Fan blade icing diagnosis method and system, wind generating set and electronic equipment

Abstract

The invention discloses a fan blade icing diagnosis method and system, a wind generating set and electronic equipment, and relates to the technical field of wind generating sets. According to the method, when the fan runs, the running parameters of the fan and the actual value of the output power are obtained, and then when the environment temperature is within the preset temperature range, the running parameters are input into a prediction model obtained through pre-training on the basis of the running parameters of the fan and the actual value of the output power of the fan corresponding to the running parameters; and obtaining a predicted value of the power output by the fan corresponding to the operation parameters, and determining that the blade is in an icing state when the predicted value is subtracted from an actual value, the obtained difference value exceeds a preset early warning value and the environment temperature of the environment where the fan is located is lower than a preset temperature threshold value. Therefore, the problems that in the prior art, when a fault diagnosis model is adopted for blade icing diagnosis, due to the fact that construction of the fault diagnosis model depends on the fault data label, the model construction workload is large, and the use cost is increased are solved.

Description

Technical Field

[0001] The present application relates to the technical field of wind turbines, and more specifically, to a method and system for diagnosing icing of wind turbine blades, a wind turbine generator set and electronic equipment. Background Art

[0002] Wind farms are generally located in high-altitude and high-humidity areas, which are prone to blade icing in winter, posing a threat to the safe operation of wind turbines. Therefore, timely detection of blade icing during wind turbine operation helps to timely adjust the wind turbine operating status or shut down the unit to ensure safe operation of the unit.

[0003] At present, there are mainly three methods for detecting abnormal icing of wind turbine blades:

[0004] 1. By deploying an ice detection device on the blade surface or inside the blade, when the sensor shows an abnormal signal of ice characteristics, it is determined that ice is present on the blade surface. This method is the most common practice at present. It is simple, direct and effective, but the cost is that it will bring hardware procurement and structural change costs, and deployment and debugging are inconvenient.

[0005] 2. Monitor the surface status of the blades by deploying a pan-tilt camera and using a visual algorithm. This method can monitor the ice and damage of the blades at the same time, but it is easily affected by the environment and also requires additional hardware equipment, making deployment and debugging inconvenient.

[0006] 3. The fault diagnosis model obtained by training the labeled fault data and normal data using machine learning algorithms outputs a fault discrimination signal when the wind turbine is running - ice or no ice. Although this method does not rely on hardware detection equipment, it is necessary to collect a large amount of normal data and ice abnormality data when building the fault diagnosis model, and mark the two types of data. When applied to different wind farms and different models, it will bring about a huge amount of data labeling work. Summary of the invention

[0007] In view of this, the embodiments of the present application are committed to providing a wind turbine blade icing diagnosis method, system, wind turbine generator set and electronic device to solve the problem that when a fault diagnosis model is used to diagnose blade icing in the prior art, the construction of the fault diagnosis model depends on the fault data label, resulting in a large workload for model construction and increased cost of use.

[0008] In a first aspect, the present invention provides a method for diagnosing icing of fan blades, comprising:

[0009] When the fan is running, obtaining the operating parameters of the fan and the actual value of the power output by the fan corresponding to the operating parameters;

[0010] Inputting the operating parameters into a trained prediction model to obtain a predicted value of the power output by the wind turbine corresponding to the operating parameters output by the prediction model;

[0011] Subtracting the predicted value from the actual value, and determining whether the obtained difference exceeds a preset warning value;

[0012] When it is determined that the difference exceeds the preset warning value and the ambient temperature of the environment in which the fan is located is lower than a preset temperature threshold, it is determined that the blades of the fan are in an icing state;

[0013] Wherein, the prediction model is trained when the ambient temperature is within a preset temperature range based on the operating parameters of the fan when it is in a target state and the actual value of the power output of the fan corresponding to the operating parameters, and is updated when the change in the ambient temperature is greater than the preset temperature change range. The target state includes: the blades are in an unfrozen state, and the fluctuation amplitude of the power output of the fan is less than a preset fluctuation threshold.

[0014] In a possible implementation manner, determining whether the obtained difference exceeds a preset warning value includes:

[0015] Inputting the difference into a residual analysis model to obtain a quantity to be evaluated output by the residual analysis model;

[0016] Determining whether the quantity to be evaluated exceeds the preset warning value;

[0017] The residual analysis model is a model obtained by a residual analysis method using the difference between the historical predicted value and the historical actual value of the power output by the wind turbine as a sample.

[0018] In a possible implementation, the preset warning value is determined by a control chart method.

[0019] In a possible implementation, the operating parameters include: wind speed, wind direction, and blade angle.

[0020] In a possible implementation, the prediction model is a BP neural network model.

[0021] In a possible implementation manner, obtaining the operating parameters of the wind turbine includes:

[0022] Obtaining target operating parameters, where the target operating parameters are operating parameters generated by the wind turbine within a preset time period;

[0023] The mean of the target operating parameters is calculated and normalized to obtain the operating parameters.

[0024] In a second aspect, the present invention provides a fan blade icing diagnosis system, comprising:

[0025] A collection unit, used for obtaining the operating parameters of the fan and the actual value of the power output of the fan corresponding to the operating parameters when the fan is running;

[0026] A processing unit, configured to input the operating parameters into a trained prediction model to obtain a predicted value of the power output by the wind turbine corresponding to the operating parameters output by the prediction model;

[0027] A comparison unit, used to make a difference between the predicted value and the actual value, and determine whether the obtained difference exceeds a preset warning value;

[0028] a diagnostic unit, configured to determine that the blades of the fan are in an icing state when it is determined that the difference exceeds the preset warning value and the ambient temperature of the environment in which the fan is located is lower than a preset temperature threshold;

[0029] Wherein, the prediction model is trained when the ambient temperature is within a preset temperature range based on the operating parameters of the fan when it is in a target state and the actual value of the power output of the fan corresponding to the operating parameters, and is updated when the change in the ambient temperature is greater than the preset temperature change range. The target state includes: the blades are in an unfrozen state, and the fluctuation amplitude of the power output of the fan is less than a preset fluctuation threshold.

[0030] In a possible implementation manner, the comparison unit is specifically configured to:

[0031] Inputting the difference into a residual analysis model to obtain a quantity to be evaluated output by the residual analysis model;

[0032] Determining whether the quantity to be evaluated exceeds the preset warning value;

[0033] The residual analysis model is a model obtained by a residual analysis method using the difference between the historical predicted value and the historical actual value of the power output by the wind turbine as a sample.

[0034] In a third aspect, the present invention provides a wind turbine generator set, comprising a wind turbine body and the wind turbine blade icing diagnosis system provided in the second aspect of the present invention.

[0035] In a fourth aspect, the present invention provides an electronic device, the electronic device comprising:

[0036] processor;

[0037] a memory for storing instructions executable by the processor;

[0038] The processor is used to execute the wind turbine blade icing diagnosis method provided in the first aspect of the present invention.

[0039] According to the icing diagnosis method for fan blades provided by the present invention, when the fan is running, the operating parameters of the fan and the actual value of the power output of the fan corresponding to the operating parameters are obtained, and then the operating parameters are input into the prediction model pre-trained based on the operating parameters when the fan is in the target state and the actual value of the power output of the fan corresponding to the operating parameters when the ambient temperature is within the preset temperature range, and the predicted value of the power output of the fan corresponding to the operating parameters is obtained, so as to make a difference between the predicted value and the actual value, and determine that the obtained difference exceeds the preset warning value, and the ambient temperature of the environment where the fan is located is lower than the preset temperature threshold, and determine that the blade of the fan is in an icing state. On the one hand, the diagnosis of whether the blade is frozen is realized by the prediction model constructed independently of the blade icing data label, which reduces the cost and improves the efficiency. On the other hand, by setting the prediction model to be updated when the change of the ambient temperature is greater than the preset temperature change range, the online adjustment of the prediction model according to the environmental change is realized, thereby reducing the model error caused by the temperature change, and then improving the accuracy of the diagnosis result. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0041] Figure 1 Shown is a flow chart of a method for diagnosing icing of fan blades provided by an embodiment of the present invention.

[0042] Figure 2 Shown is a structural diagram of the BP neural network structure provided by an embodiment of the present invention.

[0043] Figure 3 The figure shows a specific flow chart of an example of applying the fan blade icing diagnosis method provided by an embodiment of the present invention to diagnose fan blade icing.

[0044] Figure 4 Shown is a structural diagram of a fan blade icing diagnosis system provided by an embodiment of the present invention.

[0045] Figure 5 Shown is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0046] Unless otherwise defined, the technical terms or scientific terms used in the embodiments of this specification shall have the common meanings understood by persons with ordinary skills in the field to which this specification belongs. The words "first", "second" and similar words used in the embodiments of this specification do not indicate any order, quantity or importance, but are only used to avoid confusion of constituent elements.

[0047] Unless the context requires otherwise, throughout the specification, "plurality" means "at least two", and "including" is interpreted as an open, inclusive meaning, that is, "including, but not limited to". In the description of the specification, the terms "one embodiment", "some embodiments", "exemplary embodiments", "examples", "specific examples" or "some examples" are intended to indicate that a particular feature, structure, material or characteristic associated with the embodiment or example is included in at least one embodiment or example of the specification. The schematic representation of the above terms does not necessarily refer to the same embodiment or example.

[0048] The following will be combined with the drawings in the embodiments of this specification to clearly and completely describe the technical solutions in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this specification.

[0049] It is understandable that in order to avoid the high cost and deployment and debugging problems caused by deploying hardware equipment to monitor the icing condition of wind turbine blades, the current method is to use machine learning algorithms to train labeled fault data and normal data to obtain a fault diagnosis model. That is, the fault diagnosis model is used to output a fault discrimination signal when the wind turbine is running - ice or not. However, this method requires collecting a large amount of normal data and ice abnormality data when building a fault diagnosis model, and marking the two types of data. When applied to different wind farms and different models, it will bring a huge workload of data labeling.

[0050] The present invention aims to solve the above-mentioned problem by applying a prediction model which is trained based on the operating parameters of the fan when it is in a target state and the actual values ​​of the power output by the fan corresponding to the operating parameters, and is updated when the change in ambient temperature is greater than a preset temperature change range, to predict the power output by the fan, and then determine whether the fan blades are frozen based on the actual power output by the fan and the predicted power in combination with the ambient temperature, thereby realizing the diagnosis of whether the blades are frozen by using the prediction model constructed through the operating parameters of the fan when it is operating normally, reducing the workload of model construction, and at the same time, realizing online updating of the model, thereby improving the accuracy of the diagnosis results and ensuring rapid deployment between different wind farms and models.

[0051] Furthermore, the wind turbine blade icing diagnosis method provided by the present invention is executed on an electronic device, for example: the electronic device can be a controller of a wind turbine generator set, or can be an intelligent terminal such as a smart phone or a laptop computer remotely connected to the controller of the wind turbine generator set.

[0052] Based on the above content, an embodiment of the present invention provides a method for diagnosing icing of wind turbine blades. Figure 1 , Figure 1 : is a flow chart of a fan blade icing diagnosis method provided by an embodiment of the present invention. The flow chart of the fan blade icing diagnosis method provided by this embodiment may include:

[0053] S100 . When the fan is running, obtain operating parameters of the fan and actual values ​​of power output by the fan corresponding to the operating parameters.

[0054] Specifically, the operating parameters of the wind turbine are related to the operation of the wind turbine, such as wind speed, wind direction, blade angle, bearing temperature, hydraulic oil pressure, cabin temperature, etc.

[0055] In some possible embodiments, operating parameters such as wind speed and wind direction may be detected by sensors arranged on the wind turbine.

[0056] Furthermore, the purpose of obtaining the operating parameters of the wind turbine in this embodiment is to obtain the predicted value of the wind turbine output power corresponding to the operating parameters through the trained prediction model based on the operating parameters of the wind turbine. Therefore, the operating parameters obtained in this embodiment are operating parameters associated with the wind turbine output power.

[0057] Based on this, in an optional embodiment, the operating parameters include: wind speed, wind direction, and blade angle.

[0058] It is understandable that by preprocessing the operating parameters, the accuracy of the acquired data can be improved, thereby improving the accuracy of the diagnosis result.

[0059] Based on this, in an optional embodiment, obtaining the operating parameters of the fan includes:

[0060] Get the target running parameters.

[0061] The target operating parameters are operating parameters generated by the fan within a preset time period.

[0062] The mean of the target operating parameters is calculated and normalized to obtain the operating parameters.

[0063] Specifically, by obtaining the operating parameters generated by the fan within a preset time period as the target operating parameters, then calculating the mean of the target operating parameters, and then performing normalization processing, on the one hand, the accuracy of the operating parameters obtained due to data loss or jumps is avoided, and on the other hand, the amount of data that needs to be processed by the prediction model is reduced, thereby improving the diagnostic efficiency.

[0064] In some possible embodiments, the preset duration can be flexibly set according to actual needs, for example, to 45 seconds, 1 minute, etc., which is not specifically limited here.

[0065] S110, inputting the operating parameters into the trained prediction model to obtain the predicted value of the power output by the wind turbine corresponding to the operating parameters output by the prediction model.

[0066] Among them, the prediction model is trained based on the operating parameters of the fan when it is in the target state and the actual value of the power output of the fan corresponding to the operating parameters when the ambient temperature is within the preset temperature range, and is updated when the change in ambient temperature is greater than the preset temperature change range. The target states include: the blades are in an unfrozen state, and the fluctuation amplitude of the power output of the fan is less than the preset fluctuation threshold.

[0067] Specifically, the preset temperature range is usually set at around 2-10 degrees Celsius, thereby ensuring that the prediction model is trained based on the operating parameters and output power of the fan when the blades are not frozen, and on the other hand, ensuring that the operating parameters and output power of the trained prediction model are within the temperature range where there is a risk of blade icing, thereby improving the accuracy of the prediction model, that is, ensuring the accuracy of the icing diagnosis results.

[0068] More specifically, by setting a preset temperature change range, it is ensured that the prediction model can be updated in real time when the ambient temperature fluctuates greatly, thereby achieving adaptive adjustment of the prediction model as the environment changes, and further ensuring that it can be quickly deployed between different wind farms and models without the need for offline training of the model in advance.

[0069] In some possible embodiments, taking into account factors such as heat exchange during fan operation, when the ambient temperature is lower than 5 degrees Celsius, the blades are very likely to freeze, but when it is higher than 15 degrees Celsius, blade freezing generally does not occur. Therefore, the preset temperature range is set to [5,15], and the preset temperature change range is ±5 degrees Celsius. That is, when the ambient temperature is in the range of 5-15 degrees Celsius, the prediction model is obtained by training the actual values ​​of the operating parameters and output power of the fan during normal operation, and when the change in ambient temperature exceeds 5 degrees Celsius compared to the change in ambient temperature when the prediction model was obtained by the last training, and the ambient temperature is not lower than 5 degrees Celsius or higher than 15 degrees Celsius, the operating parameters and output power of the fan are obtained again to re-train the prediction model.

[0070] In a specific embodiment, the preset temperature variation range is set to ±4 degrees Celsius. If the current ambient temperature is monitored to be 5 degrees Celsius, and the ambient temperature was 10 degrees Celsius when the prediction model was last updated, that is, the change in ambient temperature compared to the last time the prediction model was updated exceeds the preset temperature variation range, then the prediction model update will be triggered, that is, the actual values ​​of the operating parameters of the fan and the power output of the fan will be re-obtained, and then a new round of online modeling process will be started based on the re-obtained operating parameters and actual values ​​of the output power.

[0071] Furthermore, by setting the operating parameters of the prediction model obtained through training and the actual values ​​of the power output of the wind turbine corresponding to the operating parameters to be the operating parameters and the actual power when the wind turbine is in the target state, the operating parameters of the training prediction model are the operating parameters when the blades are in an unfrozen state, and the actual value of the output power is the power when the fluctuation amplitude of the power output of the wind turbine is less than a preset fluctuation threshold, that is, by ensuring the reliability of the data used to train the prediction model (screening out the grid-connected power generation data when the wind turbine is not frozen, and excluding the data in the power instability stage during the start-up and shutdown process), the accuracy of the prediction model is improved.

[0072] Furthermore, by adopting any feasible neural network model structure and training data, the converged model parameters, i.e., the prediction model, can be obtained through optimization algorithms such as the gradient descent algorithm.

[0073] In an optional embodiment, the prediction model is a BP neural network model.

[0074] Specifically, the BP (Back Propagation) neural network structure can be as follows Figure 2 As shown in the figure, there are 3 input layer nodes (x1, x2, x3), representing the three features of wind speed, wind direction, and blade angle respectively. The hidden layer has at least 5 nodes to achieve a higher model prediction accuracy. The number of output layer nodes is 1 (y), representing the predicted output power of the wind turbine.

[0075] Furthermore, the input layer and the hidden layer are connected via a σ function, and the hidden layer and the output layer are connected via a linear function.

[0076] More specifically, in the online modeling process, firstly, the wind speed, wind direction, blade angle, and actual output power data of the wind turbine are obtained within a preset time (for example, 1 minute), and the data during the start-up and shutdown process and fault process are excluded. Then, the data mean is calculated and normalized, and then the prediction model is trained. Finally, when the modeling error or the number of iterations meets the set convergence error or the maximum number of iterations, the training is exited and the latest updated parameters are used as the final prediction model.

[0077] S120: Subtract the predicted value from the actual value, and determine whether the obtained difference exceeds a preset warning value.

[0078] Specifically, by setting a reasonable preset warning value, and then comparing the difference between the predicted value of the wind turbine output power and the actual value of the wind turbine output power with the preset warning value, it can be determined whether there is a risk of icing on the wind turbine blades.

[0079] In an optional embodiment, determining whether the obtained difference exceeds a preset warning value includes:

[0080] Input the difference into the residual analysis model to obtain the quantity to be evaluated output by the residual analysis model;

[0081] Determine whether the quantity to be evaluated exceeds the preset warning value;

[0082] The residual analysis model is a model obtained by the residual analysis method, which takes the difference between the historical predicted value and the historical actual value of the power output of the wind turbine as a sample.

[0083] In this embodiment, a residual analysis model is used to determine whether the difference between the predicted value and the actual value of the fan output power exceeds a preset warning value. Compared with simply comparing the difference with the set warning value, false alarms caused by model errors can be effectively avoided, thereby improving the reliability of the fan blade icing diagnosis method provided in this embodiment.

[0084] In an optional embodiment, the preset warning value is determined by a control chart method.

[0085] Specifically, the residual vector of the training set samples is obtained by subtracting the power prediction value obtained by the prediction model from the actual output power of the wind turbine, and the mean and variance of the residual of the training set samples are calculated. Then, the upper and lower limits of the control chart warning are designed using the control chart method, that is, they comply with the following expressions respectively:

[0086]

[0087] Among them, μ0 is the mean of the residual of the training set samples, L is the width of the control line, σ is the variance of the residual of the training set samples, and λ is the weight of the residual of the current training set samples in the historical statistics.

[0088] In some possible embodiments, during the prediction process, a residual analysis model can be used to calculate the residual value to be evaluated by using an exponentially weighted moving average method, a threshold method, or an equal-weighted moving average method, and then determine whether it exceeds a preset warning value based on the relationship between the residual value to be evaluated and the upper and lower limits of the control chart.

[0089] S130: When it is determined that the difference exceeds the preset warning value and the ambient temperature of the environment in which the fan is located is lower than the preset temperature threshold, it is determined that the blades of the fan are in an icing state.

[0090] It is understandable that when a fan fails, the difference between the actual value of the fan output power and the fan output power predicted by the prediction model may exceed the preset warning value.

[0091] Based on this, in this embodiment, when it is determined that the difference between the actual value and the predicted value of the fan output power exceeds the preset warning value, it is further determined whether the ambient temperature of the fan is lower than the preset temperature threshold, and only when the ambient temperature is lower than the preset temperature threshold, it is determined that the fan blades are in a frozen state, thereby effectively improving the accuracy and reliability of the fan blade ice diagnosis method provided in this embodiment to diagnose whether the blades are frozen.

[0092] It should be noted that although the temperature at which water freezes into ice is 0 degrees Celsius, considering that fans are generally located in high-altitude and high-humidity areas, and there is a huge heat exchange with the external environment during operation, when the ambient temperature is higher than 0 degrees Celsius, the blades may also freeze. Therefore, the preset temperature threshold is generally set to 4-5 degrees Celsius. That is, taking the preset environmental threshold of 5 degrees Celsius as an example, when it is determined that the difference exceeds the preset warning value and the ambient temperature is lower than 5 degrees Celsius, the blades are judged to be in a frozen state. If it is determined that the difference does not exceed the preset warning value or the ambient temperature is higher than 5 degrees Celsius, the blades are judged to be in an unfrozen state.

[0093] Figure 3 A specific flow chart of an example of applying the fan blade icing diagnosis method provided by an embodiment of the present invention to diagnose fan blade icing. Figure 3 As shown, when the fan blade icing diagnosis method provided by the embodiment of the present invention is used to perform blade icing diagnosis, a total of 5 process structures are included, from step 1 to step 5. Among them, when the prediction model is deployed for the first time in the unit, since there is no trained prediction model, it is necessary to execute steps 1 to step 3 to obtain the prediction model through online training. When the trained prediction model is obtained and the change of the ambient temperature does not exceed the preset temperature change range, steps 1, step 4 and step 5 are executed to give a normal warning of the blade icing state.

[0094] In summary, the wind turbine blade icing diagnosis method provided by the embodiment of the present invention performs blade icing state diagnosis by using the prediction model obtained by training the actual value of the power output of the wind turbine corresponding to the operating parameter based on the target state that the wind turbine blade is in an unfrozen state and the fluctuation amplitude of the power output of the wind turbine is less than the preset fluctuation threshold when the ambient temperature is within the preset temperature range, so that the data required for the training of the prediction model are the operating parameters when the wind turbine blade is not frozen, and do not rely on the fault data label, so that data collection and processing can be completed online without marking the data in advance, reducing the workload of building the prediction model, thereby reducing the cost. By setting the prediction model to be updated according to the change of ambient temperature, the model error caused by temperature change is reduced, thereby not only improving the accuracy of detection, but also enabling the prediction model to be quickly deployed between different wind farms and models without the need to train the model offline in advance, thereby improving the convenience of application. By using the residual analysis model when determining whether the difference between the predicted value and the actual value of the wind turbine output power exceeds the preset warning value, false alarms caused by model errors are further avoided, thereby improving the reliability of the prediction model.

[0095] A fan blade icing diagnosis system provided in an embodiment of the present invention is introduced below. The fan blade icing diagnosis system described below can be considered as a module architecture for implementing a fan blade icing diagnosis method provided in an embodiment of the present invention; the content described below can be cross-referenced with the above.

[0096] See also Figure 4 , Figure 4 : is a structural block diagram of a fan blade icing diagnosis system provided by an embodiment of the present invention, and the system may include:

[0097] The acquisition unit 10 is used to obtain the operating parameters of the fan and the actual value of the power output of the fan corresponding to the operating parameters when the fan is running;

[0098] The processing unit 20 is used to input the operating parameters into the trained prediction model to obtain the predicted value of the power output of the wind turbine corresponding to the operating parameters output by the prediction model;

[0099] A comparison unit 30 is used to make a difference between the predicted value and the actual value, and determine whether the obtained difference exceeds a preset warning value;

[0100] The diagnosis unit 40 is used to determine that the blades of the fan are in an icing state when it is determined that the difference exceeds a preset warning value and the ambient temperature of the environment in which the fan is located is lower than a preset temperature threshold;

[0101] Among them, the prediction model is trained based on the operating parameters of the fan when it is in the target state and the actual value of the power output of the fan corresponding to the operating parameters when the ambient temperature is within the preset temperature range, and is updated when the change in ambient temperature is greater than the preset temperature change range. The target states include: the blades are in an unfrozen state, and the fluctuation amplitude of the power output of the fan is less than the preset fluctuation threshold.

[0102] In an optional embodiment, the comparison unit 30 is specifically configured to:

[0103] Input the difference into the residual analysis model to obtain the quantity to be evaluated output by the residual analysis model;

[0104] Determine whether the quantity to be evaluated exceeds the preset warning value;

[0105] The residual analysis model is a model obtained by the residual analysis method, which takes the difference between the historical predicted value and the historical actual value of the power output of the wind turbine as a sample.

[0106] Optionally, the preset warning value is determined by a control chart method.

[0107] Optionally, the operating parameters include: wind speed, wind direction, and blade angle.

[0108] Optionally, the prediction model is a BP neural network model.

[0109] Optionally, the collection unit 10 is specifically used for:

[0110] Obtaining target operating parameters, where the target operating parameters are operating parameters generated by the fan within a preset time period;

[0111] The mean of the target operating parameters is calculated and normalized to obtain the operating parameters.

[0112] Optionally, an embodiment of the present invention further provides a wind turbine generator set, comprising a wind turbine body and a wind turbine blade icing diagnosis system as provided in any of the aforementioned embodiments.

[0113] Below, reference Figure 5 To describe the electronic device provided by the embodiment of the present application, the electronic device provided by the embodiment may include: at least one processor 100, at least one communication interface 200, at least one memory 300 and at least one communication bus 400;

[0114] In the embodiment of the present invention, the number of the processor 100, the communication interface 200, the memory 300, and the communication bus 400 is at least one, and the processor 100, the communication interface 200, and the memory 300 communicate with each other through the communication bus 400; obviously, Figure 5The communication connections shown for the processor 100, the communication interface 200, the memory 300, and the communication bus 400 are merely optional;

[0115] Optionally, the communication interface 200 may be an interface of a communication module, such as an interface of a GSM module; the processor 100 may be a central processing unit CPU, or an application specific integrated circuit ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present invention.

[0116] The memory 300 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0117] The processor 100 is specifically used to execute the application program in the memory to implement the steps of the above-mentioned wind turbine blade icing diagnosis method.

[0118] The basic principles of the present application are described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, effects, etc. mentioned in the present application are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. are required by each embodiment of the present application. In addition, the specific details disclosed above are only for the purpose of illustration and ease of understanding, not for limitation, and the above details do not limit the present application to being implemented by adopting the above specific details.

[0119] The block diagrams of the devices, apparatuses, equipment, and systems involved in this application are only illustrative examples and are not intended to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagram. As will be appreciated by those skilled in the art, these devices, apparatuses, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including", "comprising", "having", etc. are open words, referring to "including but not limited to", and can be used interchangeably with them. The words "or" and "and" used here refer to the words "and / or" and can be used interchangeably with them, unless the context clearly indicates otherwise. The words "such as" used here refer to the phrase "such as but not limited to", and can be used interchangeably with them.

[0120] It should also be noted that in the apparatus, device and method of the present application, each component or each step can be decomposed and / or recombined. Such decomposition and / or recombination should be regarded as equivalent solutions of the present application.

[0121] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present application. Therefore, the present application is not intended to be limited to the aspects shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

[0122] It should be understood that the qualifiers "first", "second", "third", "fourth", "fifth" and "sixth" used in the description of the embodiments of the present application are only used to more clearly explain the technical solutions and cannot be used to limit the scope of protection of the present application.

[0123] The above description has been given for the purpose of illustration and description. In addition, this description is not intended to limit the embodiments of the present application to the forms disclosed herein. Although multiple example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions and sub-combinations thereof.

Claims

1. A method for diagnosing icing of fan blades, characterized in that: include: When the fan is running, obtaining the operating parameters of the fan and the actual value of the power output by the fan corresponding to the operating parameters; Inputting the operating parameters into a trained prediction model to obtain a predicted value of the power output by the wind turbine corresponding to the operating parameters output by the prediction model; Subtracting the predicted value from the actual value, and determining whether the obtained difference exceeds a preset warning value; When it is determined that the difference exceeds the preset warning value and the ambient temperature of the environment in which the fan is located is lower than a preset temperature threshold, it is determined that the blades of the fan are in an icing state; Wherein, the prediction model is trained when the ambient temperature is within a preset temperature range based on the operating parameters of the fan when it is in a target state and the actual value of the power output of the fan corresponding to the operating parameters, and is updated when the change in the ambient temperature is greater than the preset temperature change range. The target state includes: the blades are in an unfrozen state, and the fluctuation amplitude of the power output of the fan is less than a preset fluctuation threshold.

2. The method according to claim 1, characterized in that The determining whether the obtained difference exceeds a preset warning value includes: Inputting the difference into a residual analysis model to obtain a quantity to be evaluated output by the residual analysis model; Determining whether the quantity to be evaluated exceeds the preset warning value; The residual analysis model is a model obtained by a residual analysis method using the difference between the historical predicted value and the historical actual value of the power output by the wind turbine as a sample.

3. The method according to claim 2, characterized in that The preset warning value is determined by a control chart method.

4. The method according to claim 1, characterized in that: The operating parameters include: wind speed, wind direction, and blade angle.

5. The method according to claim 1, characterized in that The prediction model is a BP neural network model.

6. The method according to claim 1, characterized in that The obtaining of the operating parameters of the fan includes: Obtaining target operating parameters, where the target operating parameters are operating parameters generated by the wind turbine within a preset time period; The mean of the target operating parameters is calculated and normalized to obtain the operating parameters.

7. A fan blade icing diagnosis system, characterized in that: include: A collection unit, used for obtaining the operating parameters of the fan and the actual value of the power output of the fan corresponding to the operating parameters when the fan is running; A processing unit, configured to input the operating parameters into a trained prediction model to obtain a predicted value of the power output by the wind turbine corresponding to the operating parameters output by the prediction model; A comparison unit, used to make a difference between the predicted value and the actual value, and determine whether the obtained difference exceeds a preset warning value; a diagnostic unit, configured to determine that the blades of the fan are in an icing state when it is determined that the difference exceeds the preset warning value and the ambient temperature of the environment in which the fan is located is lower than a preset temperature threshold; Wherein, the prediction model is trained when the ambient temperature is within a preset temperature range based on the operating parameters of the fan when it is in a target state and the actual value of the power output of the fan corresponding to the operating parameters, and is updated when the change in the ambient temperature is greater than the preset temperature change range. The target state includes: the blades are in an unfrozen state, and the fluctuation amplitude of the power output of the fan is less than a preset fluctuation threshold.

8. The system according to claim 7, characterized in that The comparison unit is specifically used for: Inputting the difference into a residual analysis model to obtain a quantity to be evaluated output by the residual analysis model; Determining whether the quantity to be evaluated exceeds the preset warning value; The residual analysis model is a model obtained by a residual analysis method using the difference between the historical predicted value and the historical actual value of the power output by the wind turbine as a sample.

9. A wind turbine generator set, characterized in that: It comprises a fan body and a fan blade icing diagnosis system as claimed in claim 7 or 8.

10. An electronic device, characterized in that: The electronic device comprises: processor; a memory for storing instructions executable by the processor; The processor is used to execute the method according to any one of claims 1 to 6.

Images