Wind power prediction method and device and readable storage medium

Abstract

The embodiment of the invention provides a wind power prediction method and device and a readable storage medium, an electronic device obtains multiple frames of multispectral images of a target area in a historical time period to obtain a multispectral image sequence, and air flow speeds of different height layers of the target area in the historical time period are determined according to the multispectral image sequence. Afterwards, the electronic equipment predicts the wind speed of the wind power plant in each unit time in the future time period according to the air flow speeds of different height layers of the target area, and predicts the wind power of the wind power plant in each unit time in the future time period according to the wind speed of the wind power plant in each unit time in the future time period. By adopting the scheme, the electronic equipment utilizes the multispectral image to carry out ultra-short-term prediction of the wind power, and the updating level of the multispectral image is the minute level, so that the accuracy of the ultra-short-term prediction of the wind power can be improved.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the technical field of wind farms, in particular to a wind power prediction method, device and readable storage medium. BACKGROUND

[0002] As a clean and renewable energy, wind energy is increasingly valued by countries around the world. Wind turbine, short for wind generator, is a device that generates electricity using wind energy.

[0003] Wind power refers to the effective power that wind turbines in a wind farm convert from captured wind energy into a power grid that can be received, i.e. the net output power after deducting conversion loss, usually in units of kilowatts (kW) or megawatts (MW). Ultra-short-term prediction of wind power refers to predicting wind power in the short term in the future using historical power data of a wind farm, to provide a reliable basis for real-time scheduling. In common ultra-short-term prediction methods, a mathematical model is established to simulate atmospheric movement, thereby predicting wind power in a future period of time. This method is called wind power prediction based on numerical weather prediction.

[0004] Ultra-short-term prediction of wind power requires predicting wind power of a wind farm in the short term in the future, such as 15 minutes in the future, 2 hours in the future, etc. However, the input data used by the numerical weather prediction model is usually updated every 6 hours, which is not timely, resulting in an inability to accurately achieve ultra-short-term prediction of wind power. SUMMARY

[0005] Embodiments of the present application provide a wind power prediction method, device and readable storage medium, which use multispectral images to perform ultra-short-term prediction of wind power. Since the update level of multispectral images is minute level, the accuracy of ultra-short-term prediction of wind power can be improved.

[0006] In a first aspect, embodiments of the present application provide a wind power prediction method, comprising:

[0007] obtaining a plurality of multispectral images of a target region in a historical time period to obtain a multispectral image sequence, a wind farm being located in the target region;

[0008] determining air flow velocities of different height layers of the target region in the historical time period according to the multispectral image sequence;

[0009] predicting wind speeds of the wind farm in each unit time in a future time period according to the air flow velocities of the different height layers of the target region, a time length between a start point of the future time period and an end point of the historical time period being less than a first preset time length;

[0010] predicting wind power of the wind farm in each unit time in the future time period according to the wind speeds of the wind farm in each unit time in the future time period.

[0011] In a second aspect, an embodiment of the present application provides a wind power prediction device, comprising:

[0012] an acquisition module configured to acquire a plurality of multispectral images of a target area in a historical time period, to obtain a multispectral image sequence, and a wind farm is located in the target area;

[0013] a determination module configured to determine air flow velocities of different height layers of the target area in the historical time period according to the multispectral image sequence;

[0014] a processing module configured to predict a wind speed of the wind farm per unit time in a future time period according to the air flow velocities of the different height layers of the target area, and a time length between a start point of the future time period and an end point of the historical time period is less than a first preset time length;

[0015] a prediction module configured to predict a wind power of the wind farm per unit time in the future time period according to the wind speed of the wind farm per unit time in the future time period.

[0016] In a third aspect, an embodiment of the present application provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, the electronic device implements the method in the first aspect or various possible implementation manners of the first aspect.

[0017] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, and the computer readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the computer instructions are used to implement the method in the first aspect or various possible implementation manners of the first aspect.

[0018] In a fifth aspect, an embodiment of the present application provides a computer program product comprising a computer program, and when the computer program is executed by a processor, the computer program implements the method in the first aspect or various possible implementation manners of the first aspect.

[0019] The wind power prediction method, device, and readable storage medium provided in this application involve an electronic device acquiring multiple frames of multispectral images of a target area within a historical time period to obtain a multispectral image sequence. Based on the multispectral image sequence, the airflow velocity at different altitude layers within the target area during the historical time period is determined. Subsequently, the electronic device predicts the wind speed of the wind farm per unit time within a future time period based on the airflow velocity at different altitude layers in the target area, and then predicts the wind power of the wind farm per unit time within the future time period based on the wind speed of the wind farm per unit time within the future time period. Using this scheme, the electronic device performs ultra-short-term wind power prediction using multispectral images. Since the update level of multispectral images is on the minute level, the accuracy of ultra-short-term wind power prediction can be improved. Furthermore, by dividing the airspace above the target area into different altitude layers and predicting the wind power of the wind farm within a future period based on the wind speed at different altitude layers, the accuracy of ultra-short-term wind power prediction is further improved. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of the network architecture of the wind power prediction method provided in the embodiments of this application;

[0022] Figure 2 This is a flowchart of a wind power prediction method provided in an embodiment of this application;

[0023] Figure 3 This is a frame of multispectral data in the wind power prediction method provided in the embodiments of this application;

[0024] Figure 4 This is another flowchart of the wind power prediction method provided in the embodiments of this application;

[0025] Figure 5 A schematic diagram of the wind power prediction device provided in the embodiments of this application;

[0026] Figure 6 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0027] Wind energy has randomness, volatility and other characteristics, so that the output power of the wind farm has instability, which has a great influence on the normal operation and dispatching of the power system, especially the real-time dispatching of the power system. The ultra-short-term prediction of wind power is a reliable basis for the real-time dispatching of the power grid. At present, the time scale of the ultra-short-term prediction of wind power is generally 0-6 hours in the future, and the prediction time resolution is 5-15 minutes, etc. For example, the wind power every 15 minutes in the next 2 hours is predicted.

[0028] In the common ultra-short-term prediction method of wind power, the historical power, historical wind speed, topography, numerical weather prediction, wind turbine operating state and other data of the wind farm are used as the basis to establish a prediction model of the output power of the wind farm, and the wind speed, power, numerical weather prediction and other data are used as the input of the model, and the equipment state and operating condition of the wind farm unit are combined to perform ultra-short-term prediction of the wind power of the wind farm.

[0029] The above-mentioned traditional ultra-short-term prediction method of wind power seriously depends on the data weather prediction, and the input data used by the numerical weather prediction is usually updated every 6 hours, which has poor timeliness, resulting in that the ultra-short-term prediction of wind power cannot be accurately realized.

[0030] Based on this, the embodiment of the present application provides a wind power prediction method, device and readable storage medium, which uses multi-spectral images to perform ultra-short-term prediction of wind power. Since the update level of the multi-spectral image is minute level, the accuracy of the ultra-short-term prediction of wind power can be improved.

[0031] Figure 1 is a network architecture schematic diagram of the wind power prediction method provided by the embodiment of the present application. Please refer to Figure 1 The network architecture includes a server 11, a collection device 12, a terminal device 13 and the like. The server 11 and the collection device 12 establish a network connection, and the server 11 and the terminal device 13 establish a network connection.

[0032] The server 11 is also called a cloud server, which is deployed with a deep learning model, a regression model and the like, and is used to obtain multi-frame multi-spectral images from the collection device 12, and predict the wind power of the wind farm in the future time period based on the multi-frame multi-spectral images. The server 11 has great computing power, storage capacity and the like. The server 11 can be hardware or software. When the server 11 is hardware, the server 11 is a single server or a distributed server cluster composed of multiple servers. When the server 11 is software, it can be multiple software modules or a single software module, etc., which is not limited by the embodiment of the present application.

[0033] The acquisition device 12 is used to acquire a multispectral image, which can be a meteorological satellite fast imager, a unmanned aerial vehicle multispectral camera, etc., and is used to acquire a plurality of multispectral images of a target region in a historical time period. For example, the acquisition frequency is one multispectral image per minute, and 15 multispectral images of the target region in the past 15 minutes are acquired.

[0034] The terminal device 13 includes, but is not limited to, a field station terminal device in the nacelle of a wind turbine generator unit in a wind farm, a field station terminal device on the ground of a wind farm, a mobile phone, a notebook computer, etc. of a staff of a wind farm, and is used to receive various prompt information from the server 11.

[0035] It should be understood that, Figure 1 The number of servers 11, acquisition devices 12, and terminal devices 13 in

[0036] Next, based on the network architecture shown in Figure 1 , the wind power prediction method described in the embodiments of the present application will be described in detail. For example, please refer to Figure 2 . Figure 2 is a flowchart of the wind power prediction method provided by the embodiments of the present application. The execution subject of the present embodiment is an electronic device, for example, a server in Figure 1 , and the present embodiment includes:

[0037] 201, acquiring a plurality of multispectral images of a target region in a historical time period to obtain a multispectral image sequence, and the wind farm is located in the target region.

[0038] In the embodiments of the present application, the wind power of the wind farm is predicted in a very short time, that is, the size of the effective power that the wind turbine generator unit in the wind farm will capture from the wind energy and convert into the power grid in a future time period. A wind farm deploys a plurality of wind turbine generator units, such as tens or even hundreds of wind turbine generator units. The area of a wind farm ranges from several square kilometers to hundreds of square kilometers, and is related to the terrain condition, the single machine power, the total installed power, etc.

[0039] The target region is a region containing a wind farm and the surrounding area of the wind farm, for example, a wind farm occupies an area of 20 square kilometers, the center of the target region coincides with the center of the wind farm, or the distance between the center of the target region and the center of the wind farm is less than a preset value. The range of the target region is 200 square kilometers. That is, the target region is a larger region centered on the wind farm.

[0040] The embodiment of the present application does not limit the frequency of the acquisition device collecting the multispectral image. For example, the acquisition device collects the multispectral image once per minute, that is, the update frequency of the multispectral image is at the minute level. The historical time period refers to a period of time in the past with a certain time point as the reference, and the length is 10 minutes, 15 minutes, 20 minutes, etc., which is not limited by the embodiment of the present application. For example, the reference time is 6:00 pm, and the historical time period is 5:45 pm to 6:00 pm; for another example, the reference time is 8:30 am, and the historical time period is 8:20 am to 8:30 am. The reference time can be the current time or a time before the current time.

[0041] In the embodiment of the present application, the electronic device obtains one frame of multispectral image per minute, or the length of the historical time period is 15 minutes, and the electronic device obtains the multispectral image once every 15 minutes, and obtains all multispectral images in the 15 minutes at a time. One frame of multispectral image contains data of multiple wavebands at the same time, and the multiple wavebands include but are not limited to visible light waveband, water vapor waveband, thermal infrared waveband, etc.

[0042] 202. Determine the air flow speed of different height layers of the target area in the historical time period according to the sequence of multispectral images.

[0043] After the electronic device obtains the sequence of multispectral images of the target area, the air flow speed of different height layers of the target area in the historical time period is determined, that is, the wind speed of different height layers. Different heights correspond to different temperatures, and different temperatures correspond to different cloud heights. The electronic device divides the height layers from the ground to the top of the cloud, and the number of height layers includes but is not limited to 3 layers, 5 layers, 6 layers, 7 layers, etc. For example, the different height layers are low wind layer, middle wind layer and high wind layer from bottom to top, the height of the low wind layer is 0-3 km, the height of the middle wind layer is 3 km-8 km, and the height of the high wind layer is 8 km-12 km.

[0044] The electronic device determines the wind speed of each layer in the historical time period by using the optical flow method. For example, the historical time period is the past 15 minutes, and the electronic device determines the wind speed of different height layers of the target area in the past 15 minutes.

[0045] 203. According to the air flow speed of different height layers of the target area, the wind speed of the wind farm per unit time in the future time period is predicted, and the length of time between the start point of the future time period and the end point of the historical time period is less than the first preset length.

[0046] In the embodiments of the present application, the length of the future time length, for example, future 5 minutes, future 15 minutes, future 2 hours, future 4 hours or 6 hours, etc. is not limited in the embodiments of the present application. Moreover, the time length between the start point of the future time period and the end point of the historical time period is less than the first preset time length. The first preset time length is, for example, 5 minutes, 10 minutes, etc. For example, the historical time period is from 5:45 pm to 6:00 pm, and the future time period is from 6:00 pm to 8:00 pm. For another example, the historical time period is from 5:45 pm to 6:00 pm, and the future time period is from 6:05 pm to 8:05 pm.

[0047] In the embodiments of the present application, the future time period can be divided into multiple unit times. The length of the unit time is, for example, 5 minutes, 10 minutes, 15 minutes, etc. which is not limited in the embodiments of the present application. For example, the time length of the future time period is 2 hours, and the unit time is 10 minutes. The electronic device predicts the wind speed of the wind farm every 10 minutes in the next 2 hours. In each unit time, the wind speed of the wind farm is the same, and a total of 12 wind speeds are predicted. The wind speed change graph is drawn according to the time sequence, and the change of the wind speed of the wind farm in the future time period is obtained.

[0048] 204. According to the wind speed of the wind farm in each unit time in the future time period, the wind power of the wind farm in each unit time in the future time period is predicted.

[0049] After the electronic device obtains the wind speed of the wind farm in each unit time in the future time period, the wind power of the wind farm in each unit time in the future time period is determined by looking up a table or the like. For example, a wind speed-wind power mapping table is pre-stored on the electronic device, which indicates the relationship between different wind speeds, the scale of the wind farm and the wind power. The electronic device takes the wind speed and the scale of the wind farm as input, and queries the mapping table to predict the wind power of the wind farm in each unit time in the future time period.

[0050] The wind power prediction method provided in the embodiments of the present application, the electronic device obtains a plurality of multispectral images in a target region in a historical time period to obtain a multispectral image sequence, and determines the air flow speed of different height layers in the target region in the historical time period according to the multispectral image sequence. Then, the electronic device predicts the wind speed of the wind farm in each unit time in the future time period according to the air flow speed of different height layers in the target region, and predicts the wind power of the wind farm in each unit time in the future time period according to the wind speed of the wind farm in each unit time in the future time period. By using this scheme, the electronic device uses multispectral images to make super short-term prediction of wind power. Since the update level of the multispectral image is minute level, the accuracy of the super short-term prediction of the wind power can be improved. Moreover, by dividing the space above the target region into different height layers, the wind power of the wind farm in a future period of time is predicted according to the wind speed of different height layers, which further improves the accuracy of the super short-term prediction of the wind power.

[0051] Optionally, in the above embodiment, the electronic device can flexibly predict the wind speed of the wind farm per unit time in the future time period according to the air flow speed of different height layers of the target area.

[0052] In one way, the electronic device determines the wind power density of each height layer according to the air flow speed of the height layer. Then, the electronic device averages the wind power density of each height layer to obtain the wind speed at the height of the wind turbine nacelle. The wind power density is used to indicate the wind energy resource potential of the corresponding height layer. The wind speed at the height of the wind turbine nacelle is used to indicate the overall wind energy of the blade swept area of the wind turbine generator set.

[0053] In another way, the electronic device predicts the cloud evolution over the target area in the future time period according to the air flow speed of different height layers of the target area. The cloud evolution is used to indicate the dynamic change process of the cloud layer over the target area. Then, the electronic device selects a target segment from the cloud evolution and predicts the wind speed of the wind farm per unit time in the future time period according to the target segment. The target segment is used to indicate the dynamic change process of the cloud layer over the wind farm.

[0054] In the embodiment of the present application, the target area is an area much larger than the wind farm, and the wind farm is located in the target area, for example, the wind farm is located in the middle of the target area. That is, the target area includes the wind farm and the surrounding area of the wind farm. The electronic device predicts the cloud evolution over the target area in the future time period according to the air flow speed of different height layers of the target area. Obviously, the cloud evolution over the target area includes the cloud evolution over the wind farm and the cloud evolution over the surrounding area of the wind farm. The cloud evolution over the target area is used to indicate the dynamic change process of the cloud layer over the target area in terms of shape, height, distribution or movement mode.

[0055] After the electronic device obtains the cloud evolution over the target area, it extracts a local cloud evolution from the cloud evolution, which is referred to as a target segment below. The target segment is used to indicate the dynamic change process of the cloud layer over the wind farm. Then, the electronic device predicts the wind speed of the wind farm per unit time in the future time period according to the target segment. For example, the electronic device inputs the target segment into a pre-trained wind speed model, so that the wind speed model outputs the wind speed of the wind farm per unit time in the future time period based on the target segment.

[0056] By using this scheme, the electronic device obtains the cloud evolution according to the wind speed of different height layers of the target area, extracts a local cloud evolution, i.e. a target segment, from the cloud evolution, and predicts the wind speed based on the target segment, which is fast and has high accuracy.

[0057] Optionally, in the above embodiment, in the process of predicting the wind speed of the wind farm at each unit time in the future time period according to the target segment, the electronic device first extracts flow field features from the target segment, the flow field features being used to indicate the characteristics of the wind resources of the wind farm in the future time period. Then, the electronic device predicts the wind speed at the height of the nacelle of the wind turbine in the wind farm at each unit time in the future time period according to the flow field features.

[0058] In the embodiment of the present application, the heights of the nacelles of the wind turbines in a wind farm are at the same level, that is, the heights of the nacelles of the wind turbines are substantially the same. The electronic device extracts the average flow speed, flow speed fluctuation, flow direction and other flow field features of the wind resources above the wind farm from the target segment. Then, the electronic device determines the wind speed at the height of the nacelle of the wind turbine in the wind farm at each unit time in the future time period according to the flow field features by looking up a table or the like.

[0059] By adopting this scheme, the electronic device predicts the wind speed at the height of the nacelle of the wind turbine at each unit time in the future time period by extracting the flow field features, which is fast and has high accuracy.

[0060] The embodiment of the present application does not limit the way in which the electronic device predicts the wind speed at the height of the nacelle of the wind turbine according to the fluid features. In one way, the electronic device normalizes the fluid features, corrects the fluid features according to the height of the nacelle of the wind turbine, and predicts the wind speed at the height of the nacelle of the wind turbine at each unit time in the future time period by using the corrected fluid features.

[0061] In another way, the electronic device deploys a regression model that has been trained in advance. Each time the fluid features are obtained, the electronic device inputs the fluid features into the regression model, so that the regression model outputs the wind speed at the height of the nacelle of the wind turbine in the wind farm at each unit time in the future time period.

[0062] In the process of training the regression model, the electronic device obtains sample data, the sample data including historical features extracted based on historical cloud evolution and historical wind speeds corresponding to the historical cloud evolution. Then, the electronic device trains an initial model according to the sample data by using a machine learning method, and constantly adjusts the parameters of the initial model in the training process until the value of the loss function is minimum. The model at the time when the value of the loss function is minimum is taken as the regression model.

[0063] By adopting this way, the electronic device predicts the wind speed at the height of the nacelle of the wind turbine in the wind farm by using the regression model, thereby achieving the purpose of quickly and accurately determining the wind speed at the height of the nacelle of the wind turbine in the wind farm.

[0064] Optionally, in the above embodiment, the electronic device predicts the wind speed of the wind farm in each unit time in the future time period according to the air flow speed of the different height layers of the target area, and when the wind speed of the first unit time is less than the wind speed of the second unit time and the difference of the wind speed is greater than a preset threshold in the future time period, the electronic device determines that strong convection occurs after the first unit time and sends first prompt information to the terminal device. Wherein, the time length between the first unit time and the second unit time is less than a second preset time length, and the second unit time is located after the first unit time, the time length between the target time point and the starting time point of the second unit time is a third preset time length, and the terminal device is used to control each wind turbine generator in the wind farm.

[0065] In the embodiment of the present application, the wind speed of the wind farm in the future time period refers to the wind power of the wind farm in each unit time in the future time period. Taking the future time period as 2 hours and the unit time as 15 minutes as an example, the wind speed of the wind farm in the future time period is: cutting the future 2 hours into 8 15 minutes, the wind speed of each 15 minutes is 8 wind speeds respectively. Obviously, the wind speed in each unit time is the same; by analogy, the wind power in each unit time is the same.

[0066] After the electronic device predicts the wind speed of the wind farm in each unit time in the future time period, it also judges whether strong convection occurs according to the predicted wind speed. In the judgment process, the electronic device compares the wind speed of the first unit time and the wind speed of the second unit time. The time length between the first unit time and the second unit time is less than a second preset time length, for example, 15 minutes, which is not limited in the embodiment of the present application. The time length between the first unit time and the second unit time refers to the time length between the end point of the first unit time and the starting point of the second unit time.

[0067] Continuing with the above example, assuming the future time period is from 5:00pm to 7:00pm, a total of 8 unit times, unit time 1, unit time 2,..., unit time 8, unit time 1 is 5:00-5:15, unit time 2 is 5:16-5:30,..., the first unit time and the second unit time can be two adjacent unit times, for example, the first unit time is unit time 1, and the second unit time is unit time 2. The first unit time and the second unit time can also be non-adjacent, for example, the first unit time is unit time 3, and the second unit time is unit time 5. When the wind speed difference between the first unit time and the second unit time is greater than the preset threshold, it is determined that strong convection will occur after the first unit time. Then, the electronic device sends the first prompt information to the terminal device to prompt the wind farm to enter the protection mode from the target time point. The target time point must ensure that the wind farm enters the protection mode before the second unit time. For example, the starting point of the first unit time is taken as the target time point; for another example, the middle time of the first unit time is taken as the target time point.

[0068] When the wind speed difference between the first unit time and the second unit time is less than the preset threshold, it is determined that strong convection will not occur after the first unit time.

[0069] With this scheme, the electronic device determines that strong convection will occur in the second unit time according to the wind speed of the first unit time and the second unit time in the future time period, and sends the first prompt information to the terminal device to make the wind farm enter the protection mode in advance, thereby achieving the purpose of improving the safety of the wind farm.

[0070] Optionally, in the above embodiment, the electronic device predicts the wind power of the wind farm in each unit time in the future time period according to the wind speed of the wind farm in each unit time in the future time period, and when the wind power of the third unit time is greater than the wind power of the fourth unit time, and the difference between the wind powers is greater than a preset wind power, it is determined that the wind power will decrease after the third unit time, and the second prompt information is sent to the terminal device. The time length between the third unit time and the fourth unit time is less than a second preset time length, and the fourth unit time is located after the third unit time, and the first prompt information is used to instruct to reduce the load of the wind farm, and the terminal device is used to control each wind turbine generator in the wind farm.

[0071] In the embodiment of the present application, the terminal device is, for example Figure 1The electronic device determines whether the wind power of the wind farm in the fourth unit time decreases. When the difference between the wind power of the third unit time and the wind power of the fourth unit time is greater than the preset wind power, the electronic device determines that the wind power of the wind farm in the fourth unit time decreases, and thus sends second prompt information to the terminal device to prompt the load of the wind farm to be reduced before the fourth unit time. For example, some power-consuming facilities powered by the wind farm, such as manganese-silicon electric arc furnaces, are turned off according to priorities. For another example, the energy storage device is started to power the power-consuming facilities together with the wind farm.

[0072] When the difference between the wind power of the third unit time and the wind power of the fourth unit time is less than or equal to the preset wind power, the electronic device determines that the wind power of the wind farm in the fourth unit time does not decrease substantially, and the power-consuming facilities are continued to be powered by the wind farm in the fourth unit time.

[0073] With this scheme, when the electronic device determines that the wind power decreases in the second unit time according to the wind power of the third unit time and the wind power of the fourth unit time in the future time period, the electronic device sends the second prompt information to the terminal device to prompt the load of the wind farm to be reduced in advance, thereby achieving the purpose of improving the safety of the wind farm.

[0074] Optionally, in the process in which the electronic device determines the air flow speed of different height layers of the target region in the historical time period according to the plurality of multispectral image sequences in the above embodiment, first, the electronic device determines whether the plurality of multispectral image sequences are lost frames. When the plurality of multispectral image sequences are lost frames, a plurality of multispectral images are extracted from the plurality of multispectral image sequences, and the air flow speed of different height layers of the target region in the historical time period is determined according to the plurality of multispectral images. In the plurality of multispectral images, there are a preset number of multispectral images between every two adjacent multispectral images.

[0075] In the embodiment of the present application, the historical time period is, for example, the past 15 minutes, the past 10 minutes, or the like. Taking the historical time period of 15 minutes as an example, assuming that the frequency of the collection device collecting the multispectral images is one frame per minute, theoretically, 15 frames of multispectral images can be collected. Then, the electronic device determines the air flow speed of different height layers according to the adjacent multispectral images. Taking at least three adjacent frames as an example, the adjacent multispectral images include the first frame to the third frame of multispectral images, the second frame to the fourth frame of multispectral images, the third frame to the fifth frame of multispectral images, and the like.

[0076] In practice, frame loss may occur, at which time the acquisition device cannot acquire 15 multispectral images, for example, 14 multispectral images are acquired if the second frame is lost. For another example, 13 multispectral images are acquired if the fourth frame and the sixth frame are lost. In this case, the electronic device performs frame extraction processing. Taking the loss of the second frame as an example, the electronic device extracts the first frame, the third frame, the fifth frame,..., and the thirteenth frame, a total of 7 frames, and determines the air flow velocity at different height layers according to every 3 adjacent multispectral images in the 7 multispectral images. In the 7 multispectral images, the 3 adjacent multispectral images include the first frame, the third frame, and the fifth frame; the third frame, the fifth frame, and the seventh frame, etc.

[0077] If the second frame and the fifth frame are lost, a total of 13 multispectral images are acquired, then the electronic device extracts the first frame, the fourth frame, the seventh frame,..., and the thirteenth frame, a total of 5 multispectral images, and determines the air flow velocity at different height layers according to every 3 adjacent multispectral images in the 5 multispectral images. In the 5 multispectral images, the 3 adjacent multispectral images include the first frame, the fourth frame, and the seventh frame; the fourth frame, the seventh frame, and the tenth frame, etc.

[0078] If the number of lost multispectral images is greater than the preset number of frames, it is considered that the multispectral images in the historical time period are unreliable, and the electronic device reacquires the multispectral images.

[0079] With this scheme, when frame loss occurs in the multispectral image sequence, the electronic device performs frame extraction processing, so that there are a preset number of multispectral images between every two adjacent multispectral images in the extracted multispectral images, and determines the wind speed at different height layers according to the extracted multispectral images, achieving the purpose of improving the accuracy of the predicted wind speed. At the same time, the data processing amount is reduced, and the smooth operation of the electronic device is ensured.

[0080] Optionally, in the above embodiment, in the process of acquiring the multispectral images of the target region in the historical time period to obtain the multispectral image sequence, first, the electronic device acquires multispectral data in the historical time period, and each frame of multispectral data covers a preset range including the target region. Then, the electronic device extracts the multispectral images of the target region from each frame of multispectral data according to the position of the wind farm in the preset range to obtain the multispectral image sequence.

[0081] In the embodiment of the present application, the collection range of the collection device is often large, for example, collecting multi-spectral data of a province, that is, the geographical range corresponding to the multi-spectral data is a province, the preset range is a province, and the multi-spectral image of the target region is essentially local data in the multi-spectral data. In this case, the electronic device extracts the multi-spectral image of the target region from each frame of multi-spectral data according to the position of the wind farm in the preset range, and obtains the multi-spectral image sequence. For example, please refer to Figure 3 .

[0082] Figure 3 is a frame of multi-spectral data in the wind power prediction method provided by the embodiment of the present application. Please refer to Figure 3 , the abscissa is the longitude, the ordinate is the latitude, and the range shown by the two solid lines is the preset range, for example, a province, a prefecture-level city under the jurisdiction of a province, etc. The black filled circle is a wind farm, and the range shown by the dashed line is the target region. In the figure, the center of the target region coincides with the center of the wind farm, and the target region is a range with the center of the wind farm as the center and a preset radius.

[0083] It should be noted that although Figure 3 is described by taking the target region as a circular example. However, the embodiment of the present application is not limited. For example, when the land occupied by the wind farm is irregularly shaped, the shape of the wind farm is enlarged by a preset multiple to obtain the target region, and the multi-spectral image is extracted from the multi-spectral data according to the target region.

[0084] By using this scheme, the electronic device extracts the multi-spectral image of the target region from the multi-spectral data corresponding to the preset range according to the position of the wind farm in the preset range, without collecting the multi-spectral image in units of the target region, which to some extent reduces the cost of the device.

[0085] Figure 4 is another flowchart of the wind power prediction method provided by the embodiment of the present application. The embodiment includes:

[0086] 401, obtaining a plurality of frames of multi-spectral images of a target region to obtain a multi-spectral image sequence.

[0087] For example, the collection device collects multi-spectral images. The server receives the multi-spectral images from the collection device through a network, a satellite, etc. The collection device is, for example, a meteorological satellite rapid imager, and the collection frequency is one frame of multi-spectral image per minute.

[0088] 402, determining the air flow speed of different height layers of the target region in a historical time period based on adjacent multi-spectral images.

[0089] The electronic device determines the air flow velocity of different height layers of the target area based on the adjacent multi-spectrum images by using an optical flow method or the like, and the air flow velocity is a vector, also known as a cloud layer flow vector.

[0090] 403. The cloud evolution over the target area in the future time period is deduced according to the air flow velocity of different height layers.

[0091] In the embodiment of the present application, a deep learning model is pre-deployed on the electronic device, and the deep learning model is obtained by training historical cloud images, and the historical cloud images include cloud images of different spectral bands such as visible light cloud images, infrared cloud images, and water vapor cloud images, that is, cloud images of different height layers, which are a large amount of real historical data. The deep learning model is, for example, a predictive recurrent neural network (PredRNN), a time series generative adversarial network (TSGAN), or the like, and the embodiment of the present application is not limited thereto.

[0092] The electronic device inputs the air flow velocity of different height layers into the deep learning model, and deduces the cloud evolution over the target area in the future time period by using the deep learning model.

[0093] 404. A target segment is selected from the cloud evolution over the target area, flow field features are extracted based on the target segment, and the wind speed at the height of the wind turbine nacelle is predicted according to the flow field features.

[0094] The electronic device selects a target segment above the wind farm, extracts flow field features such as average flow velocity, flow velocity fluctuation, and flow direction based on the target segment, inputs the flow field features into a regression model, so that the regression model outputs the wind speed at the height of the wind turbine nacelle in the wind farm in each unit time in the future time period.

[0095] In the embodiment of the present application, the regression model is a model learned based on historical deduced features and historical nacelle anemometer wind speeds.

[0096] 405. The wind power of the wind farm in each unit time in the future time period is predicted according to the wind speed of the wind farm in each unit time in the future time period.

[0097] For example, the electronic device calculates the wind power of the wind farm in each unit time in the future time period by using a power curve model or the like according to the predicted wind speed.

[0098] The following is an embodiment of the device of the present application, which can be used to execute the method embodiment of the present application. For details not disclosed in the device embodiment of the present application, please refer to the method embodiment of the present application.

[0099] Figure 5A schematic diagram of a wind power prediction device is provided in the embodiments of the present application. The wind power prediction device 500 comprises an acquisition module 51, a determination module 52, a processing module 53, and a prediction module 54.

[0100] The acquisition module 51 is configured to acquire a plurality of multispectral images of a target region in a historical time period, to obtain a multispectral image sequence, and a wind farm is located in the target region.

[0101] The determination module 52 is configured to determine air flow velocities of different height layers of the target region in the historical time period according to the multispectral image sequence.

[0102] The processing module 53 is configured to predict a wind speed of the wind farm per unit time in a future time period according to the air flow velocities of the different height layers of the target region, and a time length between a start point of the future time period and an end point of the historical time period is less than a first preset time length.

[0103] The prediction module 54 is configured to predict a wind power of the wind farm per unit time in the future time period according to the wind speed of the wind farm per unit time in the future time period.

[0104] In a feasible implementation manner, the processing module 53 is configured to predict a cloud evolution in the sky above the target region in the future time period according to the air flow velocities of the different height layers of the target region, the cloud evolution is used to indicate a dynamic change process of a cloud layer in the sky above the target region, select a target segment from the cloud evolution, the target segment is used to indicate a dynamic change process of a cloud layer in the sky above the wind farm, and predict the wind speed of the wind farm per unit time in the future time period according to the target segment.

[0105] In a feasible implementation manner, when the processing module 53 predicts the wind speed of the wind farm per unit time in the future time period according to the target segment, the processing module 53 is configured to extract a flow field feature according to the target segment, the flow field feature is used to indicate a feature of a wind resource of the wind farm in the future time period, and predict the wind speed of a nacelle height of a wind turbine in the wind farm per unit time in the future time period according to the flow field feature.

[0106] In a feasible implementation manner, when the processing module 53 predicts the wind speed of the nacelle height of the wind turbine in the wind farm per unit time in the future time period according to the flow field feature, the processing module 53 is configured to acquire sample data, the sample data comprises historical features extracted based on historical cloud evolutions and historical wind speeds corresponding to the historical cloud evolutions, train a regression model according to the sample data, and input the flow field feature into the regression model, so that the regression model outputs the wind speed of the nacelle height of the wind turbine in the wind farm per unit time in the future time period.

[0107] In a possible implementation, the processing module 53 is further configured to: predict wind speed of the wind farm in each unit time in a future time period according to the air flow speed of different height layers of the target region; when the wind speed in a first unit time is less than the wind speed in a second unit time, and a difference between the wind speeds is greater than a preset threshold in the future time period, determine that strong convection occurs after the first unit time, a time length between the first unit time and the second unit time is less than a second preset time length, and the second unit time is located after the first unit time; control the electronic device to send first prompt information to a terminal device, the first prompt information being used to instruct the wind farm to enter a protection mode from a target time point, a time length between the target time point and a starting time point of the second unit time is a third preset time length, and the terminal device is used to control each wind turbine generator unit in the wind farm.

[0108] In a possible implementation, the processing module 53 is further configured to: predict wind power of the wind farm in each unit time in the future time period according to the wind speed of the wind farm in each unit time in the future time period; when the wind power in a third unit time is greater than the wind power in a fourth unit time, and a difference between the wind powers is greater than a preset wind power in the future time period, determine that the wind power decreases after the third unit time, a time length between the third unit time and the fourth unit time is less than a second preset time length, and the fourth unit time is located after the third unit time; control the electronic device to send second prompt information to a terminal device, the first prompt information being used to instruct to reduce a load of the wind farm, and the terminal device being used to control each wind turbine generator unit in the wind farm.

[0109] In a possible implementation, the acquisition module 51 is configured to: determine whether the multi-spectrum image sequence is lost frames; when the multi-spectrum image sequence is lost frames, extract a plurality of multi-spectrum images from the multi-spectrum image sequence, there are a preset number of multi-spectrum images between each adjacent two multi-spectrum images in the plurality of multi-spectrum images; and determine the air flow speed of different height layers of the target region in the historical time period according to the plurality of multi-spectrum images.

[0110] In a possible implementation, the acquisition module 51 is configured to: acquire a plurality of multi-spectrum data in the historical time period, each frame of multi-spectrum data covers a preset range including the target region; extract a multi-spectrum image of the target region from each frame of multi-spectrum data according to a position of the wind farm in the preset range, to obtain the multi-spectrum image sequence.

[0111] The wind power prediction device provided by the embodiments of the present application can execute the actions of the electronic device in the above embodiments, and the implementation principle and technical effects are similar, and thus will not be described herein again.

[0112] Figure 6 FIG. 1 is a structural schematic diagram of an electronic device provided by an embodiment of the present application, for example, a server or the like. Figure 6 The electronic device 600 described in the embodiments of the present application includes at least one processor 61, at least one communication bus 62, a user interface 63, at least one network interface 65, and a memory 65.

[0113] The communication bus 62 is configured to realize the connection and communication between the components.

[0114] The user interface 63 can include a display screen (Display) and a camera (Camera), and the optional user interface 63 can further include a standard wired interface and a wireless interface.

[0115] The network interface 65 can optionally include a standard wired interface and a wireless interface (such as a WI-FI interface).

[0116] The processor 61 can include one or more processing cores. The processor 61 connects various parts in the entire electronic device 600 by using various interfaces and lines, executes various functions of the electronic device 600 and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 65, and calling data stored in the memory 65. Optionally, the processor 61 can be implemented in at least one of a hardware form of a digital signal processing (DSP), a field-programmable gate array (FPGA), and a programmable logic array (PLA). The processor 61 can be integrated with a combination of one or more of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU is mainly used to process an operating system, a user interface, and an application program; the GPU is used to render and draw a panoramic sphere required to be displayed on the display screen; and the modem is used to process wireless communication. It can be understood that the above-mentioned modem can also not be integrated into the processor 61, but can be implemented by a separate chip.

[0117] The memory 65 can include a random access memory (RAM) and a read-only memory (ROM). Optionally, the memory 65 includes a non-transitory computer-readable storage medium. The memory 65 can be used to store instructions, programs, codes, code sets, or instruction sets. The memory 65 can include a program storage area and a data storage area, where the program storage area can store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playing function, an image playing function, etc.), instructions for implementing the above various method embodiments, etc., and the data storage area can store data involved in the above various method embodiments, etc. The memory 65 can optionally be at least one storage device located away from the processor 61. As shown in FIG. 6, the memory 65, as a computer storage medium, can include an operating system, a network communication module, a user interface module, and an operating application program of the electronic device. Figure 6

[0118] The embodiments of the present application also provide a computer readable storage medium, which stores computer instructions. The computer instructions are executed by a processor to implement the wind power prediction method as described above.

[0119] The embodiments of the present application also provide a computer program product, which contains a computer program. The computer program is executed by a processor to implement the wind power prediction method as described above.

[0120] Those skilled in the art should understand that the embodiments of the present application can provide methods, systems or computer program products. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0121] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams and the combination of the flows and / or blocks can be implemented by computer program instructions. These computer program instructions can be provided to a general purpose computer, a special purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the computer or other programmable data processing device produce a device that implements the functions described in the flowcharts and / or block diagrams.​Figure 1 one or more processes and / or blocks Figure 1 means for performing the function specified by the block or blocks.

[0122] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the Figure 1 one or more processes and / or blocks Figure 1 means for performing the function specified by the block or blocks.

[0123] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the Figure 1 one or more processes and / or blocks Figure 1 means for performing the function specified by the block or blocks.

[0124] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0125] The memory can include non-persistent memory and / or volatile memory, such as random access memory (RAM) and / or cache memory. The memory can also include non-volatile memory, such as read-only memory (ROM), electrically programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), flash memory, or other non-volatile memory. Memory is an example of computer-readable media.

[0126] Computer-readable media includes permanent and non-permanent, removable and non-removable media implemented in any method or technology for storage of information such as computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile discs (DVDs) or other optical storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible to a computing device. According to the definition herein, computer-readable media does not include transitory media, such as modulated data signals and carrier waves.

[0127] It should also be noted that the terms "comprising", "comprises" or other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0128] The above embodiments are only used to illustrate the present application, but not to limit it. Instead of the above, various modifications and changes can be made to the application by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the application shall fall into the scope of the claims of the application.

Claims

1. A method for predicting wind power, characterized in that, Applied to electronic devices, the method includes: Acquire multiple frames of multispectral images of the target area within a historical time period to obtain a multispectral image sequence, wherein the wind farm is located within the target area; The airflow velocity at different altitudes in the target area during the historical time period is determined based on the multispectral image sequence. Based on the air flow velocity at different heights in the target area, the wind speed of the wind farm is predicted for each unit time in the future time period, and the duration between the start of the future time period and the end of the historical time period is less than a first preset duration. Based on the wind speed of the wind farm at each unit time within the future time period, predict the wind power of the wind farm at each unit time within the future time period.

2. The method according to claim 1, characterized in that, The step of predicting the wind speed of the wind farm per unit time within a future time period based on the air flow velocity at different heights in the target area includes: Based on the air flow velocity at different altitudes in the target area, the cloud evolution over the target area is predicted in the future time period, and the cloud evolution is used to indicate the dynamic change process of the clouds over the target area. Select a target segment from the cloud evolution, the target segment being used to indicate the dynamic change process of the clouds above the wind farm; Predict the wind speed of the wind farm for each unit time within a future time period based on the target segment.

3. The method according to claim 2, characterized in that, The step of predicting the wind speed of the wind farm for each unit time within a future time period based on the target segment includes: Flow field features are extracted based on the target segment, and the flow field features are used to indicate the characteristics of the wind resources of the wind farm in the future time period; Based on the flow field characteristics, predict the wind speed at the height of the wind turbine nacelle within the wind farm for each unit time in the future time period.

4. The method according to claim 3, characterized in that, The step of predicting the wind speed at the height of the wind turbine nacelle within the wind farm for each unit time period based on the flow field characteristics includes: Acquire sample data, which includes historical features extracted based on historical cloud evolution and historical wind speeds corresponding to the historical cloud evolution; A regression model is trained based on the sample data; The flow field characteristics are input into the regression model so that the regression model outputs the wind speed at the height of the wind turbine nacelle in the wind farm at each unit time in the future time period.

5. The method according to any one of claims 1 to 4, characterized in that, After predicting the wind speed of the wind farm per unit time within a future time period based on the air flow velocity at different heights in the target area, the method further includes: When the wind speed in the first unit time is less than the wind speed in the second unit time within the future time period, and the difference in wind speed is greater than a preset threshold, it is determined that strong convection occurs after the first unit time, the duration between the first unit time and the second unit time is less than a second preset duration, and the second unit time is after the first unit time. A first prompt message is sent to the terminal device. The first prompt message is used to instruct the wind farm to enter the protection mode from the target time point. The duration between the target time point and the start time point of the second unit time is a third preset duration. The terminal device is used to control each wind turbine generator set in the wind farm.

6. The method according to any one of claims 1 to 4, characterized in that, After predicting the wind power of the wind farm based on the wind speed of the wind farm at each unit time within the future time period, the method further includes: When the wind power in the third unit time is greater than the wind power in the fourth unit time within the future time period, and the difference in wind power is greater than the preset wind power, it is determined that the wind power will decrease after the third unit time. The duration between the third unit time and the fourth unit time is less than the second preset duration, and the fourth unit time is after the third unit time. A second prompt message is sent to the terminal device, the first prompt message being used to instruct the wind farm to reduce its load, and the terminal device being used to control each wind turbine generator unit within the wind farm.

7. The method according to any one of claims 1 to 4, characterized in that, Determining the airflow velocity at different altitudes in the target area within the historical time period based on the multispectral image sequence includes: Determine whether the multispectral image sequence has dropped frames; When the multispectral image sequence loses frames, multiple multispectral images are extracted from the multispectral image sequence. Among the extracted multiple multispectral images, there is a preset number of multispectral images between each pair of adjacent multispectral images. The airflow velocity at different altitudes in the target area during the historical time period is determined based on the multiple multispectral images.

8. The method according to any one of claims 1 to 4, characterized in that, The process of acquiring multiple frames of multispectral images of the target area within a historical time period to obtain a multispectral image sequence includes: Acquire multiple frames of multispectral data within the historical time period, with each frame of multispectral data covering a preset range including the target region; Based on the location of the wind farm within the preset range, the multispectral image of the target area is extracted from each frame of multispectral data to obtain the multispectral image sequence.

9. An electronic device comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it causes the electronic device to implement the method as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 8.

Images