data of wind farms other than the standard wind farm using the normalized proportional coefficient;

[0014] Using at least part of the labeled historical sample data as a training data set for model training includes: using at least part of the labeled normalized historical sample data as a training data set for model training.

[0015] In some possible embodiments, the predicted meteorological data is consistent with the type of meteorological data, and the meteorological data further includes one or more of the following: wind speed, precipitation, cloud cover, air pressure, and dew point temperature.

[0016] In a second aspect, an embodiment of the present application provides a device for predicting blade icing in a wind farm, including: a data acquisition module, used to obtain predicted meteorological data of the wind farm in a future preset time period, the predicted meteorological data including temperature and humidity; a model processing module, used to input the predicted meteorological data into a pre-trained blade icing prediction model, the blade icing prediction model processes the predicted meteorological data, and outputs a blade icing prediction result of the wind farm, the blade icing prediction result is used to characterize the blade icing situation in the future preset time period, wherein the blade icing prediction model is trained based on historical sample data of the wind farm, the historical sample data includes meteorological data within the historical icing time and meteorological data within the historical non-icing time, the meteorological data includes temperature and humidity, and the historical sample data has a label, and the label is used to characterize the blade icing situation corresponding to the meteorological data.

[0017] In a third aspect, an embodiment of the present application provides a device for predicting icing on blades in a wind farm, comprising: a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, the method for predicting icing on blades in a wind farm according to the first aspect is implemented.

[0018] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having computer program instructions stored thereon. When the computer program instructions are executed by a processor, the method for predicting wind farm blade icing according to the first aspect is implemented.

[0019] The embodiments of the present application provide a method, device, equipment and medium for predicting blade icing in a wind farm, which can use a pre-trained blade icing prediction model to process the predicted meteorological data of the wind farm input in a preset time period in the future, thereby obtaining a blade icing prediction result output by the blade icing prediction model. The blade icing prediction model is trained based on the labeled historical sample data of the wind farm, and the historical sample data includes the meteorological data within the historical icing time and the meteorological data within the historical non-icing time. The labels of the historical sample data can characterize the blade icing conditions corresponding to the meteorological data. The blade icing prediction model trained based on the labeled historical sample data of the wind farm can accurately describe the nonlinear dynamic process of blade icing evolution, so that the blade icing conditions in the preset time period in the future can be accurately predicted in advance based on the predicted meteorological data in the preset time period in the future, and blade icing can be detected in a timely and accurate manner, thereby reducing the safety risks of the wind farm. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solution of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0021] Figure 1 A flow chart of a method for predicting wind farm blade icing provided in an embodiment of the present application;

[0022] Figure 2 A schematic diagram of an example of a wind farm blade icing prediction hit rate provided in an embodiment of the present application;

[0023] Figure 3 A flow chart of a method for predicting wind farm blade icing provided by another embodiment of the present application;

[0024] Figure 4 A flow chart of a method for predicting wind farm blade icing provided in another embodiment of the present application;

[0025] Figure 5A schematic diagram of an example of comparison between the actual ice-covered shutdown capacity and the predicted ice-covered shutdown capacity in a certain area provided in an embodiment of the present application;

[0026] Figure 6 A schematic diagram of the structure of a wind farm blade icing prediction device provided in one embodiment of the present application;

[0027] Figure 7 A schematic diagram of the structure of a wind farm blade icing prediction device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0028] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, rather than to limit the present application. For those skilled in the art, the present application can be implemented without the need for some of these specific details. The following description of the embodiments is only to provide a better understanding of the present application by illustrating the examples of the present application.

[0029] In winter, spring and other seasons, when wind turbines are operating in low temperature environments, water droplets in the natural environment may form thin ice on the blades of the wind turbines, i.e., blade icing. After the blades are covered with ice, the ice layer will gradually grow and thicken, causing the aerodynamic characteristics of the blades to change, and the dynamic load of the wind turbine to increase, bringing safety risks to the wind turbines. Blade icing is caused by natural phenomena and is irresistible. In order to ensure the safety of wind turbines, it is necessary to monitor the state of blade icing. At present, it mostly relies on manual observation of the blade state to determine whether blade icing occurs. However, real-time observation of blade icing has a lag. When manual observation of blade icing occurs, the wind turbines in the wind farm have been running for a long time in the state of blade icing, and it is difficult to detect blade icing in time, which makes the wind farm have a greater safety risk. In some cases, in order to monitor the blade icing situation, physical mathematical models are used to simulate natural phenomena and simulate the icing formation mechanism. However, the icing formation mechanism is complex and changeable, and it is difficult to fully simulate the physical formation of blade icing. Ignored factors will become the source of error, which will reduce the accuracy of blade icing prediction. In other cases, in order to monitor the blade icing situation, the blade icing situation of wind turbines in different time periods can be statistically analyzed based on mathematical statistics to make predictions. However, due to the instability of meteorology, the meteorological conditions in the same month of different years are also different. Mathematical statistics, a single time series prediction method, cannot accurately describe the nonlinear dynamic process of blade icing evolution, and will also reduce the accuracy of blade icing prediction.

[0030] The present application provides a method, device, equipment and medium for predicting blade icing in a wind farm, which can utilize meteorological data from the historical blade icing time when blade icing occurred in the history of the wind farm and meteorological data from the historical non-icing time when blade icing did not occur in the history to train a blade icing prediction model that can output a predicted blade icing condition of the wind farm based on the meteorological data. The blade icing prediction model is trained using meteorological data from the historical time and whether blade icing occurred in the historical time, and can accurately describe the nonlinear dynamic process of blade icing evolution. By inputting the predicted meteorological data for a future preset time period into the blade icing prediction model, a blade icing prediction result that characterizes the blade icing condition for the future preset time period can be obtained, thereby realizing the prediction of blade icing in the wind farm, timely and accurately detecting blade icing, and reducing the safety risk of the wind farm.

[0031] In a first aspect, the present application provides a method for predicting blade icing in a wind farm, which can be applied to a scenario for predicting blade icing in a wind farm. The method for predicting blade icing in a wind farm can be executed by a prediction device, equipment, etc. for predicting blade icing in a wind farm. Figure 1 A flow chart of a method for predicting wind farm blade icing provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, the wind farm blade icing prediction method may include step S101 and step S102.

[0032] In step S101, forecasted meteorological data of a wind farm within a preset future time period is obtained.

[0033] The number of wind farms to be monitored is not limited here. When it is necessary to monitor a wind farm, the predicted meteorological data of the wind farm in the future preset time period is obtained; when it is necessary to monitor multiple wind farms in a preset area such as a region or a province, the predicted meteorological data of each wind farm in the future preset time period is obtained.

[0034] The predicted meteorological data includes the predicted meteorological data within a preset time period in the future, which may include but is not limited to temperature and humidity. In some examples, the predicted meteorological data may also include one or more of the following: wind speed, precipitation, cloud cover, air pressure, dew point temperature. The preset time period in the future can be set according to the scene, demand, experience, etc., and is not limited here. For example, the preset time period in the future may be the next 3 days. The predicted meteorological data can be obtained from official meteorological agencies, or a meteorological prediction model can be obtained by pre-training and predicted by the meteorological prediction model. Other channels that can obtain predicted meteorological data are also within the protection scope of the embodiments of the present application, and are not described here one by one.

[0035] In step S102, the predicted meteorological data is input into a pre-trained blade icing prediction model, and the blade icing prediction model processes the predicted meteorological data to output a blade icing prediction result of the wind farm.

[0036] The blade icing prediction model is trained based on the historical sample data of the wind farm. The blade icing prediction model belongs to a machine learning model. The blade icing prediction model may include but is not limited to one or more of the long short-term memory network (Long Short-Term Memory, LSTM) model, XGBoost model, etc. The blade icing prediction model that can be trained to output the blade icing situation according to the input meteorological conditions is within the protection scope of the embodiments of the present application. The historical sample data includes meteorological data within the historical icing time and meteorological data within the historical non-icing time. The historical icing time includes the time period in the historical time when blade icing occurs in the wind farm, and the meteorological data within the historical icing time can reflect the characteristics of the meteorology that causes blade icing. The historical non-icing time includes the time period in the historical time when blade icing does not occur in the wind farm, and the meteorological data within the historical non-icing time can reflect the characteristics of the meteorology that does not cause blade icing. Meteorological data may include temperature and humidity. The above-mentioned predicted meteorological data is consistent with the type of meteorological data. In some examples, the meteorological data may also include one or more of the following: wind speed, precipitation, cloud cover, air pressure, dew point temperature. Rich meteorological data can further improve the accuracy of predicting blade icing conditions. In some examples, the historical sample data may also include meteorological data within a first preset time period before at least one data collection moment in the historical icing time. The first preset time period may be set according to the scenario, demand, experience, etc. For example, the first preset time period may be 8 hours. The label of the meteorological data within the first preset time period before at least one data collection moment in the historical icing time may be obtained based on the statistics of blade icing conditions within the first preset time period before at least one data collection moment in the historical icing time. The meteorological data within the first preset time period before at least one data collection moment in the historical icing time may be used as additional historical sample data to increase the number of historical sample data used to train the blade icing prediction model, so as to further improve the accuracy and reliability of the trained blade icing prediction model.

[0037] The historical sample data has a label, and the label is used to characterize the blade icing condition corresponding to the meteorological data. The blade icing condition may include but is not limited to whether blade icing occurs in the wind farm and / or the proportion of wind turbines in the wind farm that are shut down due to blade icing. The label of the historical sample data can be obtained based on the blade icing condition of the wind farm collected during the time corresponding to the historical sample data. The blade icing prediction model trained with the historical sample data with labels can output the predicted blade icing condition of the wind farm based on the input meteorological data. In an embodiment of the present application, the predicted meteorological data within a preset time period in the future is input into the blade icing prediction model, and the blade icing prediction result output by the blade icing prediction model can be obtained, and the blade icing prediction result can be used to characterize the blade icing condition within a preset time period in the future.

[0038] In an embodiment of the present application, the blade icing prediction model obtained by pre-training can be used to process the predicted meteorological data of the input wind farm in a preset time period in the future, so as to obtain the blade icing prediction result output by the blade icing prediction model. The blade icing prediction model is trained based on the historical sample data with labels of the wind farm, and the historical sample data includes the meteorological data within the historical icing time and the meteorological data within the historical non-icing time. The labels of the historical sample data can characterize the blade icing conditions corresponding to the meteorological data. The blade icing prediction model obtained by training based on the historical sample data with labels of the wind farm can accurately describe the nonlinear dynamic process of blade icing evolution, so that the blade icing conditions within the preset time period in the future can be accurately predicted in advance based on the predicted meteorological data within the preset time period in the future, and the blade icing can be detected in a timely and accurate manner, thereby reducing the safety risks of the wind farm.

[0039] For example, Figure 2 A schematic diagram of an example of a wind farm blade icing prediction hit rate provided in an embodiment of the present application, as shown in FIG. Figure 2 As shown, the horizontal axis is the wind farm number, and the vertical axis is the hit rate. The method for predicting wind farm blade icing in the embodiment of the present application was applied to 51 wind farms, and the blade icing prediction results of each of the 51 wind farms were obtained. The blade icing prediction results of the wind farms were compared with the actual blade icing conditions of the wind farms to obtain the hit rate. The hit rate can characterize the consistency between the blade icing prediction results of the wind farms and the actual blade icing conditions of the wind farms. Specifically, it can be the ratio of the accurately predicted blade icing duration to the total duration, and the total duration is the sum of the accurately predicted blade icing duration and the inaccurately predicted blade icing duration. The hit rate can be calculated by TP / (TP+FP), wherein TP is the accurately predicted blade icing duration, and FP is the inaccurately predicted blade icing duration. By Figure 2It can be obtained that among the 51 wind farms, 98% of the wind farms have a hit rate of blade icing prediction of more than 60%, and 94% of the wind farms have a hit rate of blade icing prediction of more than 80%. The blade icing conditions predicted by the wind farm blade icing prediction method in the embodiment of the present application are basically consistent with the actual blade icing conditions, and the accuracy of predicting the blade icing conditions can be guaranteed.

[0040] In some embodiments, before using the blade icing prediction model to predict the future blade icing conditions of a wind farm, it is necessary to first train the blade icing prediction model. Figure 3 This is a flow chart of a method for predicting wind farm blade icing provided by another embodiment of the present application. Figure 3 and Figure 1 The difference is that Figure 3 The wind farm blade icing prediction method shown may further include steps S103 to S106.

[0041] In step S103, historical sample data and the icing shutdown ratio corresponding to the meteorological data in the historical sample data are obtained.

[0042] The ice shutdown ratio includes the ratio of the number of wind turbines that have been shut down due to blade ice to the number of wind turbines in the wind farm. When blade ice appears on a wind turbine, if the wind turbine continues to operate, it will cause great safety risks. Therefore, the wind turbine will be shut down when blade ice appears on the wind turbine. The value range of the ice shutdown ratio is [0, 1].

[0043] In step S104, a label is generated according to the icing shutdown ratio.

[0044] In some examples, the value of the tag may be equal to the ice-covered shutdown ratio, that is, the tag may represent the ice-covered shutdown ratio.

[0045] In other examples, the value of the label may be a first value or a second value, and the first value and the second value are different. The first value indicates that the blades in the wind farm are not covered with ice, and the second value indicates that the blades in the wind farm are covered with ice. If the ice-covered shutdown ratio is less than or equal to the preset ratio threshold, the value of the label takes the first value, and if the ice-covered shutdown ratio is greater than the preset ratio threshold, the value of the label takes the second value. The preset ratio threshold can be set according to scenarios, requirements, experience, etc. For example, the preset ratio threshold is 0, that is, if there is more than one wind turbine generator set in the wind farm corresponding to the historical sample data that is shut down due to blade ice, then the label corresponding to the historical sample data is the second value, and if there is no wind turbine generator set in the wind farm corresponding to the historical sample data that is shut down due to blade ice, then the label corresponding to the historical sample data is the first value.

[0046] In step S105, a corresponding relationship between the label and the historical sample data is established.

[0047] Establish a corresponding relationship between labels and historical sample data so that the historical sample data has labels.

[0048] In step S106, at least part of the historical sample data with labels is used as a training data set to perform machine learning model training to obtain a blade icing prediction model.

[0049] All the historical sample data with labels can be used as training data sets to train a blade icing prediction model. Alternatively, a portion of the historical sample data with labels can be selected as a training data set, and another portion of the historical sample data with labels can be used as a test data set. The training data set can be used to train a blade icing model, and the trained blade icing model can be tested using the test data set to determine whether the accuracy of the trained blade icing model meets the requirements. For example, 80% of the historical sample data with labels can be used as a training data set, and the remaining 20% ​​of the historical sample data with labels can be used as a test data set.

[0050] In some examples, the meteorological data in the historical icing period used to train the blade icing prediction model, ie, the training data set, may include first-category icing meteorological data, second-category icing meteorological data, and third-category icing meteorological data.

[0051] The first type of icing meteorological data includes meteorological data with corresponding labels characterizing blade icing, temperature lower than the preset icing temperature and humidity higher than the preset icing humidity. The label characterizing blade icing indicates that blade icing occurs in the wind farm when the first type of icing meteorological data is collected. The preset icing temperature and the preset icing humidity can be used to determine whether ice has begun to form, whether ice continues to form, and whether ice has begun to melt. The preset icing temperature and the preset icing humidity can be set according to the scene, demand, experience, etc., and are not limited here. For example, the preset icing temperature can be 5°C, and the preset icing humidity can be 70%. The temperature is lower than the preset icing temperature and the humidity is higher than the preset icing humidity, indicating that the wind farm is in low temperature and high humidity meteorological conditions when the first type of icing meteorological data is collected. The first type of icing meteorological data may be meteorological data in the icing stage.

[0052] The second type of icing meteorological data includes corresponding tags characterizing blade icing, tags characterizing blade icing that continue to appear within the second preset time period before, and meteorological data with a temperature lower than the preset icing temperature and a humidity lower than or equal to the preset icing humidity. Tags characterizing blade icing indicate that blade icing occurs in the wind farm when the second type of icing meteorological data is collected. Tags characterizing blade icing that continue to appear within the second preset time period before the second type of icing meteorological data is collected indicate that blade icing has occurred in the wind farm before the second type of icing meteorological data is collected. The second preset time period can be set according to scenarios, requirements, experience, etc., and is not limited here. For example, the second preset time period can be 3 hours. The specific contents of the preset icing temperature and the preset icing humidity can be referred to the relevant descriptions in the above embodiments, and will not be repeated here. The temperature is lower than the preset icing temperature and the humidity is lower than or equal to the preset icing humidity, indicating that the wind farm is in low temperature and low humidity meteorological conditions when the second type of icing meteorological data is collected. The second type of icing meteorological data may be meteorological data in the continuous icing stage.

[0053] The third type of icing meteorological data includes meteorological data in which the corresponding label characterizes blade icing, the label characterizing blade icing continues to appear within the third preset time period before, and the temperature is higher than or equal to the preset icing temperature. The label characterizes blade icing, indicating that the wind farm has blade icing when collecting the third type of icing meteorological data. The label characterizing blade icing continues to appear within the third preset time period before collecting the third type of icing meteorological data, indicating that the wind farm has already had blade icing before collecting the third type of icing meteorological data. The third preset time period can be set according to the scene, demand, experience, etc., and is not limited here. For example, the third preset time period can be 3 hours. The specific contents of the preset icing temperature and the preset icing humidity can be referred to the relevant description in the above embodiments, which will not be repeated here. The temperature is higher than or equal to the preset icing temperature, indicating that the wind farm is in non-low temperature meteorological conditions when collecting the third type of icing meteorological data. The third type of icing meteorological data can be meteorological data in the ice melting stage.

[0054] The meteorological data during the historical icing period used to train the blade icing prediction model include meteorological data in the icing stage, i.e., the first type of icing meteorological data, meteorological data in the continuous icing stage, i.e., the second type of icing meteorological data, and meteorological data in the ice melting stage, i.e., the third type of icing meteorological data. This can enrich the types of historical sample data used to train the blade icing prediction model, so that the blade icing prediction model can learn richer and more complete content of the blade icing stage during the training process, thereby improving the accuracy and completeness of the trained blade icing prediction model.

[0055] In some examples, the blade icing prediction model may include a blade icing classification prediction model and / or a blade icing ratio prediction model. The blade icing classification prediction model is used to predict whether blade icing occurs in a wind farm. When the blade icing prediction model includes the blade icing classification prediction model, the blade icing prediction result may include blade icing or blade non-icing. Correspondingly, the historical sample data used to train the blade icing prediction model has a label that can characterize blade icing or blade non-icing. The blade icing ratio prediction model is used to predict the icing shutdown ratio of wind turbines in a wind farm. When the blade icing prediction model includes the blade icing ratio prediction model, the blade icing prediction result includes the icing shutdown ratio. Correspondingly, the historical sample data used to train the blade icing prediction model has a label that can characterize the icing shutdown ratio. The icing shutdown ratio includes the ratio of the number of wind turbines in the wind farm that have been shut down due to blade icing to the number of wind turbines in the wind farm. In some cases, the blade icing classification prediction model can be obtained through the blade icing ratio prediction model, and the icing shutdown ratio output by the blade icing ratio prediction model can be converted into a prediction result of blade icing or blade non-icing according to a preset ratio threshold, and used as the output of the blade icing classification prediction model. For example, if the icing shutdown ratio is less than or equal to the preset ratio threshold, the blade icing prediction result output by the blade icing classification prediction model indicates that the blade is not iced; if the icing shutdown ratio is greater than the preset ratio threshold, the blade icing prediction result output by the blade icing classification prediction model indicates that the blade is iced.

[0056] Implementing a blade icing prediction model through a blade icing classification prediction model and / or a blade icing ratio prediction model can make the blade icing prediction results of the blade icing prediction model more abundant and flexible in representing content, and more easily meet the needs of wind farm safety management.

[0057] Since the meteorological conditions of different wind farms are different, and the differences may be large, in order to avoid the adverse effects of the large differences in meteorological conditions of different wind farms on model training, the historical sample data of each wind farm may be normalized before the above step S106. Specifically, one wind farm may be selected as a standard wind farm from more than two wind farms; the historical sample data of the standard wind farm may be normalized to obtain a normalized proportional coefficient; and the historical sample data of wind farms other than the standard wind farm may be normalized using the normalized proportional coefficient. Correspondingly, the above step S106 may be specifically refined as follows: at least part of the normalized historical sample data with labels is used as a training data set for model training to obtain a blade icing prediction model.

[0058] The standard wind farm is a wind farm used as a reference. The historical sample data of the standard wind farm can be normalized so that the values ​​of the historical sample data of the standard wind farm are all within a set range. The scaling factor that makes the values ​​of the historical sample data of the standard wind farm within the set range is the normalized proportional coefficient. For example, a scaler MinMaxScalaer can be established, and the scaler MinMaxScalaer is used to multiply the values ​​of the historical sample data of the standard wind farm with the normalized proportional coefficient so that the maximum value of the multiplied historical sample data is 1 and the minimum value is 0. The values ​​of the historical sample data of other wind farms except the standard wind farm are multiplied by the normalized proportional coefficient to realize the normalization of the historical sample data of the wind farms other than the standard wind farm. The historical sample data obtained after multiplication is the historical sample data after normalization.

[0059] The normalized scale coefficient is obtained by normalizing the standard wind farm. The historical sample data of other wind farms except the standard wind farm are proportionally scaled using the normalized scale coefficient. On the one hand, it can reduce the adverse effects of the large differences in meteorological conditions between different wind farms on model training. On the other hand, it can also retain the differences between different wind farms. Using the normalized historical sample data for model training can improve the efficiency of model training and further improve the accuracy of the blade icing prediction model obtained through training.

[0060] In some embodiments, the number of wind turbines shut down due to icing in the preset area can be obtained based on the blade icing prediction results of multiple wind farms in the preset area, thereby facilitating the adoption of protection strategies for the wind farms in the preset area. Figure 4 A flowchart of a method for predicting wind farm blade icing provided by another embodiment of the present application, Figure 4 and Figure 1 The difference is that Figure 4 The wind farm blade icing prediction method shown also includes steps S107 to S109.

[0061] In step S107, the number of wind turbines in the wind farms in the preset area that are predicted to be shut down due to icing is calculated based on the blade icing prediction results of the wind farms in the preset area and the number of wind turbines in the wind farms in the preset area.

[0062] The blade icing prediction result includes the icing shutdown ratio. The specific content of the icing shutdown ratio can be found in the relevant descriptions in the above embodiments, and will not be repeated here. The preset area can be set according to the scene, demand, experience, etc. The preset area can be a certain area that needs to be monitored. For example, the preset area can be a province, a city, etc. The product of the number of wind turbines in the wind farm and the icing shutdown ratio in the blade icing prediction result corresponding to the wind farm is the number of wind turbines in the wind farm that have been shut down due to icing.

[0063] In step S108, the number of wind turbines shut down due to icing in the preset area is determined according to the predicted number of wind turbines shut down due to icing in the wind farm in the preset area.

[0064] The sum of the numbers of wind turbine generator sets shut down due to icing in each wind farm in the preset area is the number of wind turbine generator sets shut down due to icing in the preset area.

[0065] In some examples, the predicted ice shutdown capacity in the preset area can also be calculated based on the number of wind turbines that have been shut down due to icing in the preset area and the capacity of the wind turbines in the wind farm. The predicted ice shutdown capacity in the preset area is the sum of the capacities of the wind turbines that have been shut down due to icing in the preset area. The accuracy of the ice shutdown capacity in the preset area predicted by the wind farm blade icing prediction method in the embodiment of the present application is also high. For example, Figure 5 A schematic diagram of an example of a comparison between the actual icing shutdown capacity and the predicted icing shutdown capacity in a certain area provided in an embodiment of the present application, wherein the horizontal axis is time and the vertical axis is shutdown capacity in megawatts (MW). The actual icing shutdown capacity is the icing shutdown capacity determined by actual observation, and the predicted icing shutdown capacity is the icing shutdown capacity predicted by the wind farm blade icing prediction method provided in an embodiment of the present application, such as Figure 5 As shown, in the early icing stage, i.e., January 14 and January 15, the blades began to freeze, causing the shutdown capacity in the area to grow rapidly and reach a peak; in the mid-term continuous icing stage, i.e., January 16 and January 17, because the temperature, humidity and other meteorological conditions continued to meet the icing conditions, a high proportion of wind turbines in the area were in a shutdown state; in the later ice melting stage, i.e., January 18 to January 21, as the ice melted, the shutdown capacity in the area quickly decreased to zero; the wind farm blade icing prediction method provided in the embodiment of the present application performed well, and the actual icing shutdown capacity in the area was basically consistent with the predicted icing shutdown capacity, especially in the early icing stage and the mid-term continuous icing stage, the wind farm blade icing prediction method in the embodiment of the present application has a high prediction accuracy, which can provide effective decision-making guidance for wind farms and power grids, and ensure the safe and stable operation of wind farms and power grids.

[0066] In step S109, a wind turbine protection strategy corresponding to the number of wind turbines shut down due to icing in a preset area is executed.

[0067] The number of wind turbines shut down due to icing in a preset area can provide guidance for the control strategy of the wind farm in the preset area. The correspondence between the number of wind turbines shut down due to icing in the preset area and the unit protection strategy can be preset. After obtaining the predicted number of wind turbines shut down due to icing in the preset area, the corresponding unit protection strategy can be executed. The unit protection strategy can be used to maintain the safe and stable operation of the wind farm and the power grid electrically connected to the wind farm. For example, the unit protection strategy may include but is not limited to the processing and adjustment of the wind turbines in the wind farm. Other strategies that can maintain the safe and stable operation of the wind farm and the power grid electrically connected to the wind farm in the event of icing shutdown are also within the protection scope of the embodiments of the present application.

[0068] By predicting the number of wind turbines that will be shut down due to icing in a preset area, the icing shutdown situation in the preset area can be known in advance, so as to implement an adaptive unit protection strategy, effectively reduce the economic losses caused by blade icing, extend the service life of wind turbines, and ensure the safe operation of electronic systems including wind farms and power grids.

[0069] A second aspect of the present application provides a device for predicting icing on blades in a wind farm. Figure 6 A schematic diagram of the structure of a wind farm blade icing prediction device provided in an embodiment of the present application is shown in FIG. Figure 6 As shown, the wind farm blade icing prediction device 200 may include a data acquisition module 201 and a model processing module 202 .

[0070] The data acquisition module 201 may be used to acquire the predicted meteorological data of the wind farm within a preset time period in the future, where the predicted meteorological data includes temperature and humidity.

[0071] The model processing module 202 may be used to input the predicted meteorological data into a pre-trained blade icing prediction model, and the blade icing prediction model processes the predicted meteorological data to output a blade icing prediction result of the wind farm.

[0072] The blade icing prediction result is used to characterize the blade icing condition in the future preset time period. The blade icing prediction model is trained based on the historical sample data of the wind farm. The historical sample data includes meteorological data during the historical icing period and meteorological data during the historical non-icing period. The meteorological data includes temperature and humidity. The historical sample data has a label, which is used to characterize the blade icing condition corresponding to the meteorological data.

[0073] In some examples, the predicted meteorological data is consistent with the type of meteorological data, and the meteorological data further includes one or more of the following: wind speed, precipitation, cloud cover, air pressure, and dew point temperature.

[0074] In some examples, the historical sample data also includes meteorological data within a first preset time period before at least one data collection moment in the historical ice coverage time.

[0075] In an embodiment of the present application, the blade icing prediction model obtained by pre-training can be used to process the predicted meteorological data of the input wind farm in a preset time period in the future, so as to obtain the blade icing prediction result output by the blade icing prediction model. The blade icing prediction model is trained based on the historical sample data with labels of the wind farm, and the historical sample data includes the meteorological data within the historical icing time and the meteorological data within the historical non-icing time. The labels of the historical sample data can characterize the blade icing conditions corresponding to the meteorological data. The blade icing prediction model obtained by training based on the historical sample data with labels of the wind farm can accurately describe the nonlinear dynamic process of blade icing evolution, so that the blade icing conditions within the preset time period in the future can be accurately predicted in advance based on the predicted meteorological data within the preset time period in the future, and the blade icing can be detected in a timely and accurate manner, thereby reducing the safety risks of the wind farm.

[0076] In some embodiments, the wind farm blade icing prediction device 200 may further include a model training module.

[0077] The model training module can be used to: obtain historical sample data and the icing shutdown ratio corresponding to the meteorological data in the historical sample data, the icing shutdown ratio includes the ratio of the number of wind turbines shut down due to blade icing in the wind farm to the number of wind turbines in the wind farm; generate labels according to the icing shutdown ratio; establish a corresponding relationship between the label and the historical sample data; use at least part of the historical sample data with labels as a training data set to train a machine learning model to obtain a blade icing prediction model.

[0078] In some examples, the meteorological data during the historical icing period used to train the blade icing prediction model includes the first type of icing meteorological data, the second type of icing meteorological data, and the third type of icing meteorological data. Among them, the first type of icing meteorological data includes meteorological data with corresponding labels representing blade icing, temperature lower than the preset icing temperature, and humidity higher than the preset icing humidity. The second type of icing meteorological data includes meteorological data with corresponding labels representing blade icing, continuous appearance of labels representing blade icing within the previous second preset time period, temperature lower than the preset icing temperature, and humidity lower than or equal to the preset icing humidity. The third type of icing meteorological data includes meteorological data with corresponding labels representing blade icing, continuous appearance of labels representing blade icing within the previous third preset time period, and temperature higher than or equal to the preset icing temperature.

[0079] In some examples, the blade icing prediction model includes a blade icing classification prediction model and / or a blade icing ratio prediction model.

[0080] In the case where the blade icing prediction model includes a blade icing classification prediction model, the blade icing prediction result includes blade icing or blade non-icing, and the label represents blade icing or blade non-icing.

[0081] When the blade icing prediction model includes a blade icing ratio prediction model, the blade icing prediction result includes an icing shutdown ratio, and the label represents the icing shutdown ratio. The icing shutdown ratio includes the ratio of the number of wind turbines shut down due to blade icing in the wind farm to the number of wind turbines in the wind farm.

[0082] In some embodiments, the blade icing prediction result includes an icing shutdown ratio, and the icing shutdown ratio includes the ratio of the number of wind turbines that are shut down due to blade icing in the wind farm to the number of wind turbines in the wind farm. The wind farm blade icing prediction device 200 may also include a regional protection module. The regional protection module can be used to: calculate the number of wind turbines that are shut down due to icing in the wind farm predicted in the preset area according to the blade icing prediction result of the wind farm in the preset area and the number of wind turbines in the wind farm in the preset area; determine the number of wind turbines that are shut down due to icing in the preset area according to the number of wind turbines that are shut down due to icing in the wind farm predicted in the preset area; and execute a unit protection strategy corresponding to the number of wind turbines that are shut down due to icing in the preset area.

[0083] In some embodiments, the wind farm blade icing prediction device 200 may further include a normalization processing module. The normalization processing module may be used to: select one wind farm from more than two wind farms as a standard wind farm; perform normalization processing on the historical sample data of the standard wind farm and obtain a normalization scale coefficient; and perform normalization processing on the historical sample data of wind farms other than the standard wind farm using the normalization scale coefficient.

[0084] Correspondingly, the above-mentioned model training module can be specifically used to: use at least a part of the normalized historical sample data with labels as a training data set for model training.

[0085] The third aspect of the present application also provides a wind farm blade icing prediction device. Figure 7 This is a schematic diagram of the structure of a wind farm blade icing prediction device provided in one embodiment of the present application. Figure 7 As shown, the wind farm blade icing prediction device 300 includes a memory 301 , a processor 302 , and a computer program stored in the memory 301 and executable on the processor 302 .

[0086] In some examples, the processor 302 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0087] The memory 301 may include a read-only memory (ROM), a random access memory (RAM), a disk storage medium device, an optical storage medium device, a flash memory device, an electrical, optical or other physical / tangible memory storage device. Therefore, generally, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., a memory device) encoded with software including computer executable instructions, and when the software is executed (e.g., by one or more processors), it can be operated to perform the operations described with reference to the method for predicting ice coverage of wind farm blades according to the embodiment of the present application.

[0088] The processor 302 runs a computer program corresponding to the executable program code by reading the executable program code stored in the memory 301 , so as to implement the wind farm blade icing prediction method in the above embodiment.

[0089] In some examples, the wind farm blade icing prediction device 300 may further include a communication interface 303 and a bus 304. Figure 7 As shown, the memory 301, the processor 302, and the communication interface 303 are connected via a bus 304 and communicate with each other.

[0090] The communication interface 303 is mainly used to realize the communication between the modules, devices, units and / or equipment in the embodiment of the present application. The communication interface 303 can also be used to access input devices and / or output devices.

[0091] The bus 304 includes hardware, software, or both, coupling the components of the wind farm blade ice prediction device 300 to each other. For example, but not limitation, the bus 304 may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a Memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-E) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses or a combination of two or more of these. Where appropriate, the bus 304 may include one or more buses. Although embodiments of the present application describe and illustrate a particular bus, the present application contemplates any suitable bus or interconnect.

[0092] The fourth aspect of the present application also provides a computer-readable storage medium, on which a computer program instruction is stored. When the computer program instruction is executed by a processor, the wind farm blade icing prediction method in the above embodiment can be implemented, and the same technical effect can be achieved. To avoid repetition, it is not repeated here. Among them, the above-mentioned computer-readable storage medium may include a non-transitory computer-readable storage medium, such as a read-only memory (Read-Only Memory, referred to as ROM), a random access memory (Random Access Memory, referred to as RAM), a disk or an optical disk, etc., which is not limited here.

[0093] An embodiment of the present application provides a computer program product. When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the method for predicting icing on wind farm blades in the above embodiment and can achieve the same technical effect. To avoid repetition, it will not be described here.

[0094] It should be clear that each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. For device embodiments, equipment embodiments, computer-readable storage medium embodiments, and computer program product embodiments, the relevant parts can refer to the description part of the method embodiment. The present application is not limited to the specific steps and structures described above and shown in the figures. Those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application. In addition, for the sake of brevity, a detailed description of known method technologies is omitted here.

[0095] The above reference is according to the method of the embodiment of the present application, the flow chart of the device (system) and the computer program product and / or the block diagram described various aspects of the present application.It should be understood that each square box in the flow chart and / or the block diagram and the combination of each square box in the flow chart and / or the block diagram can be realized by computer program instructions.These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer or other programmable data processing device to produce a machine so that these instructions executed by the processor of the computer or other programmable data processing device enable the realization of the function / action specified in one or more square boxes of the flow chart and / or the block diagram.Such a processor can be but is not limited to a general-purpose processor, a special-purpose processor, a special application processor or a field programmable logic circuit.It can also be understood that each square box in the block diagram and / or the flow chart and the combination of the square boxes in the block diagram and / or the flow chart can also be realized by the dedicated hardware that performs the specified function or action, or can be realized by the combination of dedicated hardware and computer instructions.

[0096] Those skilled in the art should understand that the above embodiments are exemplary rather than restrictive. Different technical features appearing in different embodiments can be combined to achieve beneficial effects. Based on the study of the drawings, the specification and the claims, those skilled in the art should be able to understand and implement other variations of the disclosed embodiments. In the claims, the term "comprising" does not exclude other devices or steps; the quantifier "one" does not exclude multiple; the terms "first" and "second" are used to indicate names rather than to indicate any specific order. Any figure marks in the claims should not be understood as limiting the scope of protection. The functions of multiple parts appearing in the claims can be implemented by a single hardware or software module. The appearance of certain technical features in different dependent claims does not mean that these technical features cannot be combined to achieve beneficial effects.

Claims

1. A method for predicting blade icing in a wind farm, characterized in that: include: Acquire forecasted meteorological data of the wind farm within a preset time period in the future, wherein the forecasted meteorological data includes temperature and humidity; The predicted meteorological data is input into a pre-trained blade icing prediction model, and the blade icing prediction model processes the predicted meteorological data to output a blade icing prediction result of the wind farm, wherein the blade icing prediction result is used to characterize the blade icing condition in a preset time period in the future. Among them, the blade icing prediction model is trained based on historical sample data of the wind farm, the historical sample data includes meteorological data during historical icing time and meteorological data during historical non-icing time, the meteorological data includes temperature and humidity, and the historical sample data has a label, and the label is used to characterize the blade icing condition corresponding to the meteorological data.

2. The method according to claim 1, characterized in that Before inputting the predicted meteorological data into the pre-trained blade icing prediction model, the method further includes: Acquire the historical sample data and the icing shutdown ratio corresponding to the meteorological data in the historical sample data, wherein the icing shutdown ratio includes the ratio of the number of wind turbines that are shut down due to blade icing in the wind farm to the number of wind turbines in the wind farm; Generating the label according to the ice-covered shutdown ratio; Establishing a corresponding relationship between the label and the historical sample data; At least part of the historical sample data with the label is used as a training data set to perform machine learning model training to obtain the blade icing prediction model.

3. The method according to claim 1, characterized in that The historical sample data also includes meteorological data within a first preset time period before at least one data collection moment in the historical ice coverage time.

4. The method according to claim 1, characterized in that: The meteorological data during the historical icing period used to train the blade icing prediction model includes first-class icing meteorological data, second-class icing meteorological data and third-class icing meteorological data. Among them, the first type of ice-covered meteorological data includes meteorological data in which the corresponding label indicates blade ice-covered, the temperature is lower than the preset ice-covered temperature and the humidity is higher than the preset ice-covered humidity; the second type of ice-covered meteorological data includes meteorological data in which the corresponding label indicates blade ice-covered, the label indicating blade ice-covered continues to appear within the previous second preset time period, the temperature is lower than the preset ice-covered temperature and the humidity is lower than or equal to the preset ice-covered humidity; the third type of ice-covered meteorological data includes meteorological data in which the corresponding label indicates blade ice-covered, the label indicating blade ice-covered continues to appear within the previous third preset time period and the temperature is higher than or equal to the preset ice-covered temperature.

5. The method according to claim 1, characterized in that The blade icing prediction model includes a blade icing classification prediction model and / or a blade icing ratio prediction model; In the case where the blade icing prediction model includes the blade icing classification prediction model, the blade icing prediction result includes blade icing or blade non-icing, and the label represents blade icing or blade non-icing; In the case where the blade icing prediction model includes the blade icing ratio prediction model, the blade icing prediction result includes an icing shutdown ratio, the label represents the icing shutdown ratio, and the icing shutdown ratio includes the ratio of the number of wind turbines shut down due to blade icing in the wind farm to the number of wind turbines in the wind farm.

6. The method according to claim 1, characterized in that The blade icing prediction result includes an icing shutdown ratio, and the icing shutdown ratio includes a ratio of the number of wind turbine generator sets that are shut down due to blade icing in the wind farm to the number of wind turbine generator sets in the wind farm. After inputting the predicted meteorological data into a pre-trained blade icing prediction model, and processing the predicted meteorological data by the blade icing prediction model to output a blade icing prediction result of the wind farm, the method further includes: The number of wind turbines in the wind farms in the preset area that are predicted to be shut down due to icing is calculated according to the blade icing prediction result of the wind farms in the preset area and the number of wind turbines in the wind farms in the preset area; Determining the number of wind turbines shut down due to icing in the preset area according to the predicted number of wind turbines shut down due to icing in the wind farm in the preset area; A unit protection strategy corresponding to the number of wind turbine generator sets shut down due to icing in the preset area is executed.

7. The method according to claim 2, characterized in that Before using at least part of the historical sample data with the label as a training data set for model training, the method further includes: Among more than two wind farms, one wind farm is selected as a standard wind farm; Normalizing the historical sample data of the standard wind farm and obtaining a normalized proportional coefficient; Using the normalization proportional coefficient, normalizing the historical sample data of wind farms other than the standard wind farm; The using at least part of the historical sample data with the label as a training data set for model training comprises: At least part of the normalized historical sample data with the label is used as the training data set for model training.

8. The method according to any one of claims 1 to 7, characterized in that The forecasted meteorological data is consistent with the type of meteorological data, and the meteorological data also includes one or more of the following: Wind speed, precipitation, cloud cover, air pressure, dew point temperature.

9. A wind farm blade icing prediction device, characterized in that: include: A data acquisition module, used to acquire the predicted meteorological data of the wind farm within a preset time period in the future, wherein the predicted meteorological data includes temperature and humidity; A model processing module is used to input the predicted meteorological data into a pre-trained blade icing prediction model, and the blade icing prediction model processes the predicted meteorological data to output a blade icing prediction result of the wind farm, wherein the blade icing prediction result is used to characterize the blade icing situation in a preset time period in the future. Among them, the blade icing prediction model is trained based on historical sample data of the wind farm, the historical sample data includes meteorological data during historical icing time and meteorological data during historical non-icing time, the meteorological data includes temperature and humidity, and the historical sample data has a label, and the label is used to characterize the blade icing condition corresponding to the meteorological data.

10. A wind farm blade icing prediction device, characterized in that: include: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, the method for predicting wind farm blade icing according to any one of claims 1 to 8 is implemented.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by a processor, the method for predicting wind farm blade icing according to any one of claims 1 to 8 is implemented.

Images