Real-time rainfall prediction method and device, electronic equipment and storage medium

Abstract

The invention discloses a real-time rainfall prediction method and apparatus, an electronic device and a storage medium. The method comprises the steps of obtaining influence factors of real-time rainfall; wherein the influence factors comprise temperature, humidity, air pressure, wind speed, dew point, cloud amount, cloud layer type, cloud top temperature, water vapor content, cloud coverage rate and terrain type; and inputting the influence factors of the real-time rainfall into a rainfall prediction model, so that the rainfall prediction model generates predicted real-time rainfall according to the influence factors of the real-time rainfall. According to the invention, the accuracy and response speed of rainfall prediction can be improved, and the capability of dealing with sudden weather events is enhanced.

Description

Technical Field

[0001] The present invention relates to the field of deep learning technology, and in particular to a real-time rainfall prediction method, device, electronic device and storage medium. Background Art

[0002] Changes in rainfall directly affect the management of water resources, the construction and maintenance of urban infrastructure, and the protection of the ecological environment. As climate change intensifies, changes in rainfall patterns are becoming increasingly complex, and the frequency and intensity of extreme rainfall events are increasing. Improving the accuracy and real-time nature of rainfall forecasts has become the key to addressing climate change and achieving sustainable development.

[0003] The lack of real-time performance of existing technologies in obtaining rainfall data has brought great challenges to the accurate reporting of precipitation and timely response to extreme weather events. Although meteorological observation stations can collect rainfall influencing factors such as temperature, air pressure, humidity, and wind speed in real time, there is still a significant delay in the acquisition of rainfall data. Traditional meteorological stations usually provide hourly or daily precipitation statistics, which are often the result of accumulation over a certain period of time and are updated at a low frequency. Especially in the face of local short-term heavy rainfall or sudden precipitation events, data delays are particularly obvious, making it difficult to reflect precipitation conditions in a timely manner, affecting the rapid response to sudden weather. Taking China's basic ground meteorological observation data as an example, precipitation data is usually the cumulative value of a time interval of three hours or longer. This delay is particularly inapplicable in emergency response and disaster warning. Summary of the invention

[0004] The embodiments of the present invention provide a real-time rainfall prediction method, device, electronic device and storage medium. By implementing the present invention, more accurate and real-time rainfall prediction can be achieved, the accuracy and response speed of rainfall prediction can be improved, and the ability to respond to sudden weather events can be enhanced.

[0005] An embodiment of the present invention provides a real-time rainfall prediction method, comprising:

[0006] Obtaining influencing factors of real-time rainfall; wherein the influencing factors include: temperature, humidity, air pressure, wind speed, dew point, cloud amount, cloud type, cloud top temperature, water vapor content, cloud coverage and terrain type;

[0007] Inputting the influencing factors of the real-time rainfall into a rainfall prediction model, so that the rainfall prediction model generates a predicted real-time rainfall according to the influencing factors of the real-time rainfall;

[0008] Wherein, the training of the rainfall prediction model includes:

[0009] Obtaining the influencing factors of rainfall at a historical moment and the corresponding historical rainfall label; the historical rainfall label is the rainfall at the historical moment;

[0010] Construct a training set based on the factors affecting rainfall at historical moments and the corresponding historical rainfall labels;

[0011] Randomly dividing the training set into several batches of training samples according to a preset batch size;

[0012] Input each batch of training samples into the rainfall prediction model in turn, and train the rainfall prediction model until a preset number of training times is reached; wherein, the rainfall prediction model outputs the rainfall corresponding to the training sample each time a batch of training samples is received; calculates the loss function value through the loss function according to the rainfall and the corresponding historical rainfall label; and uses the optimizer to update the rainfall prediction model according to the loss function value.

[0013] Furthermore, the temperature, humidity, air pressure, wind speed and dew point are observed by meteorological stations; the cloud type, cloud top temperature, water vapor content and cloud coverage are collected by meteorological satellites; and the terrain type is obtained by a geographic information system.

[0014] Furthermore, before obtaining the influencing factors of rainfall at historical moments, it also includes:

[0015] Get the original factors affecting rainfall at historical moments;

[0016] Clean the original influencing factors and align the timestamps to obtain time series data;

[0017] Convert time series data into numerical data and generate the factors affecting rainfall at historical moments.

[0018] Furthermore, after inputting the influencing factors of the real-time rainfall into the rainfall prediction model so that the rainfall prediction model generates the predicted real-time rainfall according to the influencing factors of the real-time rainfall, the method further includes:

[0019] Compare the predicted real-time rainfall with the set threshold and generate a corresponding warning signal;

[0020] According to the early warning signal, corresponding emergency response measures are generated.

[0021] Based on the above method embodiment, the present invention provides a corresponding device embodiment.

[0022] An embodiment of the present invention provides a real-time rainfall prediction device, comprising: an influencing factor acquisition module and a real-time rainfall prediction module;

[0023] The influencing factor acquisition module is used to acquire the influencing factors of real-time rainfall; wherein the influencing factors include: temperature, humidity, air pressure, wind speed, dew point, cloud amount, cloud type, cloud top temperature, water vapor content, cloud coverage and terrain type;

[0024] The real-time rainfall prediction module is used to input the influencing factors of the real-time rainfall into the rainfall prediction model, so that the rainfall prediction model generates the predicted real-time rainfall according to the influencing factors of the real-time rainfall;

[0025] Wherein, the training of the rainfall prediction model includes:

[0026] Obtaining the influencing factors of rainfall at a historical moment and the corresponding historical rainfall label; the historical rainfall label is the rainfall at the historical moment;

[0027] Construct a training set based on the factors affecting rainfall at historical moments and the corresponding historical rainfall labels;

[0028] Randomly dividing the training set into several batches of training samples according to a preset batch size;

[0029] Input each batch of training samples into the rainfall prediction model in turn, and train the rainfall prediction model until a preset number of training times is reached; wherein, the rainfall prediction model outputs the rainfall corresponding to the training sample each time a batch of training samples is received; calculates the loss function value through the loss function according to the rainfall and the corresponding historical rainfall label; and uses the optimizer to update the rainfall prediction model according to the loss function value.

[0030] Furthermore, the temperature, humidity, air pressure, wind speed and dew point are observed by meteorological stations; the cloud type, cloud top temperature, water vapor content and cloud coverage are collected by meteorological satellites; and the terrain type is obtained by a geographic information system.

[0031] Furthermore, before obtaining the influencing factors of rainfall at historical moments, it also includes:

[0032] Get the original factors affecting rainfall at historical moments;

[0033] Clean the original influencing factors and align the timestamps to obtain time series data;

[0034] Convert time series data into numerical data and generate the factors affecting rainfall at historical moments.

[0035] Furthermore, after inputting the influencing factors of the real-time rainfall into the rainfall prediction model so that the rainfall prediction model generates the predicted real-time rainfall according to the influencing factors of the real-time rainfall, the method further includes:

[0036] Compare the predicted real-time rainfall with the set threshold and generate a corresponding warning signal;

[0037] According to the early warning signal, corresponding emergency response measures are generated.

[0038] Based on the above method embodiment, the present invention provides a corresponding electronic device embodiment.

[0039] An embodiment of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the real-time rainfall prediction method described in any one of the above method embodiments can be implemented.

[0040] Based on the above method item embodiments, the present invention provides a corresponding storage medium item embodiment.

[0041] An embodiment of the present invention provides a storage medium having a computer program stored thereon. When the computer program is executed by a processor, the real-time rainfall prediction method described in any one of the above method embodiments can be implemented.

[0042] Compared with the prior art, the present invention has the following beneficial effects:

[0043] The embodiments of the present invention provide a real-time rainfall prediction method, device, electronic device and storage medium. The method trains a preset neural network model through rainfall influencing factors and rainfall at historical moments to obtain a rainfall prediction model. By using the neural network model, the present invention can adaptively learn the complex nonlinear relationship between rainfall and rainfall influencing factors, thereby improving the accuracy of the prediction. Since rainfall influencing factors can be obtained in real time, and rainfall data mainly relies on hourly or daily cumulative data acquisition, it has a certain lag. Therefore, the input of real-time rainfall influencing factors can make the prediction result of real-time rainfall more accurate, especially when dealing with short-term heavy rainfall or sudden precipitation events, it can provide timely and effective precipitation warnings. By implementing the present invention, the accuracy and real-time performance of real-time rainfall prediction can be significantly improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a flow chart of a real-time rainfall prediction method provided by one embodiment of the present invention.

[0045] Figure 2 It is a schematic diagram of the flow of training a rainfall prediction model provided by an embodiment of the present invention.

[0046] Figure 3 It is a structural schematic diagram of a real-time rainfall prediction device provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments 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 creative work are within the scope of protection of the present invention.

[0048] like Figure 1 As shown, an embodiment of the present invention provides a real-time rainfall prediction method, which at least includes the following steps:

[0049] Step S1, obtaining influencing factors of real-time rainfall; wherein the influencing factors include: temperature, humidity, air pressure, wind speed, dew point, cloud amount, cloud type, cloud top temperature, water vapor content, cloud coverage and terrain type;

[0050] Specifically, the temperature, humidity, air pressure, wind speed and dew point are observed by meteorological stations; the cloud type, cloud top temperature, water vapor content and cloud coverage are collected by meteorological satellites; and the terrain type is obtained by a geographic information system.

[0051] It is understandable that the observation data of meteorological stations can reflect the dynamic changes of ground meteorological conditions in real time, and provide stable and highly timely basic information for the prediction model; the data such as cloud type, cloud top temperature, water vapor content and cloud coverage collected by meteorological satellites can capture the structure and movement characteristics of clouds in the atmosphere from a macro perspective, effectively making up for the limitations of the ground observation range; the terrain type provided by the geographic information system can fully consider the significant impact of terrain on rainfall distribution and intensity, and further enhance the model's ability to depict local rainfall changes. Through the comprehensive acquisition and integration of these influencing factors, the present invention can overcome the limitations of a single data source in the prior art and ensure the comprehensiveness and timeliness of the input data. At the same time, this comprehensive method based on multi-source data can more accurately capture the complex nonlinear relationship between rainfall and a variety of meteorological and geographical factors, thereby providing reliable data support for the prediction model, significantly improving the accuracy and adaptability of rainfall prediction, and providing important scientific basis and technical support for the monitoring and early warning of extreme weather events.

[0052] Step S2, inputting the influencing factors of the real-time rainfall into the rainfall prediction model, so that the rainfall prediction model generates a predicted real-time rainfall according to the influencing factors of the real-time rainfall;

[0053] like Figure 2 As shown, in a preferred embodiment of the present invention, the training of the rainfall prediction model includes:

[0054] Step S2.1, obtaining the influencing factors of rainfall at historical moments and the corresponding historical rainfall labels; the historical rainfall labels are the rainfall at historical moments;

[0055] Step S2.2, constructing a training set according to the influencing factors of rainfall at historical moments and the corresponding historical rainfall labels;

[0056] Step S2.3, randomly dividing the training set into several batches of training samples according to a preset batch size;

[0057] Step S2.4, input each batch of training samples into the rainfall prediction model in sequence, and train the rainfall prediction model until a preset number of training times is reached; wherein, the rainfall prediction model outputs the rainfall corresponding to the training sample each time a batch of training samples is received; calculates the loss function value through the loss function according to the rainfall and the corresponding historical rainfall label; and uses the optimizer to update the rainfall prediction model according to the loss function value.

[0058] Optionally, the hyperparameters of the rainfall prediction model, such as a preset number of training times and a preset batch size, are optimized by a genetic algorithm, and the method specifically includes:

[0059] Construct a data set based on the factors affecting rainfall at historical moments and the corresponding historical rainfall labels;

[0060] According to the influencing factors of rainfall at historical moments and the dataset constructed by historical rainfall labels, a group of populations are randomly generated, and each individual represents a set of hyperparameter combinations, including specific values ​​of the number of training times and batch size.

[0061] The dataset is divided into a training set and a validation set. For each individual hyperparameter combination, the model is trained according to the number of training times and batch size specified in the individual, and the loss value of the model (such as mean square error) is calculated using the validation set, and the loss value is used as the fitness of the individual. According to the fitness value, the individual with higher fitness is selected as the parent, and a strategy such as roulette selection or tournament selection is used. The parent hyperparameter combination is crossovered to generate new individuals. For example, the number of training times of a parent is exchanged with the batch size of another parent to generate the next generation. The number of training times or batch size of the new generation of individuals is randomly adjusted to increase the diversity of the search, such as changing the batch size from 32 to 64. The individuals with lower fitness are replaced with the newly generated individuals to form the next generation population. The above process is repeated until the maximum number of generations is reached or the fitness value no longer changes significantly within multiple generations. The individual with the highest fitness is selected, and the number of training times and batch size it contains are used as the final optimization result to build and train the rainfall prediction model. Through this process, the genetic algorithm will optimize the number of training times and batch size suitable for the data set based on the performance of the training set and the validation set, so that the predictive performance of the model reaches the best level.

[0062] Specifically, before obtaining the influencing factors of rainfall at historical moments, it also includes:

[0063] Get the original factors affecting rainfall at historical moments;

[0064] Clean the original influencing factors and align the timestamps to obtain time series data;

[0065] Convert time series data into numerical data and generate the factors affecting rainfall at historical moments.

[0066] For example, due to different data sources, there are inconsistencies in the data format, sampling frequency, and timestamp of the original influencing factors, so the original influencing factors need to be cleaned and the timestamps need to be aligned.

[0067] The original influencing factors were cleaned and timestamps were aligned in the following ways:

[0068] The missing values, outliers and duplicate data of the original influencing factors are removed; specifically, the missing records in the original influencing factors are completed by interpolation, mean filling or Lagrange interpolation; outliers that obviously deviate from the normal range are eliminated by box plot detection or 3σ principle, and replaced by the mean of adjacent moments; for repeated timestamp records, only the first one is retained or the average of multiple records is calculated.

[0069] During the timestamp alignment process, data from different sources are uniformly mapped to a standard timeline. For example, high-frequency data is downsampled to match low-frequency data, or low-frequency data is interpolated to be consistent with high-frequency data. In addition, all data is converted to a time zone to ensure that the timestamps of data from different sources can be aligned to a unified time zone and time format.

[0070] After completing data cleaning and timestamp alignment, the aligned time series data is digitized in the following ways:

[0071] Convert non-numeric data into numerical representation using one-hot encoding or label encoding; normalize or standardize the numerical data to eliminate the dimensional differences of different physical quantities, including maximum and minimum normalization or z-score standardization.

[0072] Among them, temperature, humidity, air pressure, wind speed, dew point, cloud amount, cloud top temperature, water vapor content and cloud coverage are all numerical data; cloud type and terrain type are all non-numerical data;

[0073] Specifically, after inputting the influencing factors of the real-time rainfall into the rainfall prediction model so that the rainfall prediction model generates the predicted real-time rainfall according to the influencing factors of the real-time rainfall, the method further includes:

[0074] Compare the predicted real-time rainfall with the set threshold and generate a corresponding warning signal;

[0075] According to the early warning signal, corresponding emergency response measures are generated.

[0076] It should be noted here that the predicted real-time rainfall is compared with the set threshold to determine the intensity level of rainfall and generate a corresponding warning signal. The level of the warning signal is divided into multiple levels according to the amount of rainfall and the preset threshold, such as blue, yellow, orange and red, corresponding to light, moderate, heavy and extreme rainfall conditions respectively. After each level of warning signal is generated, it will serve as the basis for triggering different emergency response measures. According to the warning signal, combined with the predicted rainfall duration, spatial distribution and historical disaster records, the corresponding emergency response measures are generated. For example, when a blue warning signal is generated, it may only be necessary to remind the public to pay attention to traffic and water accumulation in low-lying areas; when a yellow warning signal is generated, it is necessary to start the pre-drainage and pre-discharge work of regional water conservancy facilities; for orange or red warning signals, it is necessary to arrange personnel evacuation in advance, demarcate dangerous areas and start emergency rescue plans. All emergency response measures must be promptly notified to the public and relevant departments through various channels such as broadcasting, text messages, and mobile applications to ensure that the potential hazards caused by rainfall are minimized.

[0077] like Figure 3As shown, an embodiment of the present invention provides a real-time rainfall prediction device, including: an influencing factor acquisition module 101 and a real-time rainfall prediction module 102;

[0078] The influencing factor acquisition module 101 is used to acquire the influencing factors of real-time rainfall; wherein the influencing factors include: temperature, humidity, air pressure, wind speed, dew point, cloud amount, cloud type, cloud top temperature, water vapor content, cloud coverage and terrain type;

[0079] The real-time rainfall prediction module 102 is used to input the influencing factors of the real-time rainfall into the rainfall prediction model, so that the rainfall prediction model generates predicted real-time rainfall according to the influencing factors of the real-time rainfall; wherein, the training of the rainfall prediction model includes: obtaining the influencing factors of rainfall at historical moments and the corresponding historical rainfall labels; the historical rainfall labels are the rainfall at historical moments; constructing a training set according to the influencing factors of rainfall at historical moments and the corresponding historical rainfall labels; randomly dividing the training set into several batches of training samples according to a preset batch size; inputting the training samples of each batch into the rainfall prediction model in turn, and training the rainfall prediction model until a preset number of training times is reached; wherein, the rainfall prediction model outputs the rainfall corresponding to the training sample each time a batch of training samples is received; calculating the loss function value through the loss function according to the rainfall and the corresponding historical rainfall label; and updating the rainfall prediction model according to the loss function value using an optimizer.

[0080] Exemplarily, the temperature, humidity, air pressure, wind speed and dew point are observed by meteorological stations; the cloud type, cloud top temperature, water vapor content and cloud coverage are collected by meteorological satellites; and the terrain type is obtained by a geographic information system.

[0081] Optionally, before obtaining the influencing factors of rainfall at historical moments, it also includes:

[0082] Get the original factors affecting rainfall at historical moments;

[0083] Clean the original influencing factors and align the timestamps to obtain time series data;

[0084] Convert time series data into numerical data and generate the factors affecting rainfall at historical moments.

[0085] Optionally, after inputting the influencing factors of the real-time rainfall into the rainfall prediction model so that the rainfall prediction model generates the predicted real-time rainfall according to the influencing factors of the real-time rainfall, the method further includes:

[0086] Compare the predicted real-time rainfall with the set threshold and generate a corresponding warning signal;

[0087] According to the early warning signal, corresponding emergency response measures are generated.

[0088] It should be noted that the embodiments of the device described above correspond to the above embodiments of the present invention, and can implement any of the methods described above in the present invention. In addition, the embodiments of the above-mentioned device are merely schematic, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which can be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art can understand and implement it without paying any creative work.

[0089] Based on the above method embodiment of the present invention, a corresponding electronic device embodiment is provided.

[0090] An embodiment of the present invention provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the real-time rainfall prediction method described in any one of the present invention is implemented, or when the processor executes the computer program, the functions of the modules in the above-mentioned device embodiments are implemented.

[0091] Exemplarily, the computer program may be divided into one or more modules, which are stored in the memory and executed by the processor to implement the present invention. The one or more modules may be a series of computer program instruction segments capable of implementing specific functions, which are used to describe the execution process of the computer program in the terminal device.

[0092] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0093] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the terminal device, and uses various interfaces and lines to connect various parts of the entire terminal device.

[0094] The memory can be used to store the computer program and / or module, and the processor realizes various functions of the terminal device by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc.; the data storage area can store data created according to the use of the mobile phone, etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0095] Based on the above method embodiment, the present invention provides a storage medium embodiment;

[0096] Another embodiment of the present invention provides a storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute any one of the above-mentioned real-time rainfall prediction methods of the present invention.

[0097] The storage medium is a computer-readable storage medium, and the computer program includes computer program code, which may be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0098] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.

[0099] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A real-time rainfall prediction method, characterized in that: include: Obtaining influencing factors of real-time rainfall; wherein the influencing factors include: temperature, humidity, air pressure, wind speed, dew point, cloud amount, cloud type, cloud top temperature, water vapor content, cloud coverage and terrain type; Inputting the influencing factors of the real-time rainfall into a rainfall prediction model, so that the rainfall prediction model generates a predicted real-time rainfall according to the influencing factors of the real-time rainfall; Wherein, the training of the rainfall prediction model includes: Obtaining the influencing factors of rainfall at a historical moment and the corresponding historical rainfall label; the historical rainfall label is the rainfall at the historical moment; Construct a training set based on the factors affecting rainfall at historical moments and the corresponding historical rainfall labels; Randomly dividing the training set into several batches of training samples according to a preset batch size; Input each batch of training samples into the rainfall prediction model in turn, and train the rainfall prediction model until a preset number of training times is reached; wherein, the rainfall prediction model outputs the rainfall corresponding to the training sample each time a batch of training samples is received; calculates the loss function value through the loss function according to the rainfall and the corresponding historical rainfall label; and uses the optimizer to update the rainfall prediction model according to the loss function value.

2. A real-time rainfall prediction method as claimed in claim 1, characterized in that: The temperature, humidity, air pressure, wind speed and dew point are observed by meteorological stations; the cloud type, cloud top temperature, water vapor content and cloud coverage are collected by meteorological satellites; and the terrain type is obtained by a geographic information system.

3. A real-time rainfall prediction method as claimed in claim 2, characterized in that: Before obtaining the factors affecting rainfall at historical moments, it also includes: Get the original factors affecting rainfall at historical moments; Clean the original influencing factors and align the timestamps to obtain time series data; Convert time series data into numerical data and generate the factors affecting rainfall at historical moments.

4. A real-time rainfall prediction method as claimed in claim 3, characterized in that: After inputting the influencing factors of the real-time rainfall into the rainfall prediction model so that the rainfall prediction model generates the predicted real-time rainfall according to the influencing factors of the real-time rainfall, the method further includes: Compare the predicted real-time rainfall with the set threshold and generate a corresponding warning signal; According to the early warning signal, corresponding emergency response measures are generated.

5. A real-time rainfall prediction device, characterized in that: include: Influencing factor acquisition module and real-time rainfall prediction module; The influencing factor acquisition module is used to acquire the influencing factors of real-time rainfall; wherein the influencing factors include: temperature, humidity, air pressure, wind speed, dew point, cloud amount, cloud type, cloud top temperature, water vapor content, cloud coverage and terrain type; The real-time rainfall prediction module is used to input the influencing factors of the real-time rainfall into the rainfall prediction model, so that the rainfall prediction model generates the predicted real-time rainfall according to the influencing factors of the real-time rainfall; Wherein, the training of the rainfall prediction model includes: Obtaining the influencing factors of rainfall at a historical moment and the corresponding historical rainfall label; the historical rainfall label is the rainfall at the historical moment; Construct a training set based on the factors affecting rainfall at historical moments and the corresponding historical rainfall labels; Randomly dividing the training set into several batches of training samples according to a preset batch size; Input each batch of training samples into the rainfall prediction model in turn, and train the rainfall prediction model until a preset number of training times is reached; wherein, the rainfall prediction model outputs the rainfall corresponding to the training sample each time a batch of training samples is received; calculates the loss function value through the loss function according to the rainfall and the corresponding historical rainfall label; and uses the optimizer to update the rainfall prediction model according to the loss function value.

6. A real-time rainfall prediction device as claimed in claim 5, characterized in that: The temperature, humidity, air pressure, wind speed and dew point are observed by meteorological stations; the cloud type, cloud top temperature, water vapor content and cloud coverage are collected by meteorological satellites; and the terrain type is obtained by a geographic information system.

7. A real-time rainfall prediction device as claimed in claim 6, characterized in that: Before obtaining the factors affecting rainfall at historical moments, it also includes: Get the original factors affecting rainfall at historical moments; Clean the original influencing factors and align the timestamps to obtain time series data; Convert time series data into numerical data and generate the factors affecting rainfall at historical moments.

8. A real-time rainfall prediction device as claimed in claim 7, characterized in that: After inputting the influencing factors of the real-time rainfall into the rainfall prediction model so that the rainfall prediction model generates the predicted real-time rainfall according to the influencing factors of the real-time rainfall, the method further includes: Compare the predicted real-time rainfall with the set threshold and generate a corresponding warning signal; According to the early warning signal, corresponding emergency response measures are generated.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the real-time rainfall prediction method according to any one of claims 1 to 4 can be implemented.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the real-time rainfall prediction method described in any one of claims 1 to 4 can be implemented.

Images