Data processing method and storage medium

Abstract

The present invention discloses a data processing method and storage medium. The method includes: obtaining first weather detection data, wherein the first weather detection data is the actual result obtained by detecting the weather within a first time period; extracting target features from the first weather detection data based on a data prediction model; and converting the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes a weather forecast result obtained by detecting the weather within a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result for the precipitation data within the second time period. The present invention solves the technical problem of low efficiency in predicting weather detection data and precipitation data.

Description

Technical Field

[0001] The present invention relates to the field of computers, and in particular to a data processing method and a storage medium. Background Art

[0002] Currently, weather detection data and precipitation forecasts are typically implemented independently. For example, separate models are used for forecasting. Weather detection data can be predicted using a radar extrapolation model based on cyclic prediction. Due to its iterative prediction approach, prediction time and memory usage increase linearly with the prediction duration. Precipitation forecasts can be implemented using a neural network weather model. These models are all single-task prediction models, wasting time and computing resources, resulting in low efficiency in forecasting weather detection data and precipitation data.

[0003] Currently, no effective solution has been proposed to the above-mentioned technical problem of low efficiency in predicting weather detection data and precipitation data. Summary of the Invention

[0004] Embodiments of the present invention provide a data processing method and a storage medium to at least solve the technical problem of low efficiency in predicting weather detection data and precipitation data.

[0005] According to one aspect of an embodiment of the present invention, a data processing method is provided. The method may include: obtaining first weather detection data, wherein the first weather detection data is an actual result obtained by detecting the weather in a first time period; extracting target features of the first weather detection data based on a data prediction model; and converting the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period.

[0006] According to another aspect of an embodiment of the present invention, a data processing method is also provided. The method may include: responding to a first input instruction on an operation interface, inputting first weather detection data, wherein the first weather detection data is an actual result obtained by detecting the weather in a first time period; responding to a prediction operation instruction on the operation interface, displaying second weather detection data and first target precipitation data on the operation interface, wherein the second weather detection data and the first target precipitation data are obtained by extracting target features of the first weather detection data based on a data prediction model and converting the target features, the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period.

[0007] According to another aspect of an embodiment of the present invention, a data processing method is also provided. The method may include: obtaining first weather detection data from a weather broadcast platform and displaying the first weather detection data on an operation interface, wherein the first weather detection data is an actual result obtained by detecting the weather in a first time period; in response to a prediction operation instruction on the operation interface, displaying second weather detection data and first target precipitation data on the operation interface, wherein the second weather detection data and the first target precipitation data are obtained by extracting target features of the first weather detection data based on a data prediction model and converting the target features, the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period; and returning the second weather detection data and the first target precipitation data to the weather broadcast platform.

[0008] According to one aspect of an embodiment of the present invention, a data processing device is provided. The device may include: a first acquisition unit, configured to acquire first weather detection data, wherein the first weather detection data is an actual result obtained by detecting the weather in a first time period; an extraction unit, configured to extract target features from the first weather detection data based on a data prediction model; and a conversion unit, configured to convert the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period.

[0009] According to another aspect of an embodiment of the present invention, a data processing device is further provided. The device may include: an input unit for inputting first weather detection data in response to a first input instruction applied to an operation interface, wherein the first weather detection data is an actual result obtained by detecting the weather in a first time period; and a first display unit for displaying second weather detection data and first target precipitation data on the operation interface in response to a prediction operation instruction applied to the operation interface, wherein the second weather detection data and the first target precipitation data are obtained by extracting target features of the first weather detection data based on a data prediction model and converting the target features, the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period.

[0010] According to another aspect of an embodiment of the present invention, a data processing device is also provided. The device may include: a second acquisition unit, configured to acquire first weather detection data from a weather broadcast platform and display the first weather detection data on an operation interface, wherein the first weather detection data is an actual result obtained by detecting the weather in a first time period; a second display unit, configured to respond to a forecast operation instruction on the operation interface and display second weather detection data and first target precipitation data on the operation interface, wherein the second weather detection data and the first target precipitation data are obtained by extracting target features of the first weather detection data based on a data forecast model and converting the target features, the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period; and a return unit, configured to return the second weather detection data and the first target precipitation data to the weather broadcast platform.

[0011] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium includes a stored program, wherein when the program is executed by a processor, the device where the computer-readable storage medium is located is controlled to execute the data processing method of an embodiment of the present invention.

[0012] According to another aspect of an embodiment of the present invention, a processor is provided, which can be used to run a program, wherein the program executes the data processing method of the embodiment of the present invention when it is run.

[0013] According to another aspect of an embodiment of the present invention, a data processing system is also provided. The system may include: a processor; and a memory connected to the processor and configured to provide the processor with instructions for performing the following processing steps: obtaining first weather detection data, wherein the first weather detection data is an actual result obtained by detecting weather in a first time period; extracting target features of the first weather detection data based on a data prediction model; and converting the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes a weather forecast result obtained by detecting weather in a second time period after the first time period; and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period.

[0014] In an embodiment of the present invention, first weather detection data is obtained, wherein the first weather detection data is the actual result obtained by detecting the weather in a first time period; target features of the first weather detection data are extracted based on a data prediction model; and the target features are converted into second weather detection data and first target precipitation data, wherein the second weather detection data includes the weather prediction result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes the precipitation prediction result of the precipitation data in the second time period. In other words, the present application uses radar data in a historical time period to simultaneously predict radar data and precipitation data in a future time period based on the same data prediction model, which can achieve the purpose of saving computing resources and reducing prediction time, avoids wasting computing resources and time by separately predicting radar data and precipitation data in a future time period, thereby solving the technical problem of low efficiency in predicting weather detection data and precipitation data, and achieving the technical effect of improving the efficiency of predicting weather detection data and precipitation data. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0016] Figure 1 is a hardware structure block diagram of a computer terminal (or mobile device) for implementing a data processing method according to an embodiment of the present invention;

[0017] Figure 2 is a flow chart of a data processing method according to an embodiment of the present invention;

[0018] Figure 3 is a flow chart of another data processing method according to an embodiment of the present invention;

[0019] Figure 4 is a flow chart of another data processing method according to an embodiment of the present invention;

[0020] Figure 5 is a schematic diagram of data processing of a multi-task model according to an embodiment of the present invention;

[0021] Figure 6A is a schematic diagram of a radar extrapolation and precipitation prediction scenario according to an embodiment of the present invention;

[0022] Figure 6B is a schematic diagram of another radar extrapolation and precipitation prediction scenario according to an embodiment of the present invention;

[0023] Figure 7is a schematic diagram of a data processing device according to an embodiment of the present invention;

[0024] Figure 8 is a schematic diagram of another data processing device according to an embodiment of the present invention;

[0025] Figure 9 is a schematic diagram of another data processing device according to an embodiment of the present invention;

[0026] Figure 10 It is a structural block diagram of a computer terminal according to an embodiment of the present invention. DETAILED DESCRIPTION

[0027] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0028] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0029] First, some nouns or terms that appear in the description of the embodiments of the present application are subject to the following interpretations:

[0030] Radar extrapolation, using historical radar echo sequences to predict / extrapolate future radar sequences;

[0031] Precipitation forecasting, using historical data (radar or precipitation) to predict precipitation sequences at future moments;

[0032] Multi-task: one model can simultaneously learn and predict multiple tasks;

[0033] Convolutional Neural Network (CNN) is a feedforward neural network whose artificial neurons can respond to surrounding units within a certain coverage area.

[0034] Long Short-Term Memory (LSTM) is a time-recursive neural network used to process and predict important events with relatively long intervals and delays in time series.

[0035] Example 1

[0036] According to an embodiment of the present invention, an embodiment of a data processing method is also provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0037] The method embodiment provided in the first embodiment of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1 FIG is a hardware structure block diagram of a computer terminal (or mobile device) for implementing a data processing method according to an embodiment of the present invention. Figure 1 As shown, the computer terminal 10 (or mobile device 10) may include one or more (illustrated as 102a, 102b, ..., 102n) processors 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, a power supply and / or a camera. It will be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 1 More or fewer components than shown, or with Figure 1 Different configurations shown.

[0038] It should be noted that the one or more processors 102 and / or other data processing circuits described above may generally be referred to herein as "data processing circuitry". The data processing circuitry may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuitry may be a single independent processing module, or may be incorporated in whole or in part into any of the other components of the computer terminal 10 (or mobile device). As described in the embodiments of the present application, the data processing circuitry serves as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0039] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the data processing method in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, implements the data processing method of the application described above. The memory 104 may include a high-speed random access memory and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely located relative to the processor 102, and these remote memories may be connected to the computer terminal 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0040] The transmission device 106 is used to receive or send data via a network. A specific example of the aforementioned network may include a wireless network provided by the communications provider of the computer terminal 10. In one embodiment, the transmission device 106 includes a network interface controller (NIC), which can be connected to other network devices via a base station to enable communication with the Internet. In another embodiment, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0041] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the computer terminal 10 (or mobile device).

[0042] It should be noted that, in some optional embodiments, the above Figure 1 The computer device (or mobile device) shown may include hardware elements (including circuits), software elements (including computer code stored on a computer-readable medium), or a combination of hardware elements and software elements. Figure 1This is merely one example of a particular embodiment and is intended to illustrate the types of components that may be present in the aforementioned computer device (or mobile device).

[0043] exist Figure 1 Under the operating environment shown, this application provides Figure 2 It should be noted that the data processing method of this embodiment can be Figure 1 The illustrated embodiment is executed by a mobile terminal.

[0044] Figure 2 FIG. 1 is a flow chart of a data processing method according to an embodiment of the present invention. Figure 2 As shown, the method may include:

[0045] Step S202: Acquire first weather detection data, wherein the first weather detection data is an actual result obtained by detecting the weather in a first time period.

[0046] In the technical solution provided in the above step S202 of the present invention, the first weather detection data within the first time period is obtained. The first weather detection data can be a radar data sequence (radar sequence), which is the actual result obtained by detecting the weather within the above first time period by radar. Among them, the first time period can be the time period before the current time, that is, the historical time period, so the first weather detection data of this embodiment can also be called historical radar data.

[0047] Optionally, the first weather detection data of this embodiment may be first radar data, for example, a radar data sequence within a historical time period, that is, a historical radar data sequence, which may be a radar echo sequence or a radar image sequence, wherein the radar image sequence may include multiple frames of radar images (historical images). For example, the first weather detection data may be a historical radar image sequence within a historical time period of 2 hours. When the time interval is 10 minutes, the historical radar image sequence may include 12 frames of radar images.

[0048] Optionally, the first weather detection data of this embodiment may be remote sensing data obtained by detecting the weather in the first time period through remote sensing detection technology, or infrared data obtained by detecting the weather in the first time period through infrared detection data.

[0049] Step S204: extracting target features of the first weather detection data based on the data prediction model.

[0050] In the technical solution provided in the above step S204 of the present invention, after the first weather detection data is acquired, target features of the first weather detection data may be extracted based on a data prediction model.

[0051] In this embodiment, a data prediction model can be pre-trained, and the first weather detection data can be input into the data prediction model. That is, the first weather detection data can be the input data sequence (input sequence) of the data prediction model, and then the data prediction model can be used to extract target features (features) from the first weather detection data. The target features can be multi-scale features, which can be used to simultaneously predict (forecast) weather detection data and precipitation data.

[0052] In this embodiment, the above-mentioned data prediction model is a multi-task model, which can realize the learning and prediction of multiple tasks at the same time, wherein the multiple tasks may include the task of radar data prediction and the task of precipitation data prediction. Among them, the network structure corresponding to the task of weather detection data prediction in the data prediction model can be trained simultaneously with the network structure corresponding to the task of precipitation data prediction, and they play an auxiliary role with each other.

[0053] Step S206: Convert the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period.

[0054] In the technical solution provided in the above step S206 of the present invention, after the target features of the first weather detection data are extracted based on the data prediction model, the target features can be converted into the second weather detection data and the first target precipitation data. The target features can be converted into the second weather detection data and the first target precipitation data using the above data prediction model, that is, the data sources of the predicted second weather detection data and the first target precipitation data are the same.

[0055] In this embodiment, the second weather detection data may be a radar data prediction result obtained by detecting the weather in a second time period after the first time period using radar, and may be the second radar data. That is, the second weather detection data is radar data predicting the weather in the future time period, which may be referred to as future radar data or predicted radar data. Optionally, the second weather detection data in this embodiment may be a radar data sequence, and thus the second weather detection data may be a future radar data sequence or a predicted radar data sequence. For example, the second weather detection data may be a future radar image sequence within a future time period of 3 hours. When the time interval is 10 minutes, the future radar image sequence may include 18 frames of radar images.

[0056] Optionally, the second weather detection data may be remote sensing data of the weather predicted in the second time period, or infrared data of the weather predicted in the second time period.

[0057] In this embodiment, the first target precipitation data may be the predicted precipitation data for the second time period. That is, the first target precipitation data is the predicted precipitation data for the future time period, which can be referred to as future precipitation data. Alternatively, the first target precipitation data in this embodiment may be a precipitation data sequence, and thus the first target precipitation data may be a future precipitation data sequence or a predicted precipitation data sequence. For example, the first target precipitation data may be a future precipitation image sequence for a future time period of 3 hours, and the prediction may be performed by accumulating the precipitation image sequence 10 minutes at a time.

[0058] In this embodiment, the data prediction model includes a branch that converts the target feature into the second weather detection data and a precipitation branch (Prec Branch) that converts the target feature into the first target precipitation data. That is, the two branches can share the above-mentioned target features, so that one data prediction model can not only predict the second weather detection data in the second time period based on the first weather detection data in the first time period, but also predict the first target precipitation data in the second time period based on the first weather detection data in the first time period. That is, the purposes of radar extrapolation and precipitation prediction can be achieved at the same time.

[0059] Optionally, the above-mentioned data prediction model of this embodiment can be a convolutional neural network model rather than a non-cyclic neural network iterative prediction model, so as to be able to adapt to the prediction of large-scale and long-term second weather detection data (radar data sequence) and first target precipitation data (precipitation data sequence), and shorten the training data prediction model and prediction time.

[0060] In this embodiment, the prediction of weather detection data and the prediction of precipitation data are highly correlated, and the forecast timeliness and required spatiotemporal resolution of the two can be kept highly consistent. In addition, the two can be predicted using the same data source.

[0061] Through the above steps S202 to S206 of the present application, first weather detection data is obtained, wherein the first weather detection data is the actual result obtained by detecting the weather in the first time period; target features of the first weather detection data are extracted based on the data prediction model; and the target features are converted into second weather detection data and first target precipitation data, wherein the second weather detection data includes the weather prediction result obtained by detecting the weather in the second time period after the first time period, and the first target precipitation data includes the precipitation prediction result of the precipitation data in the second time period. In other words, this embodiment uses radar data in the historical time period to simultaneously predict radar data and precipitation data in the future time period based on the same data prediction model, which can achieve the purpose of saving computing resources and reducing prediction time, avoids the waste of computing resources and time caused by separately predicting radar data and precipitation data in the future time period, thereby solving the technical problem of low efficiency in predicting weather detection data and precipitation data, and achieving the technical effect of improving the efficiency of predicting weather detection data and precipitation data.

[0062] The above method of this embodiment is further introduced below.

[0063] As an optional implementation, the data prediction model includes a feature extraction model, a first prediction model, and a second prediction model, wherein the feature extraction model is used to extract features from input data, and the first prediction model and the second prediction model are used to jointly obtain features output by the feature extraction model.

[0064] In this embodiment, the data prediction model may include a feature extraction model, which may be used to extract features from input data, such as extracting target features from the input first weather detection number. The data prediction model of this embodiment may also include a first prediction model and a second prediction model, wherein the first prediction model and the second prediction model are connected to the feature extraction model, and the first prediction model and the second prediction model share the feature extraction model and can jointly obtain the target features output by the feature extraction model.

[0065] As an optional implementation, step S204, extracting target features of the first weather detection data based on the data prediction model, including: extracting target features from the first weather detection data based on the feature extraction model; performing prediction processing on the target features based on the first prediction model to obtain second weather detection data; step S206, converting the target features into second weather detection data and first target precipitation data, including: performing prediction processing on the target features based on the second prediction model to obtain first target precipitation data, wherein the second weather detection data and the first target precipitation data are used to represent the same weather conditions.

[0066] In this embodiment, the feature extraction model may be an encoder, such as an encoder of a deep learning segmentation network model (UNet) or a network pre-training model (U2Net). The encoder may be shared by a branch that converts target features into second-order weather detection data and a precipitation branch that converts target features into first-order precipitation data. In other words, the encoder may be a shared encoder. The branch that converts target features into second-order weather detection data may be a radar branch that converts target features into predicted radar data.

[0067] In this embodiment, the first prediction model in the data prediction model can be a branch in the data prediction model that converts target features into second weather detection data, and can be used to perform predictive processing on the target features to obtain the above-mentioned second weather detection data. That is, the first prediction model of this embodiment is used to predict weather detection data, and the second weather detection data is the prediction result of the first prediction model, and can also be called the output radar data sequence (output radar sequence) of the first prediction model.

[0068] In this embodiment, the second prediction model in the data prediction model can be a precipitation branch in the data prediction model that converts the target features into the second weather detection data, and can be used to perform prediction processing on the target features to obtain the above-mentioned first target precipitation data. That is, the second prediction model of this embodiment is used to predict precipitation data, and the first target precipitation data is the prediction result of the second prediction model, and can also be called the output precipitation data sequence (output prec sequence) of the second prediction model.

[0069] In this embodiment, the first prediction model and the second prediction model share the same feature extraction model, and the first target precipitation data and the second weather detection data can be used to represent the same weather conditions. This can ensure to a certain extent that the weather prediction results of the second weather detection data and the first target precipitation data are consistent. For example, the radar image corresponding to the predicted radar data predicts a high response value in a certain area, which means that there will be heavy precipitation. Correspondingly, the precipitation image corresponding to the first target precipitation data also means that there will be heavy precipitation, thereby avoiding independent prediction of the radar data and precipitation data, which makes it difficult to ensure the consistency of the prediction results of the two.

[0070] As an optional implementation, the first network structure of the first prediction model and the second network structure of the second prediction model are the same, and the first network parameters of the first network structure are different from the second network parameters of the second network structure.

[0071] In this embodiment, the first prediction model can be composed of a first network structure, and the second prediction model can be composed of a second network structure. The first network structure and the second network structure are the same, but the first network parameters of the first network structure are different from the second network parameters of the second network structure, so that the first network structure and the second network structure may not share network parameters. The decoder part of a model such as UNet or U2Net can be used to predict the second weather detection data and the first target precipitation data based on the target features extracted from the first weather detection data.

[0072] The following further introduces the training method of the data prediction model of this embodiment.

[0073] As an optional embodiment, the first sub-model is trained based on the following data to obtain a first prediction model: first training data corresponding to the first prediction model, or the first training data and second training data corresponding to the second prediction model; and / or the second sub-model is trained based on the following data to obtain a second prediction model: second training data corresponding to the second prediction model, or the second training data and the first training data corresponding to the first prediction model.

[0074] Optionally, this embodiment can obtain first training data and train the first sub-model based on the first training data to obtain a first prediction model; obtain second training data and train the second sub-model based on the first training data and the second training data to obtain a second prediction model.

[0075] In this embodiment, the data prediction model may include a feature extraction model, a first prediction model, and a second prediction model. The training of the data prediction model in this embodiment may include a training phase. In the first training phase, multi-task training may be performed, where the feature extraction model, the first prediction model, and the second prediction model are trained simultaneously to ensure consistency in predictions made by the first prediction model and the second prediction model to a certain extent.

[0076] Optionally, this embodiment obtains first training data, which may be samples of weather detection data, such as radar sample data, for training the first sub-model to obtain a first prediction model, wherein the first sub-model may be an initial neural network model related to weather detection data processing, such as an initial neural network model related to radar data processing. This embodiment obtains second training data, which may be precipitation sample data, and trains the second sub-model using the first training data and the second training data to obtain a second prediction model. That is, the first training data of this embodiment may be additional or complementary information for the second prediction model, which can assist in the training of the second prediction model to ensure, to a certain extent, the prediction accuracy and consistency of the prediction results of the first prediction model and the second prediction model, wherein the second sub-model may be an initial neural network model related to precipitation data processing.

[0077] As an optional implementation, the method also includes: obtaining second training data, and training the second sub-model based on the second training data to obtain a second prediction model; obtaining first training data, and training the first sub-model based on the first training data and the second training data to obtain a first prediction model.

[0078] Optionally, this embodiment obtains second training data, which may be precipitation sample data, for training the second sub-model to obtain a second prediction model. This embodiment obtains first training data, which may be radar sample data, and trains the first sub-model using the first and second training data to obtain a first prediction model. That is, the second training data in this embodiment may be additional or complementary information to the first prediction model, thereby assisting in the training of the first prediction model and ensuring, to a certain extent, the consistency of radar data predictions and precipitation data predictions.

[0079] The above-mentioned first prediction model and second prediction model of this embodiment play an auxiliary role to each other during training, that is, the first prediction model can assist and improve the prediction effect of the second prediction model because the first training data provides additional or complementary information for the second prediction model, and the second prediction model can assist and improve the prediction effect of the first prediction model because the second training data provides additional or complementary information for the first prediction model, and the first prediction model and the second prediction model are trained at the same time, which ensures the consistency of the predictions of the first prediction model and the second prediction model to a certain extent.

[0080] This embodiment can simultaneously realize the prediction of weather detection data and precipitation data through the data prediction model which is a multi-task model, thereby saving computing resources, reducing prediction time, and maintaining the consistency of the prediction results of the two.

[0081] As an optional implementation, actual detection data is obtained, wherein the actual detection data is the actual result obtained by detecting the weather in the second time period and / or the actual result of precipitation data in the second time period; and the parameters of the first prediction model and / or the second prediction model are adjusted based on the actual detection data and the corresponding loss function.

[0082] In this embodiment, after converting the target features into second weather detection data and first target precipitation data within the second time period, actual detection data can be obtained, and the first prediction model and / or parameters can be fine-tuned or further optimized based on the newly obtained actual detection data. The actual detection data can be the actual results obtained by detecting the weather within the second time period, or the actual results of precipitation data within the second time period. The parameters of the first prediction model and / or the second prediction model can be adjusted based on the above actual detection results, the corresponding loss function, and the corresponding prediction results. This is further described below.

[0083] As an optional implementation, the method also includes: obtaining third weather detection data, wherein the third weather detection data is the actual result obtained by detecting the weather within the second time period; and adjusting the parameters of the first prediction model based on the third weather detection data, the second weather detection data and the first loss function.

[0084] In this embodiment, since the accuracy of the prediction results of the first prediction model can be guaranteed in the above-mentioned first training stage, in order to further improve the visualization effect of the prediction results and avoid blurring of the image corresponding to the weather detection data, for example, avoiding blurring of the image corresponding to the radar data, this embodiment designs a second training stage. This embodiment can fix the parameters of the feature extraction model of the data prediction model, can adjust the parameters of the first prediction model separately, and can use strategies such as generative adversarial networks and / or loss functions (loss) to adjust the parameters of the first prediction model to improve the degree of refinement of the weather detection data predicted by the first prediction model and the visualization effect of the corresponding image, wherein the above-mentioned adjustment can be fine-tuning and the loss function can be a perceptual loss function.

[0085] This embodiment can obtain third weather detection data, for example, obtain third radar data, and the third radar data is the actual result obtained by detecting the weather in the second time period by radar, and can be a real radar data sequence in the second time period. This embodiment can adjust the parameters of the first prediction model based on the third radar data, the second weather detection data and the first loss function. Optionally, this embodiment uses a discriminator to distinguish the true from the false of the above-mentioned second radar data (predicted radar data sequence) and the above-mentioned third radar data (real radar data sequence), and at the same time uses the first loss function to further adjust the parameters of the first prediction model to improve the authenticity of the prediction results of the first prediction model, wherein the first loss function can be a perceptual loss, which can include feature matching loss and style loss, to further improve the authenticity of the prediction results of the first prediction model, avoid the problem of blurring of the image corresponding to the first weather detection data predicted by the first prediction model, and improve the visualization effect of the image corresponding to the second weather detection data predicted by the first prediction model.

[0086] As an optional implementation, the method also includes: obtaining second target precipitation data, wherein the second target precipitation data is the actual result of precipitation data within a second time period; and adjusting the parameters of the second prediction model based on the second target precipitation data, the first target precipitation data and the second loss function.

[0087] In this embodiment, since the accuracy of the prediction results of the second prediction model can be guaranteed in the above-mentioned first training stage, in order to further improve the visualization effect of the prediction results and avoid blurring of the image corresponding to the precipitation data, this embodiment can fix the parameters of the feature extraction model of the data prediction model in the second training stage, and can adjust the parameters of the second prediction model separately. The parameters of the second prediction model can be adjusted using strategies such as generative adversarial networks and / or loss functions (loss) to improve the degree of refinement of the precipitation data predicted by the second prediction model and the visualization effect of the corresponding image, wherein the above-mentioned adjustment can be fine-tuning, and the loss function can be a perceptual loss function.

[0088] This embodiment can obtain second target precipitation data, which is the actual result of precipitation data in the second time period, and can be a real precipitation data sequence in the second time period. This embodiment can adjust the parameters of the second prediction model based on the second target precipitation data, the first target precipitation data and the second loss function. Optionally, this embodiment uses a discriminator to distinguish the true and false of the above-mentioned first target precipitation data (predicted precipitation data sequence) and the above-mentioned second target precipitation data (real precipitation data sequence), and at the same time uses the second loss function to further adjust the parameters of the second prediction model to improve the authenticity of the prediction results of the second prediction model, wherein the second loss function can be a perceptual loss, which can include feature matching loss and style loss, to further improve the authenticity of the prediction results of the second prediction model, avoid the problem of blurring of the image corresponding to the precipitation data predicted by the second prediction model, and improve the visualization effect of the image corresponding to the first target precipitation data predicted by the second prediction model.

[0089] As an optional implementation, the first target precipitation data includes multiple precipitation prediction results of precipitation data within the second time period, and the method further includes: displaying each precipitation prediction result in a graded manner according to the precipitation level of each precipitation prediction result on the operation interface.

[0090] In this embodiment, after the data prediction model is trained, target features of the first weather detection data are extracted based on the data prediction model, and the target features are converted into second weather detection data and first target precipitation data. Multiple precipitation prediction results can then be graded and fed back to the user interface. Optionally, this embodiment can determine a precipitation level for each precipitation prediction result, and on the user interface, each precipitation prediction result can be graded and displayed according to the precipitation level, thereby allowing the user to clearly understand each precipitation prediction result.

[0091] As an optional implementation, the method further includes: determining indication information for each precipitation forecast result, wherein the indication information is used to indicate the precipitation level of each precipitation forecast result; and displaying each precipitation forecast result in a graded manner, including: displaying each precipitation forecast result in a graded manner according to the indication information.

[0092] In this embodiment, each precipitation forecast result has indication information, which can be a gradient color to represent different precipitation levels. Therefore, this embodiment can display each precipitation forecast result in a graded manner on the operation interface according to the color corresponding to the precipitation level. That is, the precipitation level and color of each precipitation forecast result in this embodiment have a corresponding relationship, so that the user can clearly understand each precipitation forecast result.

[0093] As an optional implementation, the first target precipitation data includes multiple precipitation prediction results of precipitation data within the second time period, and the method further includes: displaying each precipitation prediction result on the operation interface according to the area corresponding to each precipitation prediction result.

[0094] In this embodiment, the region (area) corresponding to each precipitation forecast result can be determined. That is, the precipitation forecast result is the result obtained by predicting precipitation data for a certain region. Each precipitation forecast result can be displayed on the operation interface according to the region corresponding to each precipitation forecast result. For example, different display positions on the operation interface correspond to different regions, and the corresponding precipitation forecast result is displayed at the display position corresponding to the region, thereby achieving the purpose of partitioning and displaying multiple precipitation forecast results. Optionally, this embodiment can further display each precipitation forecast result according to the precipitation level corresponding to each precipitation forecast result for different regions, so that the user can clearly understand each precipitation forecast result.

[0095] As an optional implementation, the method also includes: displaying the precipitation probability corresponding to the precipitation forecast result on the operation interface; displaying prompt information on the operation interface based on the precipitation probability, wherein the prompt information is used to prompt a travel strategy corresponding to the precipitation probability.

[0096] In this embodiment, after converting the target features into the second weather detection data and the first target precipitation data, the precipitation probability corresponding to each precipitation prediction result included in the first target precipitation data can be obtained and displayed on the operation interface. Optionally, when the precipitation probability corresponding to each precipitation prediction result is greater than a target threshold, a prompt message can be displayed on the operation interface. This prompt message is used to indicate a travel strategy corresponding to the precipitation probability, thereby reminding the user. For example, if the user needs to travel to an area where the precipitation probability is greater than the target threshold, it is necessary to bring rain gear, etc.

[0097] As an optional implementation, the method also includes: obtaining adjustment information, wherein the adjustment information includes information used to characterize the accuracy of the precipitation prediction results and / or environmental parameters of the area corresponding to the precipitation prediction results; and adjusting the data prediction model based on the adjustment information.

[0098] In this embodiment, the data prediction model can be further adjusted. This embodiment can obtain adjustment information. For a user with some meteorological knowledge, when seeing the precipitation forecast result, the user can further determine the accuracy of the precipitation forecast result, or obtain regional environmental parameters, such as the temperature of the region, whether it is sunny or cloudy, etc., and adjust the data prediction model based on the obtained information to further iteratively optimize the data prediction model, thereby improving the accuracy of the prediction made by the data prediction model.

[0099] The embodiment of the present invention also provides another data processing method from the perspective of human-computer interaction.

[0100] Figure 3 FIG. 1 is a flow chart of another data processing method according to an embodiment of the present invention. Figure 3 As shown, the method may include the following steps:

[0101] Step S302 , in response to a first input instruction on the operation interface, first weather detection data is input, wherein the first weather detection data is an actual result obtained by detecting the weather in a first time period.

[0102] In the technical solution provided in the above step S302 of the present invention, the operation interface can be a human-computer interaction interface on the front-end client. The user can trigger the operation interface to generate a first input instruction. The first input instruction is used to input first weather detection data, thereby responding to the first input instruction and obtaining first weather detection data within a first time period. The first weather detection data can be first radar data, for example, a radar data sequence, which is the actual result obtained by detecting the weather within the above-mentioned first time period.

[0103] Optionally, the first weather detection data of this embodiment may be a radar data sequence within a historical time period, and may include multiple frames of radar images (historical images). For example, the first weather detection data may be a historical radar image sequence within a historical time period of 2 hours. When the time interval is 10 minutes, the historical radar image sequence may include 12 frames of radar images.

[0104] Optionally, the first weather detection data of this embodiment may be remote sensing data obtained by detecting the weather in the first time period through remote sensing detection technology, or infrared data obtained by detecting the weather in the first time period through infrared detection data.

[0105] Step S304, in response to the prediction operation instruction on the operation interface, the second weather detection data and the first target precipitation data are displayed on the operation interface, wherein the second weather detection data and the first target precipitation data are obtained by extracting the target features of the first weather detection data based on the data prediction model and converting the target features.

[0106] In the technical solution provided in the above step S304 of the present invention, the second weather detection data includes the weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes the precipitation forecast result of the precipitation data in the second time period.

[0107] In this embodiment, a user can trigger a forecasting operation instruction on the operation interface, which is used to forecast second weather detection data and first target precipitation data within a second target time period. In response to the forecasting operation instruction on the operation interface, the second weather detection data and first target precipitation data can be displayed on the operation interface.

[0108] In this embodiment, a data prediction model can be pre-trained, the first weather detection data can be input into the data prediction model, and then the data prediction model can be used to extract target features from the first weather detection data for simultaneous prediction of the second weather detection data and the first target precipitation data. The second weather detection data can be the second radar data, and the data prediction model is a multi-task model that can simultaneously realize the learning and prediction of multiple tasks. The data prediction model may include a branch that converts the target feature into the second weather detection data and a precipitation branch that converts the target feature into the first target precipitation data. That is, the two branches can share the above-mentioned target features, thereby achieving the purpose of not only predicting the second weather detection data in the second time period based on the first weather detection data in the first time period, but also predicting the first target precipitation data in the second time period based on the first weather detection data in the first time period through a data prediction model. For example, the purpose of radar extrapolation and precipitation prediction can be achieved simultaneously.

[0109] Optionally, the above-mentioned data prediction model of this embodiment can be a convolutional neural network model rather than a non-cyclic neural network iterative prediction model, so as to be able to adapt to the prediction of second weather detection data and first target precipitation data over a large range and for a long time, and shorten the training data prediction model and prediction time.

[0110] In this embodiment, the second weather detection data may be a weather forecast result obtained by detecting the weather in a second time period after the first time through radar. That is, the second weather detection data is radar data of the weather in the predicted future time period. For example, the second weather detection data may be a future radar image sequence within a future time period of 3 hours. When the time interval is 10 minutes, the future radar image sequence may include 18 frames of radar images.

[0111] In this embodiment, the above-mentioned first target precipitation data can be the prediction result of the precipitation data in the second time period, that is, the first target precipitation data is the predicted precipitation data of the weather in the future time period. The first target precipitation data can be the future precipitation image sequence within 3 hours in the future time period, and the precipitation image sequence can be accumulated every 10 minutes for prediction.

[0112] The above method of this embodiment is further introduced below.

[0113] As an optional embodiment, the method also includes: responding to a second input instruction on the operation interface, inputting third weather detection data, for example, third radar data, which is the actual result obtained by detecting the weather in the second time period by radar; responding to a first adjustment instruction on the operation interface, adjusting the parameters of the first prediction model based on the third weather detection data, the second weather detection data and the first loss function, wherein the first prediction model is used to perform prediction processing on the target features to obtain the second weather detection data.

[0114] In this embodiment, the user can trigger a second input command on the operation interface to input third radar data, which is the actual results of radar weather monitoring during the second time period. Optionally, if the difference between the third radar data and the second weather data is greater than a first target threshold, the parameters of the first prediction model can be adjusted to further improve the accuracy and visual display of the first prediction model's prediction results.

[0115] Optionally, this embodiment can respond to a first adjustment instruction on the operation interface and adjust the parameters of the first prediction model based on the third radar data, the second weather detection data and the first loss function. Optionally, this embodiment uses a discriminator to distinguish the second weather detection data (predicted radar data sequence) and the third weather detection data (real radar data sequence) as true or false, and at the same time uses the first loss function to further adjust the parameters of the first prediction model to improve the authenticity of the prediction results of the first prediction model.

[0116] As an optional embodiment, the method also includes: displaying a first clarity of the radar image corresponding to the second weather detection data on the operation interface; responding to a first adjustment instruction acting on the operation interface, including: responding to the first adjustment instruction acting on the operation interface when the first clarity is less than a first threshold.

[0117] In this embodiment, after the target features are converted into the second weather detection data through the data prediction model, an image corresponding to the second weather detection data can be displayed on the operation interface, and the first clarity of the image can be displayed. Optionally, this embodiment determines whether the first clarity is less than a first threshold value, which can be a critical threshold value for determining whether the image is blurred. If it is determined that the first clarity is less than the first threshold value, that is, the image is blurred, it is necessary to adjust the parameters of the first prediction model, and the first adjustment instruction acting on the operation interface can be responded to, and then the parameters of the first prediction model are adjusted based on the third radar data, the second weather detection data, and the first loss function. Optionally, if it is determined that the first clarity is not less than the first threshold value, that is, the radar image is clear, it is not necessary to adjust the parameters of the first prediction model, and it is not necessary to respond to the first adjustment instruction acting on the operation interface.

[0118] As an optional embodiment, the method also includes: responding to a third input instruction on the operation interface, inputting second target precipitation data, wherein the second target precipitation data is the actual result of precipitation data within a second time period; responding to a second adjustment instruction on the operation interface, adjusting the parameters of the second prediction model based on the second target precipitation data, the first target precipitation data and the second loss function, wherein the second prediction model is used to perform prediction processing on the target features to obtain the first target precipitation data.

[0119] In this embodiment, the user can trigger a third input command on the operation interface. This third input command can be used to input second target precipitation data, which is the actual precipitation data within the second time period. Optionally, if the difference between the second target precipitation data and the first target precipitation data is greater than a second target threshold, the parameters used for the second prediction model can be adjusted to further improve the accuracy and visual display effect of the second prediction model's prediction results.

[0120] This embodiment can respond to a third input instruction on the operation interface and input second target precipitation data, where the second target precipitation data represents the actual precipitation data within the second time period. This embodiment can respond to a second adjustment instruction on the operation interface and adjust the parameters of the second prediction model based on the second target precipitation data, the first target precipitation data, and the second loss function. Optionally, this embodiment uses a discriminator to distinguish the first target precipitation data (predicted precipitation data sequence) from the second target precipitation data (actual precipitation data sequence), and simultaneously uses a second loss function to further adjust the parameters of the second prediction model to improve the authenticity of the prediction results of the second prediction model.

[0121] As an optional embodiment, the method also includes: displaying a second clarity of the precipitation image corresponding to the first target precipitation data on the operation interface; responding to a second adjustment instruction acting on the operation interface, including: responding to the second adjustment instruction acting on the operation interface when the second clarity is less than a second threshold.

[0122] In this embodiment, after the target feature is converted into first target precipitation data through the data prediction model, a precipitation image corresponding to the first target precipitation data can be displayed on the operation interface, and a second clarity of the precipitation image can be displayed. Optionally, this embodiment determines whether the second clarity is less than a second threshold, which can be a critical threshold for determining whether the precipitation image is blurred. If it is determined that the second clarity is less than the second threshold, that is, the precipitation image is blurred, it is necessary to adjust the parameters of the second prediction model. The second adjustment instruction on the operation interface can be responded to, and the parameters of the second prediction model can be adjusted based on the second target precipitation data, the first target precipitation data, and the second loss function. Optionally, if it is determined that the second clarity is not less than the second threshold, that is, the precipitation image is clear, it is not necessary to adjust the parameters of the second prediction model, and the second adjustment instruction on the operation interface can be ignored.

[0123] The embodiment of the present invention also provides another data processing method applied to a weather broadcast platform.

[0124] Figure 4 FIG. 1 is a flow chart of another data processing method according to an embodiment of the present invention. Figure 4 As shown, the method may include:

[0125] Step S402: obtaining first weather detection data from the weather broadcast platform and displaying the first weather detection data on the operation interface, wherein the first weather detection data is an actual result obtained by detecting the weather in the first time period.

[0126] In the technical solution provided in step S402 of the present invention, the first weather detection data can come from a weather broadcast platform, which can be used to obtain weather detection data, broadcast predicted weather detection data and precipitation data, and analyze the weather detection data and precipitation data. In this embodiment, the first weather detection data within a first time period can be obtained from the weather broadcast platform. The first weather detection data can be a radar data sequence, which is the actual result obtained by detecting the weather within the first time period. Optionally, the first weather detection data in this embodiment can be a radar data sequence within a historical time period, which can include multiple frames of radar images.

[0127] Step S404 , in response to the forecasting operation instruction on the operation interface, the second weather detection data and the first target precipitation data are displayed on the operation interface.

[0128] In the technical solution provided in the above step S404 of the present invention, the second weather detection data and the first target precipitation data are obtained by extracting the target features of the first weather detection data based on the data prediction model and converting the target features. The second weather detection data includes the weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes the precipitation forecast result of the precipitation data in the second time period.

[0129] In this embodiment, a user can trigger a forecasting operation instruction on the operation interface, which is used to forecast second weather detection data and first target precipitation data within a second target time period. In response to the forecasting operation instruction on the operation interface, the second weather detection data and first target precipitation data can be displayed on the operation interface.

[0130] In this embodiment, a data prediction model can be pre-trained, the first weather detection data can be input into the data prediction model, and then the target features can be extracted from the first weather detection data through the data prediction model to simultaneously predict the second weather detection data and the first target precipitation data. The data prediction model is a multi-task model that can simultaneously realize the learning and prediction of multiple tasks. The data prediction model may include a branch that converts the target features into the second weather detection data and a precipitation branch that converts the target features into the first target precipitation data. That is, the two branches can share the above-mentioned target features, thereby achieving the purpose of not only predicting the second weather detection data in the second time period based on the first weather detection data in the first time period, but also predicting the first target precipitation data in the second time period based on the first weather detection data in the first time period through a data prediction model, that is, the purpose of radar extrapolation and precipitation prediction can be achieved simultaneously.

[0131] Step S406: Return the second weather detection data and the first target precipitation data to the weather broadcast platform.

[0132] In the technical solution provided in the above step S406 of the present invention, after the second weather detection data and the first target precipitation data are displayed on the operation interface, the second weather detection data and the first target precipitation data may be returned to the weather broadcast platform.

[0133] This embodiment returns the second weather detection data and the first target precipitation data to the weather broadcast platform, which can display the second weather detection data and the first target precipitation data on the weather broadcast platform and further analyze the second weather detection data and the first target precipitation data.

[0134] As an optional example, the data processing method of this embodiment may include: determining a first target area where the target vegetation is located; obtaining first weather detection data of the first target area, wherein the first weather detection data is the actual result obtained by detecting the weather in the first target area within a first time period; extracting target features of the first weather detection data based on a data prediction model; converting the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period; and determining whether to fertilize the target vegetation based on the second weather detection data and the first target precipitation data.

[0135] In this embodiment, the above method can be applied to scenarios requiring vegetation management, such as agricultural scenarios. Optionally, a first target area where the target vegetation is located is determined, and the target vegetation can be a crop. This embodiment can obtain the first weather detection data of the first target area within a first time period. In this embodiment, a data prediction model can be pre-trained, and the first weather detection data can be input into the data prediction model. Then, the target features can be extracted from the first weather detection data through the data prediction model to simultaneously predict the second weather detection data and the first target precipitation data within a second target time period, wherein the second time period can be the next 3 hours, and there is no specific limitation here. After obtaining the second weather detection data and the first target precipitation data, it can be determined whether to fertilize the target vegetation based on the second weather detection data and the first target precipitation data. For example, it can be determined whether the second weather detection data is greater than a data threshold and whether the first target precipitation data is greater than a precipitation threshold. If so, it can be determined that the target vegetation needs to be fertilized.

[0136] Optionally, this embodiment may continue to predict the precipitation in the second time period in the future in the area where the target vegetation is located through the above method to determine whether fertilization of the target vegetation is currently required.

[0137] As an optional embodiment, when determining to fertilize the target vegetation based on the second weather detection data and the first target precipitation data, the method also includes: outputting a prompt message, wherein the prompt message is used to prompt to fertilize the target vegetation; and / or outputting a fertilization control instruction to the fertilization equipment, wherein the fertilization control instruction is used to control the fertilization equipment to perform fertilization operations on the target vegetation.

[0138] In this embodiment, if it is determined based on the second weather detection data and the first target precipitation data that the target vegetation is to be fertilized, a prompt message can be output. The prompt message can be a voice prompt message, or it can be a prompt message such as text or icon displayed on the operation interface, and no specific limitation is made here. The prompt message of this embodiment can be used to prompt the target vegetation to be fertilized, and the user can then manually fertilize the target vegetation. Optionally, when it is determined based on the second weather detection data and the first target precipitation data that the target vegetation is to be fertilized, this embodiment can output a fertilization control instruction to the fertilization equipment. The fertilization equipment can respond to the fertilization control instruction and control the fertilization equipment to automatically fertilize the target vegetation, thereby greatly improving the efficiency of managing the target vegetation.

[0139] As an optional example, the data processing method of this embodiment may include: obtaining first weather detection data of a second target area, wherein the first weather detection data is the actual result obtained by detecting the weather in the second target area within a first time period; extracting target features of the first weather detection data based on a data prediction model; converting the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period; determining whether to push drainage strategy data based on the second weather detection data and the first target precipitation data, wherein the drainage strategy data is used to represent a strategy for draining the second target area.

[0140] The above method of this embodiment can be applied to the scenario of urban drought and flood prediction. This embodiment determines the first weather detection data within the first time period of the second target area, and the second target area can be any urban area that needs to be predicted for drought and flood. In this embodiment, a data prediction model can be pre-trained, and the first weather detection data can be input into the data prediction model. Then, the target features can be extracted from the first weather detection data through the data prediction model to simultaneously predict the second weather detection data and the first target precipitation data within the second target time period, wherein the second time period can be the next 3 hours, and there is no specific limitation here. After obtaining the second weather detection data and the first target precipitation data, it can be determined whether to push the drainage strategy data based on the second weather detection data and the first target precipitation data. The drainage strategy data can be used to represent the strategy for draining the second target area, and can be any strategy that can achieve drainage, such as waterway inspection, flood dredging measures, etc., thereby improving the safety of urban management.

[0141] As an optional example, the data processing method of this embodiment may include: obtaining first weather detection data of a third target area, wherein the first weather detection data is the actual result obtained by detecting the weather in the third target area within a first time period; extracting target features of the first weather detection data based on a data prediction model; converting the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period; determining the water level of the water area in the third target area based on the second weather detection data and the first target precipitation data; and determining whether to push flood control strategy data based on the water level, wherein the flood control strategy data is used to represent a strategy for flood control in the third target area.

[0142] The above method of this embodiment can be applied to water conservancy-related scenarios. This embodiment determines first weather detection data for a third target area within a first time period. The third target area includes water, such as a river. In this embodiment, a data prediction model can be pre-trained, and the first weather detection data can be input into the data prediction model. The data prediction model can then extract target features from the first weather detection data to simultaneously predict second weather detection data and first target precipitation data for a second target time period. The second time period can be the next three hours, and there is no specific limitation here. After obtaining the second weather detection data and the first target precipitation data, the water level of the water area in the third target area that may rise can be determined based on the second weather detection data and the first target precipitation data. It can be determined whether the water level reaches a water level threshold. If it is determined that the water level reaches the water level threshold, some flood prevention measures need to be taken, and flood prevention strategy data is determined to be pushed. Otherwise, it is determined not to push the flood prevention strategy data. The flood prevention strategy data can be any strategy used to represent flood prevention for the third target area.

[0143] The above method of this embodiment can be applied to short-term weather scenarios. By constructing a data prediction model and using the same data source for prediction, the prediction of radar data and precipitation data can be achieved simultaneously. The prediction timeliness and required spatiotemporal resolution of the two are highly consistent, achieving the purpose of simultaneous radar extrapolation and precipitation prediction. Among them, the data prediction model is designed with two training stages. The first training stage is multi-task simultaneous training, which can make the training of the first prediction model and the second prediction model of the data prediction model mutually assist each other, improve the prediction effect, and ensure the consistency of the prediction results of the two; the second training stage is to adjust the parameters of the first prediction model and / or the second prediction model to improve the refinement and visualization effect of the prediction results, avoid the waste of computing resources and time caused by separate prediction of radar data and precipitation data in the future time period, thereby solving the technical problem of low efficiency in predicting weather detection data and precipitation data, and achieving the technical effect of improving the efficiency of predicting weather detection data and precipitation data.

[0144] Example 2

[0145] The preferred implementation of the above method of this embodiment is further introduced below, specifically using the scenario of short-term weather.

[0146] In short-term weather application scenarios, radar extrapolation products and precipitation forecast products are often required at the same time, and these two products are highly correlated. The prediction timeliness and required temporal and spatial resolution of the two products are highly consistent. In addition, the two products can be predicted using the same data source.

[0147] In related technologies, radar extrapolation and precipitation prediction are usually separated and forecasted using independent models. For example, radar extrapolation can use a radar extrapolation model based on a long short-term memory network for cyclic prediction, such as convolutional LSTM (ConvLSTM), spatiotemporal sequence prediction model (PredRNN), spatiotemporal sequence prediction (PredRNN++), strong memory model (E3D LSTM), MIM, machine learning classifier (MLC-LSTM, etc.). Since the above methods use an iterative prediction method, the prediction time and video memory consumption increase linearly with the increase in prediction time, which makes it difficult to meet the needs of large-scale and long-term sequence prediction. The precipitation prediction model can be a neural weather model (MetNet, RainNet).

[0148] However, the above models are all single-task prediction models (radar extrapolation or precipitation prediction), which not only wastes time and computing resources, but also makes it difficult to maintain the consistency of the prediction results of the two.

[0149] However, this embodiment establishes a multi-task model that can simultaneously implement radar extrapolation and precipitation prediction. The method of this embodiment is further described below.

[0150] This embodiment can use historical radar data sequences to simultaneously predict future radar data sequences and precipitation data sequences. For example, a radar image sequence of 2 hours in a historical time period (a total of 12 frames of historical images when the time interval is 10 minutes) is used to predict a radar image sequence of 3 hours (a total of 18 frames) in a future time period and a 10-minute cumulative precipitation sequence.

[0151] The model structure of this embodiment is introduced below.

[0152] Figure 5 FIG. 1 is a schematic diagram of data processing of a multi-task model according to an embodiment of the present invention. Figure 5 As shown, this embodiment can adopt a multi-task approach, allowing a single model to simultaneously predict radar data and precipitation data. In this embodiment, historical radar data sequences can be input into a shared encoder to extract multi-scale features. The encoder can be the encoder portion of a model such as UNet or U2Net. The multi-scale features are then input into two independent branches. The radar branch is used to predict radar data sequences in the future time period, and the precipitation branch is used to predict precipitation data sequences in the future time period.

[0153] The above-mentioned two branches of this embodiment share the same encoder, which can ensure the consistency of the prediction results of the radar branch and the precipitation branch to a certain extent. For example, the multi-task model can predict a high response value in a certain area of ​​the radar image, which means heavy precipitation in the precipitation image.

[0154] The radar branch and precipitation branch of this embodiment can maintain the same network structure, but do not share network parameters, and can adopt the decoder part of models such as UNet or U2Net.

[0155] This embodiment can use two training phases to sequence a multi-task model. In the first training phase, multi-task training can be performed. The entire model, including a shared encoder, radar branch, and precipitation branch, can be trained simultaneously. Multi-task training enables the radar branch and precipitation branch to complement each other. That is, the precipitation branch can assist and improve the prediction performance of the radar branch because the precipitation data sequence of the precipitation branch can provide additional or complementary information to the radar branch. Furthermore, simultaneous training can also ensure, to a certain extent, the consistency of the prediction results of the radar branch and the precipitation branch.

[0156] However, although the radar images corresponding to the radar data and the precipitation images corresponding to the precipitation data obtained in the first training stage have high prediction accuracy, they both have blurring problems and poor visualization effects. Therefore, in the second training stage, the parameters of the shared encoder can be fixed, and the radar branch and precipitation branch can be fine-tuned separately. Strategies such as generative adversarial networks and perceptual loss can be used to fine-tune the radar branch and precipitation branch to improve the refinement of the predicted radar images and precipitation images.

[0157] Optionally, taking the radar branch as an example, this embodiment uses a discriminator to distinguish the true from the false between the predicted radar data sequence and the real radar data sequence, and also uses perception loss (including feature matching loss and style loss) to further adjust the parameters of the radar branch to improve the authenticity of the prediction results.

[0158] Figure 6A FIG. 1 is a schematic diagram of a radar extrapolation and precipitation prediction scenario according to an embodiment of the present invention. Figure 6A As shown, first weather detection data is input into a computing device, and the first computing device obtains the first weather detection data. The first weather detection data may be the actual result obtained by detecting the weather in a first time period. The computing device inputs the first weather detection data into a data prediction model, which may be a multi-task model for simultaneously predicting radar data and precipitation data in a future time period. In the data prediction model, target features of the first weather detection data are extracted and converted into second weather detection data and first target precipitation data. The second weather detection data includes the weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes the precipitation forecast result of the precipitation data in the second time period. After the second weather detection data and the first target precipitation data are predicted by the data prediction model, the second weather detection data and the first target precipitation data are output to a display interface of the computing device, and the second weather detection data and the first target precipitation data are displayed on the display interface.

[0159] Figure 6B FIG. 1 is a schematic diagram of another radar extrapolation and precipitation prediction scenario according to an embodiment of the present invention. Figure 6BAs shown, in response to a first input instruction acting on the operation interface, first radar detection data is input into the operation interface, and the first radar detection data is an actual result obtained by detecting the radar data in a first time period; in response to a prediction operation instruction acting on the operation interface, second radar detection data and first target precipitation data are displayed on the operation interface, wherein the second radar detection data and the first target precipitation data are obtained by extracting target features of the first radar detection data based on a data prediction model and converting the target features, the second radar detection data includes a prediction result of the radar data in a second time period after the first time period, and the first target precipitation data includes a precipitation prediction result of the precipitation data in the second time period.

[0160] The above method of this embodiment can be applied to short-term weather scenarios. By constructing a multi-task model and using the same data source for prediction, radar data and precipitation data can be predicted simultaneously. The prediction timeliness and required spatiotemporal resolution of the two are highly consistent, achieving the purpose of simultaneous radar extrapolation and precipitation prediction. Among them, the multi-task model is designed with two training stages. The first training stage is multi-task simultaneous training, which enables the training of the radar branch and precipitation branch of the multi-task model to achieve mutual assistance, improve the prediction effect, and ensure the consistency of the prediction results of the two. The second training stage is to adjust the parameters of the radar branch and / or precipitation branch to improve the refinement and visualization effect of the prediction results, avoiding the waste of computing resources and time caused by separately predicting radar data and precipitation data in the future time period, thereby solving the technical problem of low efficiency in predicting weather detection data and precipitation data, and achieving the technical effect of improving the efficiency of predicting weather detection data and precipitation data.

[0161] It should be noted that for the aforementioned method embodiments, for simplicity of description, they are all expressed as a series of action combinations. However, those skilled in the art should be aware that the present invention is not limited by the order of the actions described, because according to the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in this specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0162] Through the description of the above embodiments, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus the necessary general hardware platform, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present invention.

[0163] Example 3

[0164] According to an embodiment of the present invention, there is also provided a method for implementing the above Figure 2 It should be noted that the data processing device of this embodiment can be used to execute the data processing method of the present invention. Figure 2 The data processing method shown.

[0165] Figure 7 Schematic diagram of a data processing device according to an embodiment of the present invention. Figure 7 As shown, the data processing device 70 may include: a first acquiring unit 71 , an extracting unit 72 and a converting unit 73 .

[0166] The first acquiring unit 71 is configured to acquire first weather detection data, wherein the first weather detection data is an actual result obtained by detecting the weather in a first time period.

[0167] The extraction unit 72 is configured to extract target features of the first weather detection data based on the data prediction model.

[0168] The conversion unit 73 is used to convert the target feature into second weather detection data and first target precipitation data, wherein the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period.

[0169] It should be noted that the first acquisition unit 71, extraction unit 72, and conversion unit 73 described above correspond to steps S202 to S206 in Example 1. The examples and application scenarios implemented by the three units and the corresponding steps are the same, but are not limited to the contents disclosed in Example 1. It should be noted that the above units, as part of the device, can be run in the computer terminal 10 provided in Example 1.

[0170] According to an embodiment of the present invention, there is also provided a method for implementing the above Figure 3It should be noted that the data processing device of this embodiment can be used to execute the data processing method of the present invention. Figure 3 The data processing method shown.

[0171] Figure 8 FIG. 1 is a schematic diagram of another data processing device according to an embodiment of the present invention. Figure 8 As shown, the data processing device 80 may include: an input unit 81 and a first display unit 82 .

[0172] The input unit 81 is used to respond to a first input instruction applied to the operation interface and input first weather detection data, wherein the first weather detection data is an actual result obtained by detecting the weather in a first time period.

[0173] The first display unit 82 is used to respond to the prediction operation instruction on the operation interface and display the second weather detection data and the first target precipitation data on the operation interface, wherein the second weather detection data and the first target precipitation data are obtained by extracting the target features of the first weather detection data based on the data prediction model and converting the target features. The second weather detection data includes the weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes the precipitation forecast result of the precipitation data in the second time period.

[0174] It should be noted that the input unit 81 and the first display unit 82 correspond to steps S302 to S304 in Example 1. The examples and application scenarios implemented by the two units and the corresponding steps are the same, but are not limited to the contents disclosed in Example 1. It should be noted that the above-mentioned units, as part of the device, can be run in the computer terminal 10 provided in Example 1.

[0175] According to an embodiment of the present invention, there is also provided a method for implementing the above Figure 4 It should be noted that the data processing device of this embodiment can be used to execute the data processing method of the present invention. Figure 4 The data processing method shown.

[0176] Figure 9 FIG. 1 is a schematic diagram of another data processing device according to an embodiment of the present invention. Figure 9 As shown, the data processing device 90 may include: a second acquiring unit 91 , a second display unit 92 and a returning unit 93 .

[0177] The second acquisition unit 91 is used to acquire first weather detection data from the weather broadcast platform and display the first weather detection data on the operation interface, wherein the first weather detection data is an actual result obtained by detecting the weather in the first time period.

[0178] The second display unit 92 is used to respond to the prediction operation instruction on the operation interface and display the second weather detection data and the first target precipitation data on the operation interface, wherein the second weather detection data and the first target precipitation data are obtained by extracting the target features of the first weather detection data based on the data prediction model and converting the target features. The second weather detection data includes the weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes the precipitation forecast result of the precipitation data in the second time period.

[0179] The returning unit 93 is used to return the second weather detection data and the first target precipitation data to the weather broadcast platform.

[0180] It should be noted that the second acquisition unit 91 and the second display unit 92 correspond to steps S402 to S406 in Example 1. The examples and application scenarios implemented by the two units and the corresponding steps are the same, but are not limited to the contents disclosed in Example 1. It should be noted that the above-mentioned units, as part of the device, can be run in the computer terminal 10 provided in Example 1.

[0181] In the data processing device of this embodiment, based on the same data prediction model, radar data in a historical time period is used to simultaneously predict radar data and precipitation data in a future time period, which can save computing resources and reduce prediction time. It avoids the waste of computing resources and time caused by separate predictions of radar data and precipitation data in the future time period, thereby solving the technical problem of low efficiency in predicting weather detection data and precipitation data, and achieving the technical effect of improving the efficiency of predicting weather detection data and precipitation data.

[0182] Example 4

[0183] An embodiment of the present invention may provide a data processing system, which may include: an embodiment of the present invention may provide a computer terminal, which may be any computer terminal device in a computer terminal group. Optionally, in this embodiment, the computer terminal may be replaced by a terminal device such as a mobile terminal.

[0184] Optionally, in this embodiment, the computer terminal may be located in at least one network device among a plurality of network devices of a computer network.

[0185] In this embodiment, the above-mentioned computer terminal can execute the program code of the following steps in the data processing method of the application: obtaining first weather detection data, wherein the first weather detection data is the actual result obtained by detecting the weather in a first time period; extracting the target features of the first weather detection data based on the data prediction model; converting the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes the weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes the precipitation forecast result of the precipitation data in the second time period.

[0186] Optionally, Figure 10 1 is a block diagram of a computer terminal according to an embodiment of the present invention. Figure 10 As shown, the computer terminal A may include: a processor 1002 , a memory 1004 , and a transmission device 1006 .

[0187] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the data processing method and device in the embodiments of the present invention. The processor executes the software programs and modules stored in the memory to perform various functional applications and data processing, that is, to implement the above-mentioned data processing method. The memory may include a high-speed random access memory and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory may further include a memory remotely located relative to the processor, and these remote memories may be connected to the computer terminal A via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0188] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: obtain first weather detection data, wherein the first weather detection data is the actual result obtained by detecting the weather in a first time period; extract the target features of the first weather detection data based on the data prediction model; convert the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes the weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes the precipitation forecast result of the precipitation data in the second time period.

[0189] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: extract the target features from the first weather detection data based on the feature extraction model; predict and process the target features based on the first prediction model to obtain the second weather detection data; predict and process the target features based on the second prediction model in the data prediction model to obtain the first target precipitation data, wherein the second weather detection data and the first target precipitation data are used to represent the same weather conditions.

[0190] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: training the first sub-model based on the following data to obtain a first prediction model: first training data corresponding to the first prediction model, or the first training data and second training data corresponding to the second prediction model; and / or training the second sub-model based on the following data to obtain a second prediction model: second training data corresponding to the second prediction model, or the second training data and the first training data corresponding to the first prediction model.

[0191] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: obtain actual detection data, wherein the actual detection data is the actual result obtained by detecting the weather in the second time period and / or the actual result of precipitation data in the second time period; adjust the parameters of the first prediction model and / or the second prediction model based on the actual detection data and the corresponding loss function.

[0192] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: the first target precipitation data includes multiple precipitation prediction results of precipitation data within the second time period, and each precipitation prediction result is displayed in a graded manner on the operation interface according to the precipitation level of each precipitation prediction result.

[0193] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: determine the indication information of each precipitation forecast result, wherein the indication information is used to indicate the precipitation level of each precipitation forecast result; and display each precipitation forecast result in a graded manner according to the indication information.

[0194] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: the first target precipitation data includes multiple precipitation prediction results of precipitation data within the second time period, and each precipitation prediction result is displayed on the operation interface according to the area corresponding to each precipitation prediction result.

[0195] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: displaying the precipitation probability corresponding to the precipitation forecast result on the operation interface; displaying prompt information on the operation interface based on the precipitation probability, wherein the prompt information is used to prompt the travel strategy corresponding to the precipitation probability.

[0196] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: obtain adjustment information, wherein the adjustment information includes information used to characterize the accuracy of the precipitation prediction results and / or the environmental parameters of the area corresponding to the precipitation prediction results; adjust the data prediction model based on the adjustment information.

[0197] As an optional example, the processor can call the information and application stored in the memory through the transmission device to perform the following steps: responding to the first input instruction on the operation interface, inputting the first weather detection data, wherein the first weather detection data is the actual result obtained by detecting the weather in the first time period; responding to the prediction operation instruction on the operation interface, displaying the second weather detection data and the first target precipitation data on the operation interface, wherein the second weather detection data and the first target precipitation data are obtained by extracting the target features of the first weather detection data based on the data prediction model and converting the target features, the second weather detection data includes the weather forecast result obtained by detecting the weather in the second time period after the first time period, and the first target precipitation data includes the precipitation forecast result of the precipitation data in the second time period.

[0198] As another optional example, the processor can call the information and application stored in the memory through the transmission device to perform the following steps: obtain the first weather detection data from the weather broadcast platform, and display the first weather detection data on the operation interface, wherein the first weather detection data is the actual result obtained by detecting the weather in the first time period; in response to the prediction operation instruction acting on the operation interface, display the second weather detection data and the first target precipitation data on the operation interface, wherein the second weather detection data and the first target precipitation data are obtained by extracting the target features of the first weather detection data based on the data prediction model and converting the target features, the second weather detection data includes the weather forecast result obtained by detecting the weather in the second time period after the first time period, and the first target precipitation data includes the precipitation forecast result of the precipitation data in the second time period; return the second weather detection data and the first target precipitation data to the weather broadcast platform.

[0199] As another optional example, the processor can call the information and application stored in the memory through the transmission device to perform the following steps: determine the first target area where the target vegetation is located; obtain first weather detection data of the first target area, wherein the first weather detection data is the actual result obtained by detecting the weather in the first target area within a first time period; extract the target features of the first weather detection data based on the data prediction model; convert the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes the weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes the precipitation forecast result of the precipitation data in the second time period; determine whether to fertilize the target vegetation based on the second weather detection data and the first target precipitation data.

[0200] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: when determining to fertilize the target vegetation based on the second weather detection data and the first target precipitation data, output a prompt information, wherein the prompt information is used to prompt to fertilize the target vegetation; and / or output a fertilizer control instruction to the fertilizer equipment, wherein the fertilizer control instruction is used to control the fertilizer equipment to perform fertilization operations on the target vegetation.

[0201] As another optional example, the processor can call the information and application stored in the memory through the transmission device to perform the following steps: obtain first weather detection data of the second target area, wherein the first weather detection data is the actual result obtained by detecting the weather in the second target area within the first time period; extract the target features of the first weather detection data based on the data prediction model; convert the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes the weather forecast result obtained by detecting the weather in the second time period after the first time period, and the first target precipitation data includes the precipitation forecast result of the precipitation data in the second time period; determine whether to push drainage strategy data based on the second weather detection data and the first target precipitation data, wherein the drainage strategy data is used to represent the strategy for draining the second target area.

[0202] As another optional example, the processor can call the information and application stored in the memory through the transmission device to perform the following steps: obtain first weather detection data of the third target area, wherein the first weather detection data is the actual result obtained by detecting the weather in the third target area within the first time period; extract the target features of the first weather detection data based on the data prediction model; convert the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes the weather forecast result obtained by detecting the weather in the second time period after the first time period, and the first target precipitation data includes the precipitation forecast result of the precipitation data in the second time period; determine the water level of the water area in the third target area based on the second weather detection data and the first target precipitation data; determine whether to push flood control strategy data based on the water level, wherein the flood control strategy data is used to represent the strategy for flood control in the third target area.

[0203] An embodiment of the present invention provides a data processing solution for obtaining first weather detection data, wherein the first weather detection data is the actual result obtained by detecting the weather in a first time period; extracting target features of the first weather detection data based on a data prediction model; and converting the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes the weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes the precipitation forecast result of the precipitation data in the second time period. Based on the same data prediction model, radar data in a historical time period is used to simultaneously predict radar data and precipitation data in a future time period, thereby saving computing resources and reducing prediction time, avoiding the waste of computing resources and time caused by separately predicting radar data and precipitation data in the future time period, thereby solving the technical problem of low efficiency in predicting weather detection data and precipitation data, and achieving the technical effect of improving the efficiency of predicting weather detection data and precipitation data.

[0204] It can be understood by those skilled in the art that Figure 10 The structure shown is for illustration only. The computer terminal A may also be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a handheld computer, a mobile Internet device (MID for short), a PAD, or other terminal devices. Figure 10 It does not limit the structure of the above-mentioned computer terminal A. For example, the computer terminal A may also include Figure 10 More or fewer components (such as network interfaces, display devices, etc.) shown in, or with Figure 10 Different configurations shown.

[0205] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0206] Example 5

[0207] The embodiment of the present invention further provides a computer-readable storage medium. Optionally, in this embodiment, the computer-readable storage medium can be used to store the program code executed by the data processing method provided in the first embodiment.

[0208] Optionally, in this embodiment, the computer-readable storage medium may be located in any computer terminal in a computer terminal group in a computer network, or in any mobile terminal in a mobile terminal group.

[0209] Optionally, in this embodiment, the computer-readable storage medium is configured to store program code for performing the following steps: obtaining first weather detection data, wherein the first weather detection data is an actual result obtained by detecting the weather in a first time period; extracting target features of the first weather detection data based on a data prediction model; converting the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period.

[0210] Optionally, the computer-readable storage medium is also configured to store program code for executing the following steps: extracting target features from the first weather detection data based on a feature extraction model in the data prediction model; performing predictive processing on the target features based on the first prediction model in the data prediction model to obtain second weather detection data; performing predictive processing on the target features based on the second prediction model in the data prediction model to obtain first target precipitation data, wherein the second weather detection data and the first target precipitation data are used to represent the same weather conditions.

[0211] Optionally, the computer-readable storage medium is also configured to store program code for performing the following steps: training the first sub-model based on the following data to obtain a first prediction model: first training data corresponding to the first prediction model, or the first training data and second training data corresponding to the second prediction model; and / or training the second sub-model based on the following data to obtain a second prediction model: second training data corresponding to the second prediction model, or the second training data and the first training data corresponding to the first prediction model.

[0212] Optionally, the computer-readable storage medium is also configured to store program code for performing the following steps: obtaining actual detection data, wherein the actual detection data is the actual result obtained by detecting the weather in the second time period and / or the actual result of precipitation data in the second time period; adjusting the parameters of the first prediction model and / or the second prediction model based on the actual detection data and the corresponding loss function.

[0213] Optionally, the computer-readable storage medium is also configured to store program code for executing the following steps: the first target precipitation data includes multiple precipitation prediction results of precipitation data within the second time period, and each precipitation prediction result is displayed in a graded manner on the operation interface according to the precipitation level of each precipitation prediction result.

[0214] Optionally, the computer-readable storage medium is further configured to store program code for executing the following steps: determining indication information for each precipitation forecast result, wherein the indication information is used to indicate the precipitation level of each precipitation forecast result; and displaying each precipitation forecast result in a graded manner according to the indication information.

[0215] Optionally, the computer-readable storage medium is also configured to store program code for executing the following steps: the first target precipitation data includes multiple precipitation prediction results of precipitation data within the second time period, and each precipitation prediction result is displayed on the operation interface according to the area corresponding to each precipitation prediction result.

[0216] Optionally, the computer-readable storage medium is also configured to store program code for executing the following steps: displaying the precipitation probability corresponding to the precipitation forecast result on the operation interface; and displaying prompt information on the operation interface based on the precipitation probability, wherein the prompt information is used to prompt a travel strategy corresponding to the precipitation probability.

[0217] Optionally, the computer-readable storage medium is also configured to store program code for performing the following steps: obtaining adjustment information, wherein the adjustment information includes information used to characterize the accuracy of the precipitation prediction results and / or environmental parameters of the area corresponding to the precipitation prediction results; and adjusting the data prediction model based on the adjustment information.

[0218] As an optional example, a computer-readable storage medium is configured to store program code for performing the following steps: in response to a first input instruction acting on an operation interface, inputting first weather detection data, wherein the first weather detection data is an actual result obtained by detecting the weather within a first time period; in response to a prediction operation instruction acting on the operation interface, displaying second weather detection data and first target precipitation data on the operation interface, wherein the second weather detection data and the first target precipitation data are obtained by extracting target features of the first weather detection data based on a data prediction model and converting the target features, the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period.

[0219] As another optional example, a computer-readable storage medium is configured to store program code for performing the following steps: obtaining first weather detection data from a weather broadcast platform and displaying the first weather detection data on an operation interface, wherein the first weather detection data is an actual result obtained by detecting the weather in a first time period; in response to a prediction operation instruction acting on the operation interface, displaying second weather detection data and first target precipitation data on the operation interface, wherein the second weather detection data and the first target precipitation data are obtained by extracting target features of the first weather detection data based on a data prediction model and converting the target features, the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period; and returning the second weather detection data and the first target precipitation data to the weather broadcast platform.

[0220] As another optional example, a computer-readable storage medium is configured to store program code for performing the following steps: determining a first target area where the target vegetation is located; obtaining first weather detection data for the first target area, wherein the first weather detection data is the actual result obtained by detecting the weather in the first target area within a first time period; extracting target features of the first weather detection data based on a data prediction model; converting the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period; and determining whether to fertilize the target vegetation based on the second weather detection data and the first target precipitation data.

[0221] The computer-readable storage medium is also configured to store program code for executing the following steps: when determining to fertilize the target vegetation based on the second weather detection data and the first target precipitation data, outputting a prompt message, wherein the prompt message is used to prompt to fertilize the target vegetation; and / or outputting a fertilization control instruction to the fertilization equipment, wherein the fertilization control instruction is used to control the fertilization equipment to perform fertilization operations on the target vegetation.

[0222] As another optional example, a computer-readable storage medium is configured to store program code for performing the following steps: obtaining first weather detection data for a second target area, wherein the first weather detection data is the actual result obtained by detecting the weather in the second target area within a first time period; extracting target features of the first weather detection data based on a data prediction model; converting the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period; determining whether to push drainage strategy data based on the second weather detection data and the first target precipitation data, wherein the drainage strategy data is used to represent a strategy for draining the second target area.

[0223] As another optional example, a computer-readable storage medium is configured to store program code for performing the following steps: obtaining first weather detection data for a third target area, wherein the first weather detection data is the actual result obtained by detecting the weather in the third target area within a first time period; extracting target features of the first weather detection data based on a data prediction model; converting the target features into second weather detection data and first target precipitation data, wherein the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period; determining the water level of the water area in the third target area based on the second weather detection data and the first target precipitation data; and determining whether to push flood control strategy data based on the water level, wherein the flood control strategy data is used to represent a strategy for flood control in the third target area.

[0224] Optionally, the computer-readable storage medium is also configured to store program code for performing the following steps: in response to a second input instruction acting on the operating interface, inputting third radar data, wherein the third radar data is the actual result obtained by detecting the weather in the second time period by radar; in response to a first adjustment instruction acting on the operating interface, adjusting the parameters of the first prediction model based on the third radar data, the second weather detection data and the first loss function, wherein the first prediction model is used to perform prediction processing on the target features to obtain the second weather detection data.

[0225] Optionally, the computer-readable storage medium is also configured to store program code for executing the following steps: displaying a first clarity of the radar image corresponding to the second weather detection data on the operation interface; and responding to a first adjustment instruction acting on the operation interface when the first clarity is less than a first threshold.

[0226] Optionally, the computer-readable storage medium is further configured to store program code for executing the following steps: in response to a third input instruction acting on the operation interface, inputting second target precipitation data, wherein the second target precipitation data is the actual result of precipitation data within a second time period; in response to a second adjustment instruction acting on the operation interface, adjusting the parameters of the second prediction model based on the second target precipitation data, the first target precipitation data and the second loss function, wherein the second prediction model is used to perform prediction processing on the target features to obtain the first target precipitation data.

[0227] Optionally, the computer-readable storage medium is further configured to store program code for executing the following steps: displaying a second clarity of a precipitation image corresponding to the first target precipitation data on an operation interface; and responding to a second adjustment instruction acting on the operation interface when the second clarity is less than a second threshold.

[0228] As an optional example, a computer-readable storage medium is configured to store program code for executing the following steps: obtaining first weather detection data from a weather broadcast platform and displaying the first weather detection data on an operation interface, wherein the first weather detection data is an actual result obtained by detecting the weather in a first time period; in response to a prediction operation instruction acting on the operation interface, displaying second weather detection data and first target precipitation data on the operation interface, wherein the second weather detection data and the first target precipitation data are obtained by extracting target features of the first weather detection data based on a data prediction model and converting the target features, the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period, and the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period; and returning the second weather detection data and the first target precipitation data to the weather broadcast platform.

[0229] The serial numbers of the above embodiments of the present invention are for description only and do not represent the advantages or disadvantages of the embodiments.

[0230] In the above embodiments of the present invention, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0231] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0232] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0233] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0234] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, etc. Various media that can store program codes.

[0235] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A data processing method, characterized in that: include: Acquire first weather detection data, wherein the first weather detection data is an actual result obtained by detecting the weather in a first time period; extracting target features from the first weather detection data based on a feature extraction model included in the data prediction model; The target feature is predictively processed based on the first prediction model included in the data prediction model to obtain second weather detection data, wherein the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period; the target feature is predictively processed based on the second prediction model included in the data prediction model to obtain first target precipitation data, wherein the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period, and the second weather detection data and the first target precipitation data are used to represent the same weather conditions.

2. The method according to claim 1, characterized in that The method further comprises: The first prediction model is obtained by training the first sub-model based on the following data: first training data corresponding to the first prediction model, or the first training data and second training data corresponding to the second prediction model; and / or The second prediction model is obtained by training the second sub-model based on the following data: second training data corresponding to the second prediction model, or the second training data and first training data corresponding to the first prediction model.

3. The method according to claim 1, characterized in that The method further comprises: Acquiring actual detection data, wherein the actual detection data is an actual result obtained by detecting the weather in the second time period and / or an actual result of precipitation data in the second time period; The parameters of the first prediction model and / or the second prediction model are adjusted based on the actual detection data and the corresponding loss function.

4. The method according to claim 1, wherein The first target precipitation data includes a plurality of precipitation prediction results of precipitation data within the second time period, and the method further includes: Displaying each precipitation forecast result in a graded manner on an operation interface according to the precipitation level of each precipitation forecast result; and / or Each precipitation prediction result is displayed on the operation interface according to the area corresponding to each precipitation prediction result.

5. The method according to claim 4, characterized in that The method further includes: determining indication information of each precipitation prediction result, wherein the indication information is used to indicate the precipitation level of each precipitation prediction result; Displaying each precipitation forecast result in a hierarchical manner includes: displaying each precipitation forecast result in a hierarchical manner according to the indication information.

6. The method according to claim 1, characterized in that The method further comprises: Displaying the precipitation probability corresponding to the precipitation prediction result on an operation interface; displaying prompt information on the operation interface based on the precipitation probability, wherein the prompt information is used to prompt a travel strategy corresponding to the precipitation probability; and / or Acquire adjustment information, wherein the adjustment information includes information used to characterize the accuracy of the precipitation prediction result and / or environmental parameters of the area corresponding to the precipitation prediction result; and adjust the data prediction model based on the adjustment information.

7. A data processing method, characterized in that: include: Obtaining first weather detection data from a weather broadcast platform and displaying the first weather detection data on an operation interface, wherein the first weather detection data is an actual result obtained by detecting the weather in a first time period; In response to a prediction operation instruction on the operation interface, second weather detection data and first target precipitation data are displayed on the operation interface, wherein the second weather detection data is obtained by extracting target features of the first weather detection data based on a feature extraction model included in a data prediction model, and predicting the target features using feature extraction models included in different branches of the data prediction model; the first target precipitation data is obtained by predicting the target features using a second prediction model included in the data prediction model; the second weather detection data includes a weather forecast result obtained by detecting weather in a second time period after the first time period; the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period; and the second weather detection data and the first target precipitation data are used to represent the same weather conditions; The second weather detection data and the first target precipitation data are returned to the weather broadcast platform.

8. A data processing method, characterized in that: include: determining a first target area where target vegetation is located; Acquire first weather detection data of the first target area, wherein the first weather detection data is an actual result obtained by detecting the weather of the first target area within a first time period; extracting target features from the first weather detection data based on a feature extraction model included in the data prediction model; performing prediction processing on the target feature based on a first prediction model included in the data prediction model to obtain second weather detection data, wherein the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period; performing prediction processing on the target feature based on the second prediction model included in the data prediction model to obtain first target precipitation data, wherein the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period, and the second weather detection data and the first target precipitation data are used to represent the same weather conditions; Determining whether to fertilize the target vegetation is based on the second weather detection data and the first target precipitation data.

9. The method according to claim 8, characterized in that When determining to fertilize the target vegetation based on the second weather detection data and the first target precipitation data, the method further includes: Outputting prompt information, wherein the prompt information is used to prompt the target vegetation to be fertilized; and / or Outputting a fertilization control instruction to the fertilization equipment, wherein the fertilization control instruction is used to control the fertilization equipment to perform a fertilization operation on the target vegetation.

10. A data processing method, characterized in that: include: Acquire first weather detection data of a second target area, wherein the first weather detection data is an actual result obtained by detecting the weather of the second target area within a first time period; extracting target features from the first weather detection data based on a feature extraction model included in the data prediction model; performing prediction processing on the target feature based on a first prediction model included in the data prediction model to obtain second weather detection data, wherein the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period; performing prediction processing on the target feature based on the second prediction model included in the data prediction model to obtain first target precipitation data, wherein the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period, and the second weather detection data and the first target precipitation data are used to represent the same weather conditions; Whether to push drainage strategy data is determined based on the second weather detection data and the first target precipitation data, wherein the drainage strategy data is used to represent a strategy for draining the second target area.

11. A data processing method, characterized in that: include: Acquire first weather detection data of a third target area, wherein the first weather detection data is an actual result obtained by detecting the weather of the third target area within a first time period; extracting target features from the first weather detection data based on a feature extraction model included in the data prediction model; performing prediction processing on the target feature based on a first prediction model included in the data prediction model to obtain second weather detection data, wherein the second weather detection data includes a weather forecast result obtained by detecting the weather in a second time period after the first time period; performing prediction processing on the target feature based on the second prediction model included in the data prediction model to obtain first target precipitation data, wherein the first target precipitation data includes a precipitation forecast result of the precipitation data in the second time period, and the second weather detection data and the first target precipitation data are used to represent the same weather conditions; determining a water level of the water area in the third target area based on the second weather detection data and the first target precipitation data; Determine whether to push flood control strategy data based on the water level, wherein the flood control strategy data is used to represent a strategy for flood control of the third target area.

12. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored program, wherein when the program is executed by a processor, the device where the computer-readable storage medium is located is controlled to execute the method according to any one of claims 1 to 11.

Images