Wind power plant wake flow control method and system

Abstract

The invention provides a wind power plant wake flow control method and system, and the method comprises the steps: 1, building a wind power plant wake flow control model; 2, collecting input wind direction data through a wind direction sensor preset on a wind power plant fan; 3, according to the input wind direction data corresponding to different wind power plant fan positions, obtaining input factors; and 4, inputting the input factor into the wind power plant wake flow control model, and carrying out linkage control on the wind power plant fan. According to the wind power plant wake flow control method and system, the wind power plant wake flow control model for automatically controlling the yaw angle of the fan according to the wind direction data detected by the wind direction sensor to enable the wind wheel to be always in the windward state is established, and the input factors are constructed through the input wind direction data collected in real time; the input factors are input into the wind power plant wake flow control model to realize linkage control of the wind power plant fans, yaw control is performed according to the real-time induced wind direction, wake flow interference of upstream fans is avoided, and the productivity efficiency of the wind power plant is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of wind farms, and particularly relates to a wind farm wake control method and system. Background Art

[0002] During wind power generation, the blades of a wind turbine will capture part of the wind energy when the wind blows through the wind turbine and convert it into electrical energy. In this process, the wind speed and direction will be affected by the wind turbine, thus forming a wake region downstream of the wind turbine. The wake will reduce the wind speed of downstream wind turbines, reduce their energy capture ability, and thus reduce the power generation efficiency. Therefore, measures need to be taken to reduce or optimize the wake effect between wind turbines, so as to improve the power generation efficiency and reliability of the entire wind farm.

[0003] The invention patent with the application number: CN202310800293.0 discloses a wake steering control method and related equipment for an offshore wind farm. The method includes: establishing a field group digital twin model based on the layout of the wind farm fleet, where the field group digital twin model includes the location layout information of each wind turbine in the wind farm and the state parameter information of each wind turbine; generating a wind turbine target control matrix according to the main wind direction monitoring data and the field group digital twin model, where the wind turbine target control matrix is the control matrix corresponding to the theoretical minimum wake negative impact of the wind farm calculated based on the main wind direction monitoring data; determining the wind turbine target yaw control matrix based on the wind turbine target control matrix and the state parameter information of each wind turbine.

[0004] The above-mentioned prior art controls the wind turbines by calculating the control matrix corresponding to the theoretical minimum wake negative impact of the wind farm based on the main wind direction monitoring data. The control effect highly depends on the calculation accuracy of the control matrix corresponding to the theoretical minimum wake negative impact. When the calculation accuracy is poor, it is easy to cause inaccurate control problems, thereby reducing the power generation capacity of the wind farm.

[0005] In view of this, there is an urgent need for a wind farm wake control method and system to at least solve the above deficiencies. Summary of the Invention

[0006] One of the objectives of the present invention is to provide a wind farm wake control method, which establishes a wind farm wake control model that automatically controls the yaw angle of the wind turbine according to the wind direction data detected by the wind direction sensor so that the wind wheel is always in the upwind state. Input factors are constructed based on the real-time collected input wind direction data, and the input factors are input into the wind farm wake control model to realize the coordinated control of the wind farm wind turbines. Yaw control is performed according to the real-time sensed wind direction, avoiding the wake interference of upstream wind turbines and improving the power generation efficiency of the wind farm.

[0007] A wind farm wake control method provided by an embodiment of the present invention includes:

[0008] Step 1: Establish a wake control model for the wind farm;

[0009] Step 2: Collect input wind direction data through the preset wind direction sensors on the wind turbines in the wind farm;

[0010] Step 3: Obtain input factors according to the input wind direction data corresponding to different wind turbine positions in the wind farm;

[0011] Step 4: Input the input factors into the wake control model of the wind farm to perform coordinated control of the wind turbines in the wind farm.

[0012] Preferably, establishing the wake control model of the wind farm includes:

[0013] Based on the preset construction rules of the wind turbine position description matrix, construct the wind turbine position description matrix according to the wind turbine positions in the wind farm;

[0014] Obtain the yaw control command library for the wind turbines at different wind turbine positions in the wind farm, and the yaw control command library includes: one-to-one corresponding wind turbine position wind direction data and yaw control commands;

[0015] Determine the wind turbine position wind direction description factors and the yaw control command set according to the wind turbine position wind direction data and the wind turbine position description matrix;

[0016] Use the wind turbine position wind direction description factors as the input of the preset neural network control model and the yaw control command set as the output of the neural network control model to train the wake control model of the wind farm.

[0017] Preferably, collecting the input wind direction data through the preset wind direction sensors on the wind turbines in the wind farm includes:

[0018] Obtain the installation position of the wind direction sensor relative to the wind turbine;

[0019] Obtain the three-dimensional model of the wind turbine;

[0020] According to the installation position and the three-dimensional model of the wind turbine, simulate the wind direction sensor to detect different wind directions, and obtain the simulated detection wind direction data corresponding to different wind directions;

[0021] Obtain the pre-corrected wind direction data sensed by the wind direction sensor;

[0022] Determine the simulated wind direction data corresponding to the simulated detection wind direction data when the simulated detection wind direction data is consistent with the pre-corrected wind direction data, and use the simulated wind direction data as the input wind direction data.

[0023] Preferably, obtaining the input factors according to the input wind direction data corresponding to different wind turbine positions in the wind farm includes:

[0024] Determine the input factors according to the input wind direction data corresponding to different wind turbine positions in the wind farm and the wind turbine position description matrix.

[0025] A method for controlling the wake of a wind farm provided by an embodiment of the present invention further includes:

[0026] Obtain historical production records, and perform abnormal detection of wind turbines according to the historical production records and the real-time production capacity of the wind farm.

[0027] Preferably, performing abnormal detection of wind turbines according to the historical production records and the real-time production capacity of the wind farm includes:

[0028] Train a production capacity prediction model according to the historical production records;

[0029] Obtain real-time meteorological wind data;

[0030] Obtain real-time predicted production capacity according to the real-time meteorological wind data and the production capacity prediction model;

[0031] Obtain the real-time production capacity of the wind farm;

[0032] If the production capacity difference between the real-time production capacity of the wind farm and the real-time predicted production capacity is greater than or equal to a preset production capacity difference threshold, perform abnormal detection of the wind turbines.

[0033] Preferably, if the production capacity difference between the real-time production capacity of the wind farm and the real-time predicted production capacity is greater than or equal to a preset production capacity difference threshold, performing abnormal detection of the wind turbines includes:

[0034] Determine the production capacity description factor for predicting the position of the wind turbine according to the real-time meteorological wind data and the production capacity prediction model;

[0035] Based on the same construction rule of the production capacity description factor for predicting the position of the wind turbine, determine the real-time production capacity description factor for the position of the wind turbine according to the real-time production capacity of the wind farm;

[0036] Determine the production capacity numerical difference at the factor position according to the production capacity description factor for predicting the position of the wind turbine and the real-time production capacity description factor for the position of the wind turbine, and use the factor position where the production capacity numerical difference at the factor position is greater than or equal to the preset difference threshold as the first target factor position;

[0037] Locate the position of the abnormal detection wind turbine according to the first target factor position, and send the positioning information to the wind turbine maintenance personnel.

[0038] A method for controlling the wake of a wind farm provided by an embodiment of the present invention further includes:

[0039] When the wind turbine maintenance personnel respond to the abnormal detection task, determine the shutdown wind turbine according to the first target factor position. After the shutdown wind turbine shuts down, notify the wind turbine maintenance personnel to enter the site for detection.

[0040] Preferably, determining the shutdown wind turbine according to the first target factor position includes:

[0041] Determine the second target factor position corresponding to the first target factor position in the input factors;

[0042] Obtain the first position relationship between the second target factor position and the input factor position in the input factors;

[0043] Obtain the input factor value of the input factor position;

[0044] Determine the input factor position to be calculated according to the first position relationship and the input factor value;

[0045] Determine the second position relationship between the second target factor position and the input factor position to be calculated;

[0046] Obtain the wind force parameter of the candidate fan corresponding to the input factor position to be calculated;

[0047] Based on a preset wind field simulation model, calculate the arriving wind force data at the abnormal detection fan position according to the wind force parameter of the candidate fan and the second position relationship;

[0048] Obtain the maintenance specification wind force data;

[0049] If the arriving wind force data meets the maintenance specification wind force data, use the fan corresponding to the abnormal detection fan position as the shutdown fan;

[0050] If the arriving wind force data does not meet the maintenance specification wind force data, obtain the shutdown impact value of the candidate fan;

[0051] Perform a simulated shutdown operation on the candidate fans in ascending order of the shutdown impact value, and recalculate the arriving wind force data until the recalculated arriving wind force data meets the maintenance specification wind force data;

[0052] Use all the fans with simulated shutdown and the fan corresponding to the abnormal detection fan position together as the shutdown fans.

[0053] Preferably, obtaining the shutdown impact value of the candidate fan includes:

[0054] Obtain the power supply side information of the power supply side of the candidate fan, and the power supply side information includes: power supply type, power supply side service type, and power supply side support degree;

[0055] Determine the type weight corresponding to the power supply side service type according to the preset power supply side service type - type weight comparison library;

[0056] Multiply the type weight of each power supply side service type by the corresponding power supply side support degree and sum them to obtain the target value;

[0057] If the power supply type is indirect power supply, obtain the ratio of the average maintenance duration to the remaining indirect power supply duration;

[0058] If the ratio is less than 1, the corresponding ratio is taken as the shutdown impact value;

[0059] If the ratio is greater than 1 or the power supply type is direct power supply, the target value is taken as the shutdown impact value.

[0060] A wake control system for a wind farm provided by an embodiment of the present invention includes:

[0061] A wake control model establishment subsystem for establishing a wake control model for a wind farm;

[0062] An input wind direction data acquisition subsystem for acquiring input wind direction data through a preset wind direction sensor on a wind farm fan;

[0063] An input factor acquisition subsystem for acquiring input factors according to input wind direction data corresponding to different wind farm fan positions;

[0064] A fan control subsystem for inputting the input factors into the wake control model of the wind farm to perform coordinated control of the wind farm fans.

[0065] The beneficial effects of the present invention are:

[0066] The present invention establishes a wake control model for a wind farm that automatically controls the yaw angle of the fan according to the wind direction data detected by the wind direction sensor so that the wind wheel is always in the upwind state. The input factors are constructed based on the input wind direction data collected in real time, and the input factors are input into the wake control model of the wind farm to achieve coordinated control of the wind farm fans. Yaw control is performed according to the real-time sensed wind direction, avoiding the wake interference of the upstream fans and improving the power generation efficiency of the wind farm.

[0067] Other features and advantages of the present invention will be described in the following specification, and, in part, will be obvious from the specification, or will be understood by practicing the present invention. The objectives and other advantages of the present invention can be achieved and obtained by the structures specifically pointed out in this application document.

[0068] The technical solutions of the present invention will be further described in detail below through the drawings and embodiments. Description of the Drawings

[0069] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation to the present invention. In the drawings:

[0070] Figure 1 is a schematic diagram of a wake control method for a wind farm in an embodiment of the present invention;

[0071] Figure 2 is a schematic diagram of a wake control system for a wind farm in an embodiment of the present invention. Detailed Embodiments

[0072] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only for the purpose of illustrating and explaining the present invention, and are not intended to limit the present invention.

[0073] An embodiment of the present invention provides a wind farm wake control method, as Figure 1 shown, including:

[0074] Step 1: Establish a wind farm wake control model; wherein, the wind farm wake control model is an AI control model that automatically controls the yaw angle of the wind turbine according to the wind direction data detected by the wind direction sensor so that the wind wheel is always in the upwind state;

[0075] Step 2: Collect input wind direction data through a preset wind direction sensor on the wind farm wind turbine; wherein, the input wind direction data is: the wind direction information sensed by the wind direction sensor, such as: 45°, northeast wind;

[0076] Step 3: Obtain input factors according to the input wind direction data corresponding to different wind farm wind turbine positions; wherein, the input factors are matrices, and each matrix element position represents a wind turbine position, and the element value at the matrix element position is the input wind direction data collected by the wind direction sensor on the wind turbine at the corresponding wind turbine position, and the corresponding relationship between the matrix element position and the wind turbine position is preset manually;

[0077] Step 4: Input the input factors into the wind farm wake control model to perform the coordinated control of the wind farm wind turbines.

[0078] The working principle and beneficial effects of the above technical solution are as follows:

[0079] The present invention establishes a wind farm wake control model that automatically controls the yaw angle of the wind turbine according to the wind direction data detected by the wind direction sensor so that the wind wheel is always in the upwind state, constructs input factors through the real-time collected input wind direction data, inputs the input factors into the wind farm wake control model to realize the coordinated control of the wind farm wind turbines, and performs yaw control according to the real-time sensed wind direction, avoiding the wake interference of the upstream wind turbines and improving the production efficiency of the wind farm.

[0080] In one embodiment, establishing a wind farm wake control model includes:

[0081] Based on the preset construction rule of the fan position description matrix, construct the fan position description matrix according to the fan positions in the wind farm; wherein, the construction rule of the fan position description matrix is set manually, specifically: the relationship between the fan positions in the wind farm and the matrix element positions of the fan position description matrix, and the matrix element values at the matrix element positions, for example: when there is a relative position relationship between the fan positions in the wind farm and the matrix element positions of the fan position description matrix, the matrix element value at the corresponding matrix element position of the fan position description matrix is 1, otherwise it is 0;

[0082] Obtain the yaw control command library of the fans at the fan positions in different wind farms, and the yaw control command library includes: the wind direction data of the fan positions in one-to-one correspondence and the yaw control commands, and the yaw control command library is constructed according to the yaw control records with the historical power generation efficiency reaching the specified power generation efficiency standard;

[0083] Determine the fan position wind direction description factor and the yaw control command set according to the fan position wind direction data and the fan position description matrix; wherein, the fan position wind direction description factor is: the matrix obtained by replacing the matrix element values at the matrix element positions with matrix element value of 1 in the fan position description matrix with the wind direction data of the fan positions associated with the corresponding matrix element positions;

[0084] Use the fan position wind direction description factor as the input of the preset neural network control model and the yaw control command set as the output of the neural network control model to train the wind farm wake control model. Among them, the preset neural network control model is: an AI control model combining a CNN neural network model and a PID control model.

[0085] The working principle and beneficial effects of the above technical solution are as follows:

[0086] The present invention introduces the construction rule of the fan position description matrix, constructs the fan position description matrix according to the fan positions in the wind farm; constructs a yaw control command library including the wind direction data of the fan positions in one-to-one correspondence and the yaw control commands according to the yaw control records with the historical power generation efficiency reaching the specified power generation efficiency standard; replaces the matrix element values at the matrix element positions with matrix element value of 1 in the fan position description matrix with the wind direction data of the fan positions associated with the corresponding matrix element positions to obtain the fan position wind direction description factor, determines the fan position wind direction description factor and the yaw control command set according to the replaced wind direction data of the fan positions in the fan position wind direction description factor and their corresponding yaw control commands, and uses them as training samples to train the AI control model combining the CNN neural network model and the PID control model, obtains the trained wind farm wake control model, and the model can directly and intelligently control the yaw angle according to the sensed wind direction data at each fan position in the future, which is more user-friendly.

[0087] In one embodiment, input wind direction data is collected through a preset wind direction sensor on a wind farm fan, including:

[0088] Obtain the installation position of the wind direction sensor relative to the fan; wherein, the installation position is: the position of the wind direction sensor relative to the fan, such as: the top of the fan's tower barrel or the hub of the fan;

[0089] Obtain the three-dimensional model of the fan; wherein, the three-dimensional model of the fan is: the three-dimensional point cloud model of the fan;

[0090] According to the installation position and the three-dimensional model of the fan, simulate the wind direction sensor to detect different wind directions, and obtain the simulated detection wind direction data corresponding to different wind directions; wherein, the simulated detection wind direction data is: for different simulated wind directions, the simulated induction data of the simulated wind direction sensor;

[0091] Obtain the pre-corrected wind direction data sensed by the wind direction sensor; wherein, the pre-corrected wind direction data is: the wind direction data directly read from the actual wind direction sensor;

[0092] Determine the simulated wind direction data corresponding to the simulated detection wind direction data when the simulated detection wind direction data and the pre-corrected wind direction data are consistent, and use the simulated wind direction data as the input wind direction data.

[0093] The working principle and beneficial effects of the above technical solution are:

[0094] The installation position of the wind speed sensor on the fan also greatly affects the detection accuracy of the wind direction. According to the installation position of the wind direction sensor relative to the fan and the three-dimensional model of the fan, this invention simulates the wind direction sensor to detect different wind directions to obtain the simulated detection wind direction data corresponding to different wind directions; matches the pre-corrected wind direction data directly sensed by the wind direction sensor with the simulated detection wind direction data, and determines the simulated wind direction data corresponding to the simulated detection wind direction data when they are consistent as the input wind direction data, improving the accuracy and reliability of wind direction measurement.

[0095] In one embodiment, according to the input wind direction data corresponding to different wind farm fan positions, input factors are obtained, including:

[0096] Determine the input factors according to the input wind direction data corresponding to different wind farm fan positions and the fan position description matrix. Among them, the construction rule of the input factors is the same as that of the fan position wind direction description factors.

[0097] The working principle and beneficial effects of the above technical solution are:

[0098] This invention matrixizes the input wind direction data corresponding to different wind farm fan positions, improving the standardization of input wind direction description and further enhancing the subsequent control accuracy.

[0099] An embodiment of the present invention provides a wake control method for a wind farm, further including:

[0100] Obtain historical production records, and perform abnormal detection of wind turbines according to the historical production records and the real-time production capacity of the wind farm. Among them, the historical production record is: the power generation records of individual wind turbines corresponding to the wind power data of different entities in history (the wind power data of the entity includes: the wind speed of the entity and the wind direction of the entity); the real-time production capacity of the wind farm is: the power generation power of the wind turbines monitored in real time.

[0101] The working principle and beneficial effects of the above technical solution are:

[0102] The present invention introduces historical production records, extracts wind power data similar to the current situation in the historical production records, compares the historical power generation situation with the current power generation situation, and judges the abnormality of the wind turbine, which is more reasonable.

[0103] In one embodiment, performing abnormal detection of wind turbines according to the historical production records and the real-time production capacity of the wind farm includes:

[0104] Train a production capacity prediction model according to the historical production records; among them, the production capacity prediction model is: an AI model for predicting the power generation power of wind turbines according to real-time wind power data. During training, the historical wind power data is used as the model input, and the power generation power of each wind turbine in history is used as the model output;

[0105] Obtain real-time meteorological wind power data; among them, the real-time meteorological wind power data is: the real-time wind direction and wind speed;

[0106] Obtain real-time predicted production capacity according to the real-time meteorological wind power data and the production capacity prediction model;

[0107] Obtain the real-time production capacity of the wind farm; among them, the real-time production capacity of the wind farm is: the total production capacity of the wind farm monitored in real time, such as: the total power generation power of the units;

[0108] If the production capacity difference between the real-time production capacity of the wind farm and the real-time predicted production capacity is greater than or equal to a preset production capacity difference threshold, perform abnormal detection of the wind turbine. Among them, the preset production capacity difference threshold is set manually in advance.

[0109] The working principle and beneficial effects of the above technical solution are:

[0110] The present invention introduces historical production records to train a production capacity prediction model, then inputs the real-time meteorological wind power data obtained in real time into the production capacity prediction model to obtain real-time predicted production capacity, and compares the real-time predicted production capacity with the real-time production capacity of the wind farm actually measured. When the power generation efficiency of the wind turbine is quite different from the past under the same wind power data situation, it is determined that the wind turbine needs to be abnormally detected, which improves the suitability of the abnormal detection.

[0111] In one embodiment, if the production capacity difference between the real-time production capacity and the real-time predicted production capacity of the wind farm is greater than or equal to a preset production capacity difference threshold, fan anomaly detection is performed, including:

[0112] According to the real-time meteorological wind data and the production capacity prediction model, determine the fan position predicted production capacity description factor; wherein, the fan position predicted production capacity description factor is a matrix obtained by replacing the matrix element values at the matrix element positions with a value of 1 in the fan position description matrix with the fan predicted power generation power of the fan at the corresponding matrix element position;

[0113] Based on the same construction rule of the fan position predicted production capacity description factor, determine the fan position real-time production capacity description factor according to the real-time production capacity of the wind farm; wherein, the fan position real-time production capacity description factor is a matrix obtained by replacing the matrix element values at the matrix element positions with a value of 1 in the fan position description matrix with the measured fan power generation power of the fan at the corresponding matrix element position;

[0114] According to the fan position predicted production capacity description factor and the fan position real-time production capacity description factor, determine the factor position production capacity numerical difference, and take the factor position where the factor position production capacity numerical difference is greater than or equal to the preset difference threshold as the first target factor position; wherein, the factor position production capacity numerical difference is the difference between the matrix element values at the same matrix element position in the fan position predicted production capacity description factor and the fan position real-time production capacity description factor, and the preset difference threshold is set manually in advance;

[0115] According to the first target factor position, locate the fan position for anomaly detection, and send the location information to the fan maintenance personnel. When locating the fan position for anomaly detection according to the first target factor position, it is determined according to the matrix element position corresponding to the first target factor position and the fan position description matrix construction rule.

[0116] The working principle and beneficial effects of the above technical solution are as follows:

[0117] The present invention constructs a fan position predicted production capacity description factor and a fan position real-time production capacity description factor, quantifies the first target factor position where the factor position production capacity numerical difference is greater than the difference threshold, and can directly locate the fan position for anomaly detection according to the first target factor position, improving the system positioning speed and enhancing the timeliness of anomaly monitoring.

[0118] An embodiment of the present invention provides a wind farm wake control method, which further includes:

[0119] When the fan maintenance personnel respond to the anomaly detection task, determine the shutdown fan according to the first target factor position, and notify the fan maintenance personnel to enter the site for detection after the shutdown fan shuts down.

[0120] The working principle and beneficial effects of the above technical solution are as follows:

[0121] When maintaining an abnormal fan, in order to ensure the safety of maintenance personnel, it is necessary to stop the fans in the operation area. However, shutting down all fans will delay the power generation progress of the wind farm, and if only some fans are shut down, the risk during maintenance will increase sharply. Therefore, the fans to be shut down are determined according to the position adaptability of the first target factor. After the fans to be shut down are stopped, the fan maintenance personnel are notified to enter the site for inspection, which improves the safety of personnel.

[0122] In one embodiment, determining the fans to be shut down according to the first target factor position includes:

[0123] Determine the second target factor position corresponding to the first target factor position in the input factor; wherein, the second target factor position is: the matrix element position in the input factor when the matrix element positions corresponding to the input factor and the first target factor position are the same. For example: if the first target factor position is the 2nd row and the 3rd column, the second target factor position is: the 2nd row and the 3rd column of the input factor;

[0124] Obtain the first position relationship between the second target factor position and the input factor positions in the input factor; wherein, the first position relationship is: the relative direction of the fan positions represented by the second target factor position and the other input factor positions in the input factor;

[0125] Obtain the input factor value of the input factor position; wherein, the input factor position is: the wind direction data of the fan position corresponding to the input factor position;

[0126] Determine the input factor position to be calculated according to the first position relationship and the input factor value; wherein, the input factor position to be calculated is: the input factor position where the fan wake will flow to the abnormal detection fan position determined according to the first position relationship and the input factor value, and the fan wake data is obtained by simulating according to the wake model, the first position relationship and the input factor value;

[0127] Determine the second position relationship between the second target factor position and the input factor position to be calculated; wherein, the second position relationship is: the relative direction and distance of the fan positions represented by the second target factor position and the input factor position to be calculated;

[0128] Obtain the wind power parameters of the candidate fans corresponding to the input factor position to be calculated; wherein, the wind power parameters include: the wind direction and wind speed of the wake of the candidate fans, obtained according to the simulation results of the wake model;

[0129] Based on the preset wind farm simulation model, calculate the arriving wind power data at the abnormal detection fan position according to the wind power parameters of the candidate fans and the second position relationship; wherein, the wind farm simulation model is: the wind farm simulation mechanism model;

[0130] Obtain the maintenance specification wind power data; among them, the maintenance specification wind power data is the requirement of the wind power data allowed for maintenance, for example: the wind speed is not higher than 8 m / s;

[0131] If the arriving wind power data meets the maintenance specification wind power data, the fan corresponding to the abnormal detection fan position will be used as the shutdown fan;

[0132] If the arriving wind power data does not meet the maintenance specification wind power data, obtain the shutdown impact value of the candidate fan; among them, the shutdown impact value represents the impact degree of the shutdown of the candidate fan on the corresponding power supply end, and the larger the shutdown impact value, the greater the corresponding impact degree;

[0133] Perform simulated shutdown operations on the candidate fans in ascending order of the shutdown impact value, and recalculate the arriving wind power data until the recalculated arriving wind power data meets the maintenance specification wind power data;

[0134] Take all the fans with simulated shutdown and the fans corresponding to the abnormal detection fan positions together as the shutdown fans.

[0135] The working principle and beneficial effects of the above technical solution are as follows:

[0136] The present invention determines the position of the second target factor in the input factor according to the matrix element position of the fan with abnormal power generation, obtains the first position relationship between the position of the second target factor and the fan positions represented by other input factor positions in the input factor and the wind direction data of the input factor positions, and determines the input factor positions (input factor positions to be calculated) where the fan wake will flow to the abnormal detection fan position; determines the relative direction and distance between the positions of the fans represented by the second target factor position and the input factor positions to be calculated. In addition, the wind power parameters of the candidate fans are introduced, and based on the wind field simulation mechanism model, according to the wind power parameters of the candidate fans and the second position relationship, wind field simulation is carried out to obtain the arriving wind power data at the abnormal detection fan position; obtain the maintenance specification wind power data allowed for maintenance. If the wind power conditions in the corresponding area have met the maintenance requirements after the fan at the abnormal detection fan position is shut down, there is no need to perform shutdown operations on other fans, and normal operation can be maintained; if the arriving wind power data does not meet the maintenance specification wind power data, introduce the shutdown impact value of the candidate fan, perform simulated shutdown operations on the candidate fans in ascending order of the shutdown impact value, and recalculate the arriving wind power data based on the wind field simulation model every time a fan is simulated to shut down until the recalculated arriving wind power data meets the maintenance specification wind power data. Record all the simulated shutdown fans at this time, and take the simulated shutdown fans and the fans corresponding to the abnormal detection fan positions together as the shutdown fans. While ensuring the safety of maintenance personnel, the impact on the power supply end is greatly reduced, and the shutdown plan is more reasonable.

[0137] In one embodiment, obtaining the shutdown impact value of the candidate fan includes:

[0138] Obtain the power supply side information of the candidate fan on the power supply side. The power supply side information includes: power supply type, power supply side service type, and power supply side support degree. Among them, the power supply side is the supply end of the power generation capacity of the candidate fan. The power supply type includes direct power supply (the candidate fan is directly connected to the power consumption side) and indirect power supply (the candidate fan first stores energy and then is connected to the power consumption side). The power supply side service type is the type of power supply side services, such as: residential side, school side, industrial production side, etc. The power supply side support degree represents the proportion of the electric energy provided by the candidate fan in the total power consumed on the corresponding power supply side.

[0139] According to the preset power supply side service type - type weight comparison library, determine the type weight corresponding to the power supply side service type. Among them, the preset power supply side service type - type weight comparison library is preset manually. For example, when the power supply side service type is the residential side, the type weight is 0.3; when the power supply side service type is the industrial production side, the type weight is 0.5.

[0140] Multiply the type weight of each power supply side service type by the corresponding power supply side support degree and then sum them up to obtain the target value.

[0141] If the power supply type is indirect power supply, obtain the ratio of the average maintenance duration to the remaining indirect power supply duration. Among them, the average maintenance duration is the average time spent by maintenance personnel for a maintenance operation in history. The remaining indirect power supply duration is the time length from the current time to the power supply time of the energy storage battery of the candidate fan's energy storage closest to the current time.

[0142] If the ratio is less than 1, use the corresponding ratio as the shutdown impact value.

[0143] If the ratio is greater than 1 or the power supply type is direct power supply, use the target value as the shutdown impact value.

[0144] The working principle and beneficial effects of the above technical solution are as follows:

[0145] When selecting a shutdown fan in the present invention, considering the impact of shutdown, the fan with less impact is preferentially considered for shutdown. Specifically, when obtaining the shutdown impact value characterizing the shutdown impact degree of the fan to be selected, the power supply type, the power supply side service type, and the power supply side support degree of the fan to be selected are obtained, the power supply side service type - type weight comparison library is introduced to determine the service type weight, and the product of the type weight of each power supply side service type and the corresponding power supply side support degree is multiplied and summed to obtain the target value; the power supply is divided into direct power supply and indirect power supply. Therefore, the average maintenance duration and the remaining indirect power supply duration are obtained. If the ratio is less than 1, it indicates that there is enough time for personnel maintenance, and the longer the remaining indirect power supply duration, the less it will delay the subsequent power supply. Therefore, the ratio of the average maintenance duration to the remaining indirect power supply duration is used as the shutdown impact value. Otherwise, the target value is directly used as the shutdown impact value, and the acquisition of the shutdown impact value is more reasonable and accurate.

[0146] An embodiment of the present invention provides a wake control system for a wind farm, as Figure 2 shown, including:

[0147] A wake control model establishment subsystem 1 for establishing a wake control model for the wind farm;

[0148] An input wind direction data acquisition subsystem 2 for acquiring input wind direction data through a preset wind direction sensor on a wind farm fan;

[0149] An input factor acquisition subsystem 3 for acquiring input factors according to the input wind direction data corresponding to the positions of different wind farm fans;

[0150] A fan control subsystem 4 for inputting the input factors into the wake control model of the wind farm to perform linkage control of the wind farm fans.

[0151] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these changes and modifications.

Claims

1. A method for controlling a wind farm wake, characterized in that: include: Step 1: Establish a wind farm wake control model; Step 2: Collect input wind direction data through the wind direction sensor preset on the wind turbine of the wind farm; Step 3: Obtain input factors based on input wind direction data corresponding to wind turbine locations in different wind farms; Step 4: Input the input factors into the wind farm wake control model to perform linkage control of wind farm wind turbines.

2. A wind farm wake control method according to claim 1, characterized in that: Establish a wind farm wake control model, including: Based on the preset wind turbine location description matrix construction rules, a wind turbine location description matrix is ​​constructed according to the wind farm wind turbine locations; Obtaining a yaw control command library of wind turbines at different wind turbine positions in wind farms, wherein the yaw control command library includes: one-to-one corresponding wind turbine position wind direction data and yaw control commands; Determine the wind direction description factor of the wind turbine position and the yaw control command set according to the wind turbine position wind direction data and the wind turbine position description matrix; The wind turbine position and wind direction description factors are used as the input of the preset neural network control model, and the yaw control command set is used as the output of the neural network control model to train the wind farm wake control model.

3. A wind farm wake control method according to claim 1, characterized in that: The wind direction data is collected through the preset wind direction sensors on the wind turbines in the wind farm, including: Get the installation position of the wind direction sensor relative to the wind turbine; Obtain a three-dimensional model of the wind turbine; According to the installation position and the three-dimensional model of the wind turbine, the wind direction sensor is simulated to detect different wind directions, and the simulated detection wind direction data corresponding to different wind directions are obtained; Acquire pre-corrected wind direction data sensed by a wind direction sensor; Determine the simulated wind direction data corresponding to the simulated detection wind direction data when the simulated detection wind direction data and the pre-correction wind direction data are consistent, and use the simulated wind direction data as input wind direction data.

4. A wind farm wake control method according to claim 1, characterized in that: According to the input wind direction data corresponding to the wind turbine positions in different wind farms, the input factors are obtained, including: The input factors are determined based on the input wind direction data corresponding to the wind turbine locations in different wind farms and the wind turbine location description matrix.

5. A wind farm wake control method according to claim 1, characterized in that: Also includes: Obtain historical capacity records and perform wind turbine anomaly detection based on historical capacity records and the real-time capacity of the wind farm.

6. A wind farm wake control method according to claim 5, characterized in that: Based on historical production capacity records and real-time production capacity of wind farms, wind turbine anomaly detection is performed, including: Train the capacity prediction model based on historical capacity records; Get real-time weather wind data; Obtain real-time predicted capacity based on real-time meteorological wind data and capacity prediction model; Obtain real-time production capacity of wind farms; If the capacity difference between the real-time capacity of the wind farm and the real-time predicted capacity is greater than or equal to a preset capacity difference threshold, wind turbine abnormality detection is performed.

7. A method for controlling the wake of a wind farm according to claim 6, characterized in that: If the capacity difference between the real-time capacity of the wind farm and the real-time predicted capacity is greater than or equal to the preset capacity difference threshold, wind turbine abnormality detection is performed, including: Determine the predicted capacity description factor of the wind turbine location based on real-time meteorological wind data and capacity prediction model; Based on the same construction rules of wind turbine location prediction capacity description factors, the real-time capacity description factors of wind turbine locations are determined according to the real-time capacity of wind farms; Determine the capacity value difference of the factor position according to the predicted capacity description factor of the wind turbine position and the real-time capacity description factor of the wind turbine position, and take the factor position whose capacity value difference of the factor position is greater than or equal to a preset difference threshold as the first target factor position; According to the position of the first target factor, the position of the abnormal detection fan is located, and the positioning information is sent to the fan maintenance personnel.

8. A wind farm wake control method according to claim 7, characterized in that: Also includes: When the fan maintenance personnel respond to the abnormal detection task, the shutdown fan is determined according to the position of the first target factor. After the shutdown fan is shut down, the fan maintenance personnel is notified to enter the site for inspection.

9. A wind farm wake control method according to claim 8, characterized in that: According to the position of the first target factor, the shutdown fan is determined, including: Determining a second target factor position in the input factor corresponding to the first target factor position; Obtaining a first position relationship between a second target factor position and an input factor position in the input factors; Get the input factor value at the input factor position; Determining the position of the input factor to be calculated according to the first position relationship and the input factor value; Determine a second positional relationship between a second target factor position and a position of an input factor to be calculated; Obtain the wind force parameters of the selected wind turbine corresponding to the position of the input factor to be calculated; Based on the preset wind field simulation model, according to the wind parameters of the selected wind turbine and the second position relationship, the arrival wind data at the abnormal detection wind turbine position is calculated; Obtain maintenance specification wind data; If the wind speed data reaches the standard of maintenance, the fan corresponding to the abnormal detection fan position will be used as the shutdown fan; If the wind power data reaches the specified level, the shutdown impact value of the wind turbine to be selected is obtained; Perform simulated shutdown operations on the selected wind turbines in the order of shutdown impact values ​​from small to large, and recalculate the arrival wind data until the recalculated arrival wind data meets the maintenance standard wind data; All fans that are simulated to be shut down and fans corresponding to the abnormal detection fan positions are collectively regarded as shut down fans.

10. A wind farm wake control system, characterized in that: include: The wake control model establishment subsystem is used to establish the wind farm wake control model; The input wind direction data acquisition subsystem is used to collect input wind direction data through the wind direction sensors preset on the wind turbines in the wind farm; An input factor acquisition subsystem is used to obtain input factors according to input wind direction data corresponding to wind turbine positions in different wind farms; The wind turbine control subsystem is used to input the input factors into the wind farm wake control model to perform linkage control of the wind farm wind turbines.

Images