Method, device and equipment for predicting wind power climbing event and medium

By using a combined prediction model of coding, attention mechanism and decoding sub-model in offshore wind farms, wind speed and power information are extracted, and the attribute information of hill climbing events is accurately predicted, and offshore wind power power hill climbing events in the prior art are solved. The problem of low prediction accuracy improves the safety and stability of the power system.

CN118964929AActive Publication Date: 2024-11-15STATE GRID JIANGSU ELECTRIC POWER CO LTD +1

Patent Information

Application Number
CN202410979843.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-22
Publication Date
2024-11-15
Estimated Expiration
2044-07-22

AI Technical Summary

Technical Problem

The existing wind power climbing event prediction methods are not suitable for the complex weather systems of offshore wind farms, resulting in low prediction accuracy and affecting the safe and stable operation of the power system.

Method used

The preset hill climb event prediction model consisting of coding sub-model, attention mechanism sub-model and multiple parallel decoding sub-models is used to obtain wind speed and power information of offshore wind farms, and the characteristic information of hill climb event is extracted to accurately predict climbing. Attribute information for slope events, including category, initial power, start time, climb amplitude, climb direction, and duration.

Benefits of technology

It improves the prediction accuracy of offshore wind power climbing events and ensures the safe and stable operation of the power system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118964929A_ABST
    Figure CN118964929A_ABST
Patent Text Reader

Abstract

The embodiment of the invention discloses a prediction method and device for a wind power climbing event, equipment and a medium. Comprising the steps of obtaining wind speed information and power information of an offshore wind plant; inputting the wind speed information and the power information into a preset climbing event prediction model, and obtaining attribute information of a climbing event output by the preset climbing event prediction model; the preset climbing event prediction model comprises a coding sub-model, an attention mechanism sub-model and a plurality of parallel decoding sub-models, and the output end of the coding sub-model is connected with the input end of the attention mechanism sub-model; the output end of the attention mechanism sub-model is connected with the input end of each decoding sub-model, and different decoding sub-models are used for outputting different types of attribute information of the climbing event. According to the scheme, the attribute information of the climbing event can be accurately obtained, the prediction precision of the offshore wind power climbing event is improved, and then safe and stable operation of a power system is guaranteed.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The embodiments of the present invention relate to the technical field of offshore wind power prediction, and in particular to a method, device, equipment and medium for predicting wind power ramp events. Background Art

[0002] With the continuous development of offshore wind power generation technology, during the wind power generation process, the change in wind speed may cause the wind power to fluctuate greatly in a short period of time. This phenomenon is called a wind power ramp event. Therefore, in order to ensure the safe and stable operation of the power system, it is urgent to predict offshore wind power ramp events.

[0003] Existing prediction methods for wind power ramp events are mainly aimed at predicting wind power ramp events in onshore wind farms, and are not suitable for predicting wind power ramp events in offshore wind farms with more complex weather systems, resulting in lower prediction accuracy for offshore wind power ramp events. Summary of the invention

[0004] The embodiments of the present invention provide a method, device, equipment and medium for predicting wind power ramping events, which can predict offshore wind power ramping events, improve prediction accuracy, and thus ensure safe and stable operation of the power system.

[0005] In a first aspect, an embodiment of the present invention provides a method for predicting a wind power ramp event, comprising:

[0006] Obtain wind speed and power information of offshore wind farms;

[0007] The wind speed information and the power information are input into a preset climbing event prediction model to obtain attribute information of the climbing event output by the preset climbing event prediction model, wherein the attribute information includes a climbing event category; the preset climbing event prediction model includes an encoding sub-model, an attention mechanism sub-model and multiple parallel decoding sub-models, wherein the output end of the encoding sub-model is connected to the input end of the attention mechanism sub-model; the output end of the attention mechanism sub-model is respectively connected to the input end of each decoding sub-model, and different decoding sub-models are used to output different types of attribute information of the climbing event.

[0008] In a second aspect, an embodiment of the present invention further provides a device for predicting wind power ramp events, the device comprising:

[0009] An information acquisition module is used to obtain wind speed information and power information of an offshore wind farm;

[0010] A climbing event prediction module is used to input the wind speed information and the power information into a preset climbing event prediction model, and obtain the attribute information of the climbing event output by the preset climbing event prediction model, wherein the attribute information includes the climbing event category; the preset climbing event prediction model includes an encoding sub-model, an attention mechanism sub-model and multiple parallel decoding sub-models, wherein the output end of the encoding sub-model is connected to the input end of the attention mechanism sub-model; the output end of the attention mechanism sub-model is respectively connected to the input end of each decoding sub-model, and different decoding sub-models are used to output different types of attribute information of the climbing event.

[0011] In a third aspect, an embodiment of the present invention further provides an electronic device, the electronic device comprising:

[0012] at least one processor; and

[0013] a memory communicatively connected to the at least one processor; wherein,

[0014] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the method for predicting wind power ramping events described in the embodiment of the present invention.

[0015] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for predicting wind power ramping events described in the embodiment of the present invention when executed.

[0016] In a fifth aspect, an embodiment of the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements the method for predicting wind power ramping events described in an embodiment of the present invention.

[0017] The embodiment of the present invention discloses a prediction method, device, equipment and medium for wind power ramp events, including: obtaining wind speed information and power information of an offshore wind farm; inputting the wind speed information and the power information into a preset ramp event prediction model, and obtaining attribute information of the ramp event output by the preset ramp event prediction model, wherein the attribute information includes a ramp event category; the preset ramp event prediction model includes an encoding sub-model, an attention mechanism sub-model and multiple parallel decoding sub-models, wherein the output end of the encoding sub-model is connected to the input end of the attention mechanism sub-model; the output end of the attention mechanism sub-model is respectively connected to the input end of each decoding sub-model, and different decoding sub-models are used to output different types of attribute information of the ramp event. The above technical solution can realize the prediction of offshore wind power ramping events, specifically considering the influence of wind speed and power on the ramping events, so as to obtain the wind speed information and power information of the offshore wind farm; at the same time, it also considers inputting the acquired wind speed information and power information into the preset ramping event prediction model trained in advance to obtain the attribute information of the ramping event; wherein, the preset ramping event prediction model composed of the encoding sub-model, the attention mechanism sub-model and multiple parallel decoding sub-models has a high feature extraction capability, and can extract feature information about the ramping event from the wind speed information and power information of the offshore wind farm, so as to accurately obtain the attribute information of the ramping event, and improve the prediction accuracy of the offshore wind power ramping event. Compared with the prior art, the technical solution provided by this embodiment can accurately obtain the attribute information of the ramping event, improve the prediction accuracy of the offshore wind power ramping event, and thus ensure the safe and stable operation of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a flow chart of a method for predicting wind power ramping events in the first embodiment of the present invention;

[0019] Figure 2 A schematic diagram of a hill climbing event provided by an embodiment of the present invention;

[0020] Figure 3 A schematic diagram of the structure of a preset hill climbing event prediction model provided by an embodiment of the present invention;

[0021] Figure 4 is a flow chart of a method for predicting wind power ramp events in Embodiment 2 of the present invention;

[0022] Figure 5 It is a structural schematic diagram of a device for predicting wind power ramp events in Embodiment 3 of the present invention;

[0023] Figure 6 It is a structural schematic diagram of an electronic device in Embodiment 4 of the present invention. DETAILED DESCRIPTION

[0024] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.

[0025] Embodiment 1

[0026] Figure 1 This is a flow chart of a method for predicting a wind power ramp event provided in the first embodiment of the present invention. This embodiment is applicable to the case of predicting an offshore wind power ramp event. The method can be executed by a wind power ramp event prediction device, which can be implemented in the form of software and / or hardware, and optionally, by an electronic device, which can be a mobile terminal, a PC or a server, etc. Specifically, it includes the following steps:

[0027] S110: Obtain wind speed information and power information of the offshore wind farm.

[0028] In this embodiment, an offshore wind farm can be understood as an area where wind turbines are installed in a marine environment; wherein multiple wind farms are arranged at sea, and each wind farm generates wind energy through a wind turbine. The wind speed information is the wind speed of the offshore wind farm, and the power information can be understood as the power of converting wind energy into electrical energy through the wind turbine. Specifically, the wind speed information and power information of multiple offshore wind farms can be obtained.

[0029] S120: Input wind speed information and power information into a preset climbing event prediction model, and obtain attribute information of the climbing event output by the preset climbing event prediction model.

[0030] In this embodiment, the preset ramp event prediction model is a pre-trained model for realizing the prediction of ramp events. The ramp event can be understood as a phenomenon in which the power of an offshore wind farm fluctuates greatly due to changes in wind speed; the attribute information can be understood as key information of the ramp event, which can include the ramp event category and at least one of the initial power, start time, ramp amplitude, ramp direction and duration when the ramp event occurs.

[0031] Continuing with the above description, the climbing event category can be understood as the type of climbing event, such as: large-amplitude climbing event, small-amplitude climbing event, long-duration climbing event, etc. The initial power can be understood as the power corresponding to the beginning of the climbing event; the start time can be understood as the time when the climbing event begins to occur; the climbing amplitude can be understood as the power change amplitude after the climbing event occurs; the climbing direction can be understood as the direction of power change within the duration of the climbing event, and the power increase within the duration of the climbing event is represented as up-climbing, and the power decrease is represented as down-climbing; the duration can be understood as the total time from the occurrence to the end of the climbing event.

[0032] Optionally, a ramp event can be characterized by the initial power, start time, ramp amplitude, ramp direction and duration when the ramp event occurs. For better understanding, Figure 2 A schematic diagram of a hill climbing event provided by an embodiment of the present invention is shown in FIG. Figure 2 As shown, the power curve is used to represent the power change of the ramp event, and the initial power of the ramp event is recorded as P0, the start time is recorded as t0, and the ramp amplitude is recorded as P A , the climbing direction is recorded as P F And it is climbing uphill, the duration is recorded as t A .

[0033] In this embodiment, Figure 3 A schematic diagram of the structure of a preset hill climbing event prediction model provided by an embodiment of the present invention, such as Figure 3 As shown, the preset climbing event prediction model includes an encoding sub-model, an attention mechanism sub-model and multiple parallel decoding sub-models. The output end of the encoding sub-model is connected to the input end of the attention mechanism sub-model; the output end of the attention mechanism sub-model is respectively connected to the input end of each decoding sub-model, and different decoding sub-models are used to output different types of attribute information of the climbing event.

[0034] Continuing with the above description, the encoding sub-model is composed of a three-layer graph convolutional neural network (GCN); the attention mechanism sub-model can be a neural network model; each decoding sub-model is composed of a three-layer convolutional neural network (CNN), and different decoding sub-models are used to output the climbing event category of the climbing event, the initial power when the climbing event occurs, the start time, the climbing amplitude, the climbing direction and the duration.

[0035] Specifically, for inputting wind speed information and power information into a preset climbing event prediction model to obtain the attribute information of the climbing event output by the preset climbing event prediction model, one implementation method can be described as: inputting the wind speed information and power information of multiple offshore wind farms into the encoding sub-model to obtain the encoding information corresponding to the wind speed information and the power information; inputting the encoding information into the attention mechanism sub-model to obtain the characteristic information of the climbing event; inputting the characteristic information of the climbing event into multiple parallel decoding sub-models to respectively obtain the climbing event category, the initial power when the climbing event occurs, the start time, the climbing amplitude, the climbing direction and the duration.

[0036] The technical solution of this embodiment obtains the wind speed information and power information of the offshore wind farm; inputs the wind speed information and power information into the preset climbing event prediction model, and obtains the attribute information of the climbing event output by the preset climbing event prediction model. The above technical solution takes into account the influence of wind speed and power on the climbing event, thereby obtaining the wind speed information and power information of the offshore wind farm; at the same time, it also considers inputting the acquired wind speed information and power information into the preset climbing event prediction model trained in advance to obtain the attribute information of the climbing event; wherein, the preset climbing event prediction model composed of the encoding sub-model, the attention mechanism sub-model and multiple parallel decoding sub-models has a high feature extraction capability, and can extract feature information about the climbing event from the wind speed information and power information of the offshore wind farm, so as to accurately obtain the attribute information of the climbing event, and improve the prediction accuracy of the offshore wind power climbing event. The technical solution provided by this embodiment can accurately obtain the attribute information of the climbing event, improve the prediction accuracy of the offshore wind power climbing event, and thus ensure the safe and stable operation of the power system.

[0037] Embodiment 2

[0038] Figure 4 This is a flow chart of a method for predicting a wind power ramp event provided in the second embodiment of the present invention. Based on the above embodiment, the method includes the following steps:

[0039] S210. Collect power sequence samples of the offshore wind farm.

[0040] In this embodiment, the power sequence samples can be understood as power data arranged in chronological order. Moreover, in the power sequence samples arranged in chronological order, there may be a situation where the power changes significantly within a certain period of time, that is, a wind power ramp event.

[0041] S220. Decompose the power sequence samples using a variational mode decomposition method to obtain a plurality of power components arranged from high frequency to low frequency.

[0042] In this embodiment, the variational modal decomposition method is a signal processing method, which can decompose the original signal to obtain multiple sub-signal components of different frequencies; the power component can be understood as a power signal curve that fluctuates according to a certain period, the high-frequency power component is a power signal curve with a shorter period, and the low-frequency power component is a power signal curve with a longer period.

[0043] Exemplarily, the collected power sequence samples are decomposed using a variational mode decomposition method to obtain k power components arranged from high frequency to low frequency.

[0044] S230: Construct a hill climbing event sample set based on multiple power components.

[0045] In this embodiment, the ramp event sample set is used to train a preset ramp event prediction model, so that the preset ramp event prediction model can accurately predict offshore wind power ramp events. The ramp event sample set includes historical wind speed samples and historical power samples, and attribute information samples of real ramp events corresponding to the historical wind speed samples and historical power samples.

[0046] Furthermore, the historical wind speed samples are the wind speed information of the offshore wind farm within the historical time period, and the historical power samples are the power information of the offshore wind farm within the historical time period; the attribute information samples of the real climbing events can be understood as the attribute information of the actual climbing events corresponding to the historical wind speed and historical power.

[0047] In this embodiment, for constructing a hill climbing event sample set based on multiple power components, one implementation method can be described as the following steps:

[0048] a1) extracting at least one of an initial power sample, a start time sample, a climbing amplitude sample, a climbing direction sample and a duration sample of each real climbing event in the power sequence sample based on multiple power components;

[0049] In this embodiment, since there are multiple situations where the power changes significantly in the power sequence samples, this situation can be understood as a power climbing event, and can be simply understood as the presence of multiple power climbing events in the power sequence samples. Therefore, the initial power sample, start time sample, climbing amplitude sample, climbing direction sample and duration sample of each real climbing event can be extracted from the power sequence samples.

[0050] In this embodiment, for extracting at least one of the initial power sample, start time sample, climbing amplitude sample, climbing direction sample and duration sample of each real climbing event in the power sequence sample based on multiple power components, one implementation method can be described as: according to the arrangement order of multiple power components, a set number of power components are obtained from back to front as the power components to be fused; the power components to be fused are fused to obtain the fused power components; the first time point corresponding to the local maximum value and the second time point corresponding to the local minimum value in the fused power components are obtained; based on the power data between the adjacent first time point and the second time point in the power sequence sample, at least one of the initial power sample, start time sample, climbing amplitude sample, climbing direction sample and duration sample of the real climbing event is determined.

[0051] Continuing with the above description, the set number is a value set in advance, and the set number must be less than the number of power components, for example, obtaining 2 power components from the back to the front, that is, obtaining the last two power components. The power components to be fused are the set number of power components obtained; the fusion can be understood as addition, and the fused power components can be understood as the power components obtained by adding the set number of power components. The power component is a power signal curve that fluctuates according to a certain period, then the local maximum value in the fused power component can be understood as the value of a certain power signal in the power signal curve is greater than the power signal value on the left and the power signal value on the right of the power signal; the local minimum value can be understood as the value of a certain power signal in the power signal curve is less than the power signal value on the left and the power signal value on the right of the power signal.

[0052] Furthermore, the first time point can be understood as the time corresponding to the local maximum value in the fused power component, and the second time point can be understood as the time corresponding to the local minimum value in the fused power component; and there are multiple local maxima and multiple local minima in the fused power component, so multiple first time points and multiple second time points can be obtained. The power data between the adjacent first time point and the second time point in the power sequence sample is a power that changes greatly. Since there are multiple first time points and multiple second time points, there are multiple groups of power data between the adjacent first time point and the second time point in the power sequence sample, that is, multiple groups of power that changes greatly; and the power that changes greatly between the adjacent first time point and the second time point can be regarded as a wind power ramp event.

[0053] Exemplarily, an implementation method for extracting at least one of the initial power sample, start time sample, climbing amplitude sample, climbing direction sample and duration sample of each real climbing event in the power sequence samples based on multiple power components may be: from multiple power components arranged from high frequency to low frequency, two power components are obtained from back to front as power components to be fused; the two power components to be fused are added to obtain the added power component; in the added power component, a first time point corresponding to multiple local maxima and a second time point corresponding to multiple local minima are obtained; in the power sequence samples, power data between multiple groups of adjacent first time points and second time points are extracted, and the initial power sample, start time sample, climbing amplitude sample, climbing direction sample and duration sample of each group of corresponding real climbing events are determined according to each group of power data. Exemplarily, for a certain set of power data, assuming that the adjacent first time point and second time point are t1 and t2 respectively, the power corresponding to time t1 is p1, and the power corresponding to time t2 is p2, then the initial power sample of the actual climbing event corresponding to the group is p1, the start time sample is t1, the climbing amplitude sample is (p2-p1), the climbing direction sample is downward, and the duration sample is (t2-t1).

[0054] b1) Obtaining historical wind speed samples and historical power samples corresponding to each real climbing event;

[0055] In this embodiment, after obtaining the initial power sample, start time sample, climbing amplitude sample, climbing direction sample and duration sample of each real climbing event, for each real climbing event, it is also necessary to obtain the historical wind speed sample and historical power sample corresponding to the real climbing event.

[0056] c1) classifying each real climbing event to obtain a real category sample of each real climbing event;

[0057] In this embodiment, the real category samples can be understood as the types to which the real climbing events belong. Optionally, a K-means clustering algorithm can be used to classify the real climbing events to obtain the real category samples of the real climbing events, or other clustering algorithms can be used for classification, which is not limited here.

[0058] d1) The historical wind speed samples, historical power samples and attribute information samples of real climbing events constitute a climbing event sample set.

[0059] In this embodiment, the attribute information samples include at least one of an initial power sample, a start time sample, a climbing amplitude sample, a climbing direction sample, and a duration sample, and a true category sample.

[0060] In this embodiment, the climbing event sample set includes multiple historical wind speed samples, historical power samples and attribute information samples of real climbing events; for each historical wind speed sample and historical power sample, there is a corresponding attribute information sample of the real climbing event.

[0061] S240: Training a preset hill climbing event prediction model based on the hill climbing event sample set.

[0062] In this embodiment, before using the preset ramp event prediction model to predict the future ramp events of the offshore wind farm, the preset ramp event prediction model needs to be trained first, so that the trained preset ramp event prediction model can accurately predict the ramp events.

[0063] In this embodiment, for training a preset hill climbing event prediction model based on a hill climbing event sample set, one implementation method can be described as the following steps:

[0064] a2) inputting historical wind speed samples and historical power samples into a preset climbing event prediction model to obtain attribute information of the predicted climbing event output by the preset climbing event prediction model;

[0065] In this embodiment, historical wind speed samples and historical power samples are input into a preset climbing event prediction model to obtain attribute information of the predicted climbing event, namely, predicted initial power, predicted start time, predicted climbing amplitude, predicted climbing direction, predicted duration and predicted category.

[0066] b2) determining a loss function based on attribute information of predicted climbing events and attribute information samples of actual climbing events;

[0067] In this embodiment, the attribute information of the predicted climbing event may be predicted initial power, predicted start time, predicted climbing amplitude, predicted climbing direction, predicted duration and predicted category, and the attribute information samples of the real climbing event may be initial power samples, start time samples, climbing amplitude samples, climbing direction samples, duration samples and real category samples.

[0068] Furthermore, the loss function may be a mean square error loss function, a binary cross entropy loss function, and a cross entropy loss function. Among them, the loss function of initial power, start time, climbing amplitude, and duration is a mean square error loss function; the loss function of climbing direction is a binary cross entropy loss function, and the loss function of climbing event category is a cross entropy loss function.

[0069] Specifically, one implementation method for determining the loss function based on the attribute information of the predicted climbing event and the attribute information sample of the real climbing event can be described as: constructing a first mean square error loss function between the predicted initial power and the initial power sample; constructing a second mean square error loss function between the predicted start time and the start time sample; constructing a third mean square error loss function between the predicted climbing amplitude and the climbing amplitude sample; constructing a fourth mean square error loss function between the predicted duration and the duration sample; constructing a binary cross entropy loss function between the predicted climbing direction and the climbing direction sample; constructing a cross entropy loss function between the predicted category and the real category sample; performing weighted summation on the above-mentioned first mean square error loss function, second mean square error loss function, third mean square error loss function, fourth mean square error loss function, binary cross entropy loss function and cross entropy loss function to obtain the weighted sum loss function. Among them, the weight corresponding to each loss function is a value set in advance.

[0070] Exemplarily, assuming that the first mean square error loss function, the second mean square error loss function, the third mean square error loss function, the fourth mean square error loss function, the binary cross entropy loss function and the cross entropy loss function are represented by f1, f2, f3, f4, f5 and f6 respectively, and their corresponding weights are w1, w2, w3, w4, w5 and w6 respectively. Then the weighted summation loss function is (f1*w1+f2*w2+f3*w3+f4*w4+f5*w5+f6*w6).

[0071] c2) adjusting model parameters of a preset hill climbing event prediction model based on the loss function.

[0072] In this embodiment, the model parameters may be weights and bias values ​​in the model.

[0073] In this embodiment, the model parameters in the preset climbing event prediction model are reversely adjusted according to the above-mentioned weighted summation loss function to obtain the preset climbing event prediction model after parameter adjustment; the historical wind speed samples and the historical power samples are continuously input into the preset climbing event prediction model after parameter adjustment to obtain the attribute information of the predicted climbing event output by the preset climbing event prediction model after parameter adjustment; then a new loss function is determined based on the attribute information of the predicted climbing event and the attribute information samples of the actual climbing event, and the model parameters are further reversely adjusted according to the new loss function; the above steps are repeated until the preset climbing event prediction model converges, indicating that the training is completed.

[0074] S250: Obtain wind speed information and power information of the offshore wind farm.

[0075] In this embodiment, after the preset ramp event prediction model is trained, wind speed information and power information of the offshore wind farm are acquired in real time.

[0076] S260: Input the wind speed information and the power information into a preset climbing event prediction model, and obtain the attribute information of the climbing event output by the preset climbing event prediction model.

[0077] In this embodiment, the acquired wind speed information and power information are input into a trained preset climbing event prediction model, so that the attribute information of the climbing event can be accurately obtained, namely, the climbing event category, the initial power when the climbing event occurs, the start time, the climbing amplitude, the climbing direction and the duration.

[0078] The technical solution of this embodiment is to collect power sequence samples of offshore wind farms; decompose the power sequence samples using a variational mode decomposition method to obtain multiple power components arranged from high frequency to low frequency; construct a climbing event sample set based on the multiple power components; train a preset climbing event prediction model based on the climbing event sample set; obtain wind speed information and power information of the offshore wind farm; input the wind speed information and power information into the preset climbing event prediction model to obtain attribute information of the climbing event output by the preset climbing event prediction model. The above technical solution takes into account the need to train a preset climbing event prediction model in advance, so as to collect power sequence samples of offshore wind farms; at the same time, the variational mode decomposition method is used to decompose the power sequence samples to obtain multiple power components, and a climbing event sample set is constructed based on the multiple power components, so that climbing event sample data can be obtained quickly and accurately; then, the preset climbing event prediction model is trained based on the constructed climbing event sample set, so that the preset climbing event prediction model can accurately predict the climbing event; finally, the wind speed information and power information are input into the trained preset climbing event prediction model, so that the attribute information of the climbing event can be accurately obtained. The technical solution provided in this embodiment can quickly and accurately obtain climbing event sample data, so that the preset climbing event prediction model can accurately predict the climbing event, so that the attribute information of the climbing event can be accurately obtained, which improves the prediction accuracy of offshore wind power climbing events, thereby ensuring the safe and stable operation of the power system.

[0079] Embodiment 3

[0080] Figure 5 A schematic diagram of a wind power ramp event prediction device provided in the third embodiment of the present invention is shown in FIG. Figure 5 As shown, the device comprises:

[0081] The information acquisition module 310 is used to acquire wind speed information and power information of the offshore wind farm;

[0082] The climbing event prediction module 320 is used to input wind speed information and power information into a preset climbing event prediction model, and obtain attribute information of the climbing event output by the preset climbing event prediction model, wherein the attribute information includes the climbing event category; the preset climbing event prediction model includes an encoding sub-model, an attention mechanism sub-model and multiple parallel decoding sub-models, wherein the output end of the encoding sub-model is connected to the input end of the attention mechanism sub-model; the output end of the attention mechanism sub-model is respectively connected to the input end of each decoding sub-model, and different decoding sub-models are used to output different types of attribute information of climbing events.

[0083] The technical solution of this embodiment obtains the wind speed information and power information of the offshore wind farm through the information acquisition module; the wind speed information and power information are input into the preset climbing event prediction model through the climbing event prediction module to obtain the attribute information of the climbing event output by the preset climbing event prediction model. The above technical solution takes into account the influence of wind speed and power on the climbing event, so as to obtain the wind speed information and power information of the offshore wind farm through the information acquisition module; at the same time, it also considers that the acquired wind speed information and power information are input into the preset climbing event prediction model trained in advance through the climbing event prediction module to obtain the attribute information of the climbing event; wherein the preset climbing event prediction model has a high feature extraction capability, and can extract feature information about the climbing event from the wind speed information and power information of the offshore wind farm, so as to accurately obtain the attribute information of the climbing event, and improve the prediction accuracy of the offshore wind power climbing event. The technical solution provided by this embodiment can accurately obtain the attribute information of the climbing event, improve the prediction accuracy of the offshore wind power climbing event, and thus ensure the safe and stable operation of the power system.

[0084] Optionally, the attribute information also includes at least one of initial power, start time, climbing amplitude, climbing direction and duration when the climbing event occurs.

[0085] Optionally, the wind power ramp event prediction device may further include:

[0086] Power sequence sample collection module, used to collect power sequence samples of offshore wind farms;

[0087] A power component acquisition module is used to decompose the power sequence samples using a variational mode decomposition method to obtain multiple power components arranged from high frequency to low frequency;

[0088] A climbing event sample set construction module is used to construct a climbing event sample set based on multiple power components; wherein the climbing event sample set includes historical wind speed samples and historical power samples, and attribute information samples of real climbing events corresponding to the historical wind speed samples and historical power samples;

[0089] The model training module is used to train a preset climbing event prediction model based on a climbing event sample set.

[0090] Optionally, the climbing event sample set construction module may specifically include:

[0091] A sample extraction unit, configured to extract at least one of an initial power sample, a start time sample, a climbing amplitude sample, a climbing direction sample and a duration sample of each real climbing event in the power sequence sample based on the multiple power components;

[0092] A historical wind speed sample and historical power sample acquisition unit, used to acquire historical wind speed samples and historical power samples corresponding to each real climbing event;

[0093] A climbing event classification unit, used to classify each real climbing event and obtain a real category sample of each real climbing event;

[0094] The sample set acquisition unit is used to form a climbing event sample set by combining historical wind speed samples, historical power samples and attribute information samples of real climbing events; the attribute information samples include at least one of the initial power sample, start time sample, climbing amplitude sample, climbing direction sample and duration sample and the real category sample.

[0095] Optionally, the sample extraction unit may be specifically used for:

[0096] According to the arrangement order of multiple power components, a set number of power components are obtained from back to front as power components to be fused; the power components to be fused are fused to obtain fused power components; a first time point corresponding to a local maximum value and a second time point corresponding to a local minimum value in the fused power components are obtained; and at least one of an initial power sample, a start time sample, a climbing amplitude sample, a climbing direction sample and a duration sample of a real climbing event is determined based on power data between adjacent first and second time points in the power sequence samples.

[0097] Optionally, the model training module may be specifically used for:

[0098] Input historical wind speed samples and historical power samples into a preset climbing event prediction model to obtain attribute information of the predicted climbing event output by the preset climbing event prediction model; determine a loss function based on the attribute information of the predicted climbing event and attribute information samples of the actual climbing event; and adjust model parameters of the preset climbing event prediction model based on the loss function.

[0099] The above device can execute the methods provided by all the above embodiments of the present invention, and has the corresponding functional modules and beneficial effects of executing the above methods. For technical details not described in detail in this embodiment, please refer to the methods provided by all the above embodiments of the present invention.

[0100] Embodiment 4

[0101] Figure 6 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0102] like Figure 6 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0103] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0104] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as a prediction method for wind power ramp events.

[0105] In some embodiments, the prediction method of the wind power ramp event may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the prediction method of the wind power ramp event described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to execute the prediction method of the wind power ramp event in any other appropriate manner (e.g., by means of firmware).

[0106] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0107] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0108] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0109] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0110] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0111] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0112] An embodiment of the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements a method for predicting wind power ramping events as provided in any embodiment of the present application.

[0113] In the process of implementation, the computer program product can be written in one or more programming languages ​​or a combination thereof to perform the computer program code of the present invention, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0114] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0115] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for predicting wind power ramp events, characterized in that: include: Obtain wind speed and power information of offshore wind farms; The wind speed information and the power information are input into a preset climbing event prediction model to obtain attribute information of the climbing event output by the preset climbing event prediction model, wherein the attribute information includes a climbing event category; the preset climbing event prediction model includes an encoding sub-model, an attention mechanism sub-model and multiple parallel decoding sub-models, wherein the output end of the encoding sub-model is connected to the input end of the attention mechanism sub-model; the output end of the attention mechanism sub-model is respectively connected to the input end of each decoding sub-model, and different decoding sub-models are used to output different types of attribute information of the climbing event.

2. The method according to claim 1, characterized in that The attribute information also includes at least one of initial power, start time, climbing amplitude, climbing direction and duration when the climbing event occurs.

3. The method according to claim 1, characterized in that Before obtaining the wind speed information and power information of the offshore wind farm, the method further includes: collecting power series samples of the offshore wind farm; Decomposing the power sequence samples by using a variational mode decomposition method to obtain a plurality of power components arranged from high frequency to low frequency; Constructing a climbing event sample set based on the multiple power components; wherein the climbing event sample set includes historical wind speed samples and historical power samples, and attribute information samples of real climbing events corresponding to the historical wind speed samples and historical power samples; The preset hill climbing event prediction model is trained based on the hill climbing event sample set.

4. The method according to claim 3, characterized in that The step of constructing a hill climbing event sample set based on the multiple power components comprises: Extracting at least one of an initial power sample, a start time sample, a climbing amplitude sample, a climbing direction sample, and a duration sample of each real climbing event in the power sequence samples based on the multiple power components; Obtaining historical wind speed samples and historical power samples corresponding to each of the real climbing events; Classifying each of the real climbing events to obtain a real category sample of each of the real climbing events; The historical wind speed samples, the historical power samples and the attribute information samples of the real climbing events constitute the climbing event sample set; the attribute information samples include at least one of the initial power samples, the start time samples, the climbing amplitude samples, the climbing direction samples and the duration samples, as well as the real category samples.

5. The method according to claim 4, characterized in that The extracting at least one of an initial power sample, a start time sample, a climbing amplitude sample, a climbing direction sample, and a duration sample of each real climbing event in the power sequence sample based on the multiple power components comprises: According to the arrangement order of the multiple power components, a set number of power components are acquired from back to front as the power components to be fused; Fusing the power components to be fused to obtain fused power components; Obtaining a first time point corresponding to a local maximum value and a second time point corresponding to a local minimum value in the fused power component; At least one of an initial power sample, a start time sample, a climbing amplitude sample, a climbing direction sample and a duration sample of a real climbing event is determined based on power data between the first adjacent time point and the second adjacent time point in the power sequence samples.

6. The method according to claim 3, characterized in that: The step of training the preset hill climbing event prediction model based on the hill climbing event sample set includes: Inputting the historical wind speed samples and the historical power samples into the preset climbing event prediction model, and obtaining attribute information of the predicted climbing event output by the preset climbing event prediction model; Determining a loss function based on the attribute information of the predicted hill climbing event and the attribute information sample of the actual hill climbing event; Model parameters of the preset hill climbing event prediction model are adjusted based on the loss function.

7. A prediction device for wind power ramp events, characterized in that: include: An information acquisition module is used to obtain wind speed information and power information of an offshore wind farm; A climbing event prediction module is used to input the wind speed information and the power information into a preset climbing event prediction model, and obtain the attribute information of the climbing event output by the preset climbing event prediction model, wherein the attribute information includes the climbing event category; the preset climbing event prediction model includes an encoding sub-model, an attention mechanism sub-model and multiple parallel decoding sub-models, wherein the output end of the encoding sub-model is connected to the input end of the attention mechanism sub-model; the output end of the attention mechanism sub-model is respectively connected to the input end of each decoding sub-model, and different decoding sub-models are used to output different types of attribute information of the climbing event.

8. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the method for predicting a wind power ramping event according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for predicting wind power ramping events according to any one of claims 1 to 6 when executed.

10. A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the method for predicting a wind power ramping event according to any one of claims 1 to 6 is implemented.

Citation Information

Patent Citations

  • Fluctuation feature identification-based wind power combination prediction method

    CN106933778A

  • Offshore wind power ultra-short-term prediction method and device based on climbing identification

    CN116739172A

  • Method and device for predicting wind power generation climbing event caused by severe convection

    CN117236502A

  • Wind speed and power hybrid prediction method and device considering wind power climbing

    CN117293825A

Cited By

  • Offshore wind power climbing prediction method and system

    CN119669865A

  • Offshore wind power ramp prediction method and system

    CN119669865B