Wind power climbing prediction system in extreme weather

Abstract

The invention discloses a wind power climbing prediction system in extreme weather, relates to the technical field of wind power climbing prediction, and solves the technical problem of low prediction precision of wind power climbing in extreme weather such as low temperature and icing in the prior art. According to the invention, environment data and wind power data of a target area are collected in real time; training an artificial intelligence model based on the historical environment data to obtain a wind power prediction model; and predicting the environmental data based on the wind power prediction model to obtain a prediction result, and constructing a correction model based on the wind power data, correcting the wind power output power in the prediction result based on the correction model to obtain target wind power output power, and performing early warning prompt by judging whether the difference value between the target wind power output power and the wind power output power at the initial moment is greater than a preset wind power climbing threshold value or not. The stability of wind power output power is ensured, and the technical problem is solved.

Description

Technical Field

[0001] The invention belongs to the field of wind power ramp prediction, in particular to a wind power ramp prediction system in extreme weather. Background Art

[0002] Wind power ramping refers to the phenomenon of rapid increase or decrease in wind power output within a short time scale during wind power generation. When wind power increases (climbs) rapidly, if the power grid cannot adjust the output of other power sources in time, it may cause overload. On the contrary, when wind power decreases (declines) rapidly, if there is not enough spare capacity to make up for the missing power, it may cause the power grid frequency to decrease, and in severe cases even cause power outages.

[0003] The prior art (the invention application with the publication number CN117578404A) discloses a method for predicting the short-term power ramp of wind power, which relates to the field of wind power prediction. The method comprises the following steps: obtaining and preprocessing prediction data, the prediction data including the historical actual power of wind power, numerical weather forecast data, the theoretical power data of the wind tower and the wind turbine, and the startup capacity data of the wind turbine; inputting the preprocessed prediction data into the CNN-LSTM-AM power prediction model for training, and obtaining the power prediction sequence data based on the trained CNN-LSTM-AM power prediction model; through accurate wind power prediction, identification and evaluation of ramp events, and maintenance of model accuracy, effectively planning and managing the operation of wind farms, making the best use of wind energy resources, and improving power generation efficiency; through identification and intensity evaluation of ramp events, wind energy fluctuations can be predicted and grid stability can be improved. However, the prior art only considers the prediction of wind power ramp under normal weather and environmental conditions, and does not consider the influence of extreme weather, such as low temperature and icing, on wind power ramp, resulting in the problem of low prediction accuracy of wind power ramp under extreme weather conditions.

[0004] Therefore, the present invention proposes a wind power ramp prediction system for extreme weather to solve the above problems. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a wind power ramp prediction system for extreme weather, which is used to solve the technical problem that the prior art only considers the prediction of wind power ramp under normal weather and environmental conditions, but does not consider the impact of extreme weather, such as low temperature and icing, on wind power ramp, resulting in low prediction accuracy of wind power ramp under extreme weather conditions.

[0006] To achieve the above-mentioned object, a first aspect of the present invention provides a wind power ramp prediction system for extreme weather, comprising: a data acquisition module, a data prediction module and a data early warning module;

[0007] Data acquisition module: used to obtain environmental data and wind power data of the target area;

[0008] Data prediction module: trains an artificial intelligence model based on historical environmental data to obtain a wind power prediction model; and,

[0009] Predict environmental data based on the wind power prediction model to obtain prediction results;

[0010] Data warning module: corrects the prediction results based on wind power data to obtain the target wind power output power; and

[0011] Based on the target wind power output power and the preset wind power ramp threshold, determine whether to send an early warning message; if yes, send the early warning message to the client of the wind farm manager; if not, continue to monitor and judge.

[0012] Preferably, the data prediction module is in communication and / or electrically connected with the data acquisition module and the data early warning module respectively.

[0013] Preferably, the obtaining of environmental data and wind power data of the target area includes:

[0014] The environmental data and wind power data of the target area are collected in real time through data acquisition equipment; the target area refers to a wind farm in low temperature and icy weather; the environmental data include: ambient temperature, ambient humidity, wind speed, wind direction, air density, atmospheric pressure and wind power output power; the wind power data include: blade vibration frequency and the average thickness of ice on the blade.

[0015] It should be noted that icing will change the aerodynamic shape of the blades, which directly affects the lift-to-drag ratio of the blades, that is, the ratio of lift to drag. Under normal circumstances, the blades of wind turbines are designed to have optimal aerodynamic characteristics and can generate maximum lift and minimum drag within a specific wind speed range. Ice coating will make the surface of the blades rough, destroying the originally smooth airflow boundary layer, thereby reducing lift and increasing drag, resulting in a decrease in the amount of wind energy that can be captured by the iced blades and a decrease in power output. At the same time, icing will also increase the weight of the blades and the air resistance during rotation. The increase in weight changes the dynamic balance of the blades, increasing the energy required to drive the blades to rotate and increasing the requirements for the generator and transmission system. The increase in air resistance reduces the effective wind-catching area of ​​the blades and further reduces the power output capacity. Ice coating on the blades will cause an imbalance in the weight of the blades, which will generate asymmetric torque during rotation, increase mechanical stress, and increase the vibration frequency of the blades, thereby reducing the efficiency of the wind turbine and resulting in a decrease in power output.

[0016] Preferably, the training of the artificial intelligence model based on historical environmental data includes:

[0017] Integrate the historical environmental data at the initial moment into standard input data; integrate the historical environmental data at the preset prediction moment into standard output data; wherein the historical environmental data is the historical data of the environmental data of the target area;

[0018] An artificial intelligence model is trained based on standard input data and standard output data to obtain a wind power prediction model; wherein the artificial intelligence model includes: a convolutional neural network or a deep belief network.

[0019] It should be noted that wind power ramping refers to the phenomenon of rapid increase or decrease in wind power output power within a shorter time scale; the preset prediction moment is obtained based on the actual preset time scale; wherein the preset time scale is equal to the time period between the preset prediction moment and the initial moment; the preset time scale is set according to the requirements of the actual wind farm.

[0020] Preferably, the predicting of environmental data based on the wind power prediction model includes:

[0021] The environmental data collected by the data acquisition device in the target area at the initial moment is input into the wind power prediction model to obtain the prediction result; wherein the prediction result refers to the environmental data at the initial moment corresponding to the preset prediction moment.

[0022] The present invention obtains environmental data of the target area; obtains a wind power prediction model by training an artificial intelligence model based on historical environmental data; and predicts environmental data based on the wind power prediction model to obtain the wind power output power at the prediction moment; this is beneficial for monitoring the phenomenon of rapid increase or decrease in wind power output power within a shorter time scale.

[0023] Preferably, the correcting the prediction result based on wind power data includes:

[0024] Correct the wind power output power in the forecast results based on wind power data;

[0025] S1: The vibration frequency of the blade is marked as Z, the average thickness of ice on the blade is marked as H, and the wind power output power in the prediction result is marked as P;

[0026] S2: Take Z, H and P as independent variables, and take the corrected target wind power output as the dependent variable, marked as ZP;

[0027] S3: Fit the independent variables and dependent variables by polynomial fitting to construct a revised model; the revised model is specifically:

[0028] ZP=P[1+A×e^Z / (e^Z+1)+B×tanh(H)×ln(H+1)];

[0029] Where A and B are proportional coefficients, and A < 0, B < 0, tanh(*) is the hyperbolic tangent function, and ln(*) is the logarithmic function with the natural number e as the base;

[0030] S4: The wind power output power in the prediction result is corrected based on the correction model to obtain the target wind power output power.

[0031] The vibration of the blades will destroy the aerodynamic characteristics of the blade surface, resulting in a decrease in the wind energy conversion efficiency. This is because when the blades vibrate, their angle of attack (the angle relative to the wind direction) will be unstable, thereby reducing the blades' ability to capture energy from the wind; at the same time, icing increases the weight of the blades, which not only changes the dynamic characteristics of the blades, such as the natural frequency and vibration mode, but also increases the load on the transmission system and the tower, resulting in a decrease in the overall mechanical efficiency; the prior art only considers the prediction of wind power ramp under normal weather and environmental conditions, and does not consider the effects of extreme weather, such as low temperature and icing, on wind power ramp, resulting in the problem of low prediction accuracy of wind power ramp under extreme weather conditions; the present invention obtains a wind power prediction model by training an artificial intelligence model based on historical environmental data; and, predicts environmental data based on the wind power prediction model to obtain a prediction result; and corrects the wind power output power in the prediction result based on the wind power data, thereby solving the above problems.

[0032] Preferably, the determining whether to send warning information based on the target wind power output power and a preset wind power ramp threshold value includes:

[0033] Determine whether the difference between the target wind power output power and the wind power output power at the initial moment is greater than the preset wind power ramp threshold; if yes, generate an early warning message and send it to the client of the wind farm manager; if not, continue to monitor and judge; wherein, the wind power output power at the initial moment refers to the wind power output power in the environmental data input into the wind power prediction model at the initial moment; the client includes: mobile phone or computer.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] The prior art only considers the prediction of wind power ramp under normal weather and environmental conditions, but does not consider the impact of extreme weather, such as low temperature and icing, on wind power ramp, resulting in low prediction accuracy of wind power ramp under extreme weather conditions. The present invention collects environmental data and wind power data of the target area in real time; trains an artificial intelligence model based on historical environmental data to obtain a wind power prediction model; and predicts environmental data based on the wind power prediction model to obtain prediction results; constructs a correction model based on wind power data, and corrects the wind power output power in the prediction result based on the correction model to obtain the target wind power output power. By judging whether the difference between the target wind power output power and the wind power output power at the initial moment is greater than the preset wind power ramp threshold, an early warning prompt is issued, thereby ensuring the stability of the wind power output power and solving the problem of inaccurate wind power ramp prediction under low temperature and icing conditions in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention 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 some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0037] Figure 1 A schematic diagram of a system module according to an embodiment of the present invention;

[0038] Figure 2 A schematic diagram of a specific process of an embodiment of the present invention;

[0039] Figure 3 Schematic diagram of wind power ramp warning judgment according to an embodiment of the present invention. DETAILED DESCRIPTION

[0040] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] See also Figure 1-Figure 2 , the first aspect of the present invention provides a wind power ramp prediction system for extreme weather, including: a data acquisition module, a data prediction module and a data early warning module;

[0042] Data acquisition module: used to obtain environmental data and wind power data of the target area;

[0043] Data prediction module: training an artificial intelligence model based on historical environmental data to obtain a wind power prediction model; and predicting environmental data based on the wind power prediction model to obtain prediction results;

[0044] Data warning module: correct the prediction results based on wind power data to obtain the target wind power output power; and determine whether to send a warning message based on the target wind power output power and the preset wind power climbing threshold; if yes, send the warning message to the client of the wind farm manager; if not, continue to monitor and judge.

[0045] Obtain environmental data and wind power data for the target area, including:

[0046] The environmental data and wind power data of the target area are collected in real time through data acquisition equipment; the target area refers to a wind farm in low temperature and icy weather; the environmental data include: ambient temperature, ambient humidity, wind speed, wind direction, air density, atmospheric pressure and wind power output power; the wind power data include: blade vibration frequency and the average thickness of ice on the blade.

[0047] Training AI models based on historical environmental data, including:

[0048] Integrate the historical environmental data at the initial moment into standard input data; integrate the historical environmental data at the preset prediction moment into standard output data; wherein the historical environmental data is the historical data of the environmental data of the target area;

[0049] An artificial intelligence model is trained based on standard input data and standard output data to obtain a wind power prediction model; wherein the artificial intelligence model includes: a convolutional neural network or a deep belief network.

[0050] Predict environmental data based on wind power prediction model, including:

[0051] The environmental data collected by the data acquisition device in the target area at the initial moment is input into the wind power prediction model to obtain the prediction result; wherein the prediction result refers to the environmental data at the initial moment corresponding to the preset prediction moment.

[0052] The forecast results are corrected based on wind power data, including:

[0053] Correct the wind power output power in the forecast results based on wind power data;

[0054] S1: The vibration frequency of the blade is marked as Z, the average thickness of ice on the blade is marked as H, and the wind power output power in the prediction result is marked as P;

[0055] S2: Take Z, H and P as independent variables, and take the corrected target wind power output as the dependent variable, marked as ZP;

[0056] S3: Fit the independent variables and dependent variables by polynomial fitting to construct a revised model; the revised model is specifically:

[0057] ZP=P[1+A×e^Z / (e^Z+1)+B×tanh(H)×ln(H+1)];

[0058] Where A and B are proportional coefficients, and A < 0, B < 0, tanh(*) is the hyperbolic tangent function, and ln(*) is the logarithmic function with the natural number e as the base;

[0059] S4: The wind power output power in the prediction result is corrected based on the correction model to obtain the target wind power output power.

[0060] See also Figure 3 , based on the target wind power output power and the preset wind power ramp threshold, determining whether to send an early warning message includes:

[0061] Determine whether the difference between the target wind power output power and the wind power output power at the initial moment is greater than the preset wind power ramp threshold; if yes, generate an early warning message and send it to the client of the wind farm manager; if not, continue to monitor and judge; wherein, the wind power output power at the initial moment refers to the wind power output power in the environmental data input into the wind power prediction model at the initial moment; the client includes: mobile phone or computer.

[0062] For example, suppose a wind farm is located in a high-latitude area that often experiences low temperatures and icy weather. It is necessary to predict the wind power ramp of the wind farm as follows:

[0063] 1. Data collection;

[0064] Ambient temperature and humidity:

[0065] Real-time measurements are taken using temperature and humidity sensors. These sensors are usually installed at different locations in the wind farm, such as at the tower base, inside the nacelle, or at a nearby weather station to ensure representative data.

[0066] Wind speed and direction:

[0067] Wind speed is measured by anemometers (such as cup anemometers or ultrasonic anemometers), which are usually installed on wind turbine towers and can accurately capture the changes in wind speed at different heights.

[0068] Wind direction is obtained through a wind vane or in combination with the built-in wind direction measurement function of an anemometer.

[0069] Air Density and Atmospheric Pressure:

[0070] The air density in the wind farm environment is measured by an air density sensor.

[0071] Atmospheric pressure is measured by a barometer, which can be installed at a weather station or on a wind tower at the wind farm.

[0072] Wind power output power:

[0073] Wind power output power is the direct output of the wind turbine and is obtained in real time through the wind turbine control system or power monitoring system.

[0074] Vibration frequency of blades:

[0075] Vibration sensors (such as accelerometers) are installed on the blades to monitor the vibration of the blades in real time.

[0076] Average thickness of ice on blades:

[0077] The thickness of ice covering the blade can be obtained by a variety of non-contact measurement technologies, such as laser ranging or infrared thermal imaging technology.

[0078] Assume that the collected data is:

[0079] Environmental data:

[0080] Ambient temperature: -5℃;

[0081] Ambient humidity: 60%;

[0082] Wind speed: 10m / s;

[0083] Wind direction: northeast;

[0084] Air density: 1.25kg / m 3 ;

[0085] Atmospheric pressure: 1013hPa;

[0086] Current wind power output: 1500kW;

[0087] Wind power data:

[0088] Blade vibration frequency (Z): 10Hz;

[0089] Average thickness of ice on blades (H): 5mm;

[0090] 2. Data prediction;

[0091] Based on an artificial intelligence model trained with historical environmental data (for example, using a convolutional neural network (CNN)), a wind power prediction model is obtained, which can predict the wind power output power in future periods.

[0092] Set the preset prediction time to 1 hour in the future and the preset time scale to 1 hour.

[0093] Assume that the wind power prediction model predicts that the wind power output power in the next hour is 1800kW (uncorrected).

[0094] 3. Data early warning;

[0095] A correction model is constructed based on wind power data, specifically:

[0096] ZP=P[1+A×e^Z / (e^Z+1)+B×tanh(H)×ln(H+1)];

[0097] Assume that the proportionality coefficients A = -0.05, B = -0.01 (negative numbers indicate that ice accumulation and increased vibration frequency will reduce output power).

[0098] Modified model application:

[0099] Substituting the predicted wind power output power (P) = 1800 kW, the blade vibration frequency (Z) = 10 Hz, and the average thickness of ice on the blades (H) = 5 mm, the corrected target wind power output power (ZP) is calculated.

[0100] ZP=1800×[1+(-0.05)×e^10 / (e^10+1)+(-0.01)×tanh(5)×ln(5+1)]≈1767.72kW;

[0101] Early warning judgment:

[0102] The initial wind power output power is 1500kW, and the corrected target wind power output power is 1767.72kW.

[0103] Assuming that the preset wind power ramp threshold is ±200kW / h, the difference between the corrected target wind power output power and the initial wind power output power is 267.72kW.

[0104] Since the difference exceeds the preset threshold, the system will generate an early warning message and send it to the client of the wind farm manager.

[0105] The present invention uses an artificial intelligence model based on historical data combined with a correction model for wind power data to more accurately predict changes in wind power and issue early warnings in a timely manner to avoid potential power supply fluctuations and equipment damage risks. This prediction and correction mechanism is crucial to improving the operational efficiency and reliability of wind farms.

[0106] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0107] Working principle of the present invention:

[0108] The present invention collects environmental data and wind power data of the target area in real time; trains an artificial intelligence model based on historical environmental data to obtain a wind power prediction model; and predicts environmental data based on the wind power prediction model to obtain a prediction result; constructs a correction model based on the wind power data, and corrects the wind power output power in the prediction result based on the correction model to obtain the target wind power output power, and issues an early warning prompt by judging whether the difference between the target wind power output power and the wind power output power at an initial moment is greater than a preset wind power climbing threshold, thereby ensuring the stability of the wind power output power and solving the problem of inaccurate wind power climbing prediction in the prior art under low temperature and icing conditions.

[0109] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A wind power ramp prediction system for extreme weather, characterized in that: include: Data collection module, data prediction module and data early warning module; Data acquisition module: used to obtain environmental data and wind power data of the target area; Data prediction module: trains the artificial intelligence model based on historical environmental data to obtain a wind power prediction model; as well as, Predict environmental data based on the wind power prediction model to obtain prediction results; Data warning module: corrects the prediction results based on wind power data to obtain the target wind power output power; as well as, Determine whether to send warning information based on the target wind power output power and the preset wind power ramp threshold; If yes, then an early warning message is sent to the client of the wind farm manager; If not, continue to monitor and judge.

2. The extreme weather wind power ramp prediction system according to claim 1, characterized in that: The step of obtaining environmental data and wind power data of the target area includes: The environmental data and wind power data of the target area are collected in real time through data acquisition equipment; the target area refers to a wind farm in low temperature and icy weather; the environmental data include: ambient temperature, ambient humidity, wind speed, wind direction, air density, atmospheric pressure and wind power output power; the wind power data include: blade vibration frequency and the average thickness of ice on the blade.

3. The extreme weather wind power ramp prediction system according to claim 1, characterized in that: The training of the artificial intelligence model based on historical environmental data includes: Integrate the historical environmental data at the initial moment into standard input data; integrate the historical environmental data at the preset prediction moment into standard output data; wherein the historical environmental data is the historical data of the environmental data of the target area; An artificial intelligence model is trained based on standard input data and standard output data to obtain a wind power prediction model; wherein the artificial intelligence model includes: a convolutional neural network or a deep belief network.

4. The extreme weather wind power ramp prediction system according to claim 1, characterized in that: The predicting of environmental data based on the wind power prediction model includes: The environmental data collected by the data acquisition device in the target area at the initial moment is input into the wind power prediction model to obtain the prediction result; wherein the prediction result refers to the environmental data at the initial moment corresponding to the preset prediction moment.

5. The extreme weather wind power ramp prediction system according to claim 1, characterized in that: The correction of the prediction result based on the wind power data includes: Correct the wind power output power in the forecast results based on wind power data; S1: The vibration frequency of the blade is marked as Z, the average thickness of ice on the blade is marked as H, and the wind power output power in the prediction result is marked as P; S2: Take Z, H and P as independent variables, and take the corrected target wind power output as the dependent variable, marked as ZP; S3: Fit the independent variables and dependent variables by polynomial fitting to construct a revised model; the revised model is specifically: ZP=P[1+A×e^Z / (e^Z+1)+B×tanh(H)×ln(H+1)]; Where A and B are proportional coefficients, and A < 0, B < 0, tanh(*) is the hyperbolic tangent function, and ln(*) is the logarithmic function with the natural number e as the base; S4: The wind power output power in the prediction result is corrected based on the correction model to obtain the target wind power output power.

6. The extreme weather wind power ramp prediction system according to claim 1, characterized in that: The determining whether to send warning information based on the target wind power output power and the preset wind power ramp threshold includes: Determine whether the difference between the target wind power output power and the wind power output power at the initial moment is greater than the preset wind power ramp threshold; if yes, generate an early warning message and send it to the client of the wind farm manager; if not, continue to monitor and judge; wherein, the wind power output power at the initial moment refers to the wind power output power in the environmental data input into the wind power prediction model at the initial moment; the client includes: mobile phone or computer.

Images