Lightning short-term and imminent forecasting method and device
Through the combination of audio monitoring and three-dimensional lightning detection network, the lightning short-term forecast model is verified and optimized, and the problem of inaccurate position prediction in the short-term forecast of strong convective weather is solved, and a higher accuracy and accuracy of lightning short-term forecast is achieved.
Patent Information
- Application Number
- CN202410980494.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-22
- Publication Date
- 2024-11-22
AI Technical Summary
In the short-term forecast of strong convective weather, the problems of inaccurate position prediction, short radar echo extrapolation time, lack of physical constraints in deep learning models and large sample data errors, making it difficult to achieve accurate short-term forecast of lightning.
Audio monitoring is combined with a three-dimensional lightning detection network, and accurate thunder position is obtained through noise cancellers, and short-term lightning forecast model is established using audio data verification and satellite and radar data to improve position accuracy and forecast accuracy.
It improves the position recognition accuracy of lightning short-term forecasts, reduces the impact of errors, enhances the accuracy of forecasts and resource utilization efficiency, and provides a more accurate lightning short-term warning.
Smart Images

Figure CN119001919A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of meteorological forecast application, and in particular relates to a method and a device for short-term and impending lightning forecasting. Background Art
[0002] Compared with short-term forecasts, short-term forecasts have a shorter forecast time limit, mainly concentrated in 0 hours to 12 hours, and focus on the forecast of small and medium-scale weather systems, especially severe convective weather systems. Compared with large-scale weather systems, severe convective weather systems have the characteristics of short life history and strong suddenness. Their life cycle is as short as a few minutes to tens of minutes, and the longest is no more than ten hours. Due to the rapid generation and disappearance, forecasting is more difficult; currently, the short-term forecast of severe convection mainly adopts the potential forecast method to predict the occurrence area and probability of severe convection. It is still difficult to accurately predict the specific location and time of severe convection. However, short-term forecasts several hours in advance can clearly capture the location of severe convective weather systems, so that the specific location of severe convection can be predicted, and warning forecasts can be issued several hours in advance, thereby reducing the hazards of severe convective weather.
[0003] Due to the chaotic characteristics of the atmosphere, the numerical forecast model is extremely sensitive to the initial value. A small error in the initial value of the input meteorological element may lead to completely different results. The small error will have a great impact on the prediction of the position of severe convection. In addition, due to the spin-up problem, the numerical forecast needs to be balanced and adjusted in the first few hours, and the forecast results cannot be used. Therefore, it is difficult for the numerical forecast technology based on physical equations to make effective short-term forecasts for severe convection at the kilometer scale. At present, the short-term forecast of severe convection adopts traditional radar echo extrapolation methods, such as cross-correlation algorithm, monomer centroid method and optical flow method. These methods only infer the echo position at the next moment based on the radar echo images at several moments, and ignore the nonlinear motion of the small and medium-scale atmospheric systems in the radar echo in actual situations. There are problems such as insufficient utilization of historical radar data and short extrapolation time; the second is to use artificial intelligence methods, especially deep learning methods. Although deep learning methods are good at modeling nonlinear systems, statistical models have inherent small sample over-smoothing problems, the forecast solution process lacks physical conservation law constraints, and the generated data images are seriously blurred and distorted. If there is a large error in the lightning detection sample data, the model established based on the sample data will lead to a larger deviation in the forecast. Therefore, it is necessary to design a method and device for short-term lightning forecasting that can cooperate with traditional lightning detection through audio monitoring to obtain a more accurate location. Summary of the invention
[0004] In order to solve the above problems, the purpose of the present invention is to provide a method and device for short-term lightning forecasting, which are used to obtain the precise lightning strike location through an audio monitoring and verification module, and to more accurately forecast short-term lightning.
[0005] In order to achieve the above object, the technical solution of the present invention is as follows:
[0006] On the one hand, a method and device for short-term lightning forecasting are provided, comprising the following steps:
[0007] Step 1: Detect lightning and obtain the lightning strike location. When lightning is detected, extract the lightning strike location as the first lightning strike location.
[0008] Step 2: The audio within the coverage area is collected in real time by the set audio monitoring module, and according to the set preset verification range, the audio data collected by the audio monitoring module within the preset verification range is extracted with the first lightning strike point as the center, and processed into the first audio data;
[0009] Step 3: verifying the first lightning strike location based on the first audio data, and when the position verification deviation is less than the deviation threshold, directly using the first lightning strike location as the predicted lightning strike location;
[0010] When the verification position deviation is greater than or equal to the deviation threshold, the first lightning landing point position is updated to the second lightning landing point position based on the position information of each audio monitoring module and the collected first audio data, and the second lightning landing point position is used as the predicted lightning landing point position;
[0011] Step 4: Obtain prediction-related data with the predicted lightning strike point as the center, and input the prediction-related data into the lightning short-term forecast model, and output the lightning short-term forecast result through the lightning short-term forecast model.
[0012] Furthermore, in step one, lightning detection and monitoring of lightning strike locations are both performed through a three-dimensional lightning detection network; the three-dimensional lightning monitoring network detects lightning activity parameters in real time, and the lightning activity parameters include: the time, type, longitude and latitude, altitude and peak current intensity of lightning occurrence; and the lightning strike location is obtained based on the longitude and latitude of the lightning occurrence.
[0013] Further, in step 2, the audio monitoring module is based on whether it is located in the short-term lightning forecast area at the previous moment; if it is located in the short-term lightning forecast area, the corresponding audio monitoring module is started and real-time audio collection is performed.
[0014] Further, it is characterized in that: verifying the first lightning strike point position based on the first audio data comprises the following steps:
[0015] The time when the audio monitoring module is expected to receive the first audio data is calculated by the following formula:
[0016] Δt=d / c
[0017] Where Δt is the time when the audio monitoring module is expected to receive the first audio data, d is the distance between the audio monitoring module that receives the first audio data and the first lightning strike point, and c is the sound propagation speed, which is a constant of 340 m / s;
[0018] Compare the time when the audio monitoring module actually receives the first audio data with the time when the audio monitoring module expects to receive the first audio data. If the time difference between the two is less than the threshold, then it is determined that the position verification deviation is less than the deviation threshold; if the time difference between the two is greater than or equal to the threshold, then it is determined that the position verification deviation is greater than or equal to the deviation threshold.
[0019] Further, based on the position information of each audio monitoring module and the collected first audio data, updating the first lightning landing position to the second lightning landing position includes the following steps: randomly selecting three audio monitoring modules that receive the first audio data, based on the following formula:
[0020] d1=c·ΔT1
[0021] d2=c·ΔT2
[0022] d3=c·ΔT3
[0023] Where c is the speed of sound propagation, which is a constant of 340 m / s; d1, d2, and d3 are the distances between the three audio monitoring modules and the actual lightning strike location; ΔT1, ΔT2, and ΔT3 are the times when the three audio monitoring modules actually receive the audio;
[0024] The radii d1, d2 and d3 calculated by the above formula are used as the circle center, and the circular area is divided with the positions of the three audio monitoring modules as the center. The intersection of the three circular areas is selected as the second lightning landing point.
[0025] Further, in step 4, a lightning short-term forecast model is established based on satellite, radar and lightning observation data. Establishing the lightning short-term forecast model includes the following steps:
[0026] Data collection steps: Collect satellite, radar and lightning observation data;
[0027] Feature extraction step: extracting features related to lightning activity from the collected satellite, radar and lightning observation data;
[0028] Model selection and establishment steps: Select and establish a lightning short-term forecast model;
[0029] Model training steps: Use the collected satellite, radar and lightning observation data to establish a data set, use the data set to train the lightning short-term forecast model and optimize the parameters of the lightning short-term forecast model;
[0030] Evaluation step: Evaluate the performance of the trained lightning short-term forecast model, and put the lightning short-term forecast model that passes the performance evaluation into use.
[0031] Furthermore, the audio monitoring module is arranged on a fixed component, the fixed component includes a fixed shaft, a lightning rod is fixedly connected to the top end of the fixed shaft, a solar panel is fixedly connected to the fixed shaft, the solar panel is electrically connected to a battery pack, the solar panel and the battery pack are electrically connected, and an energy storage box is also fixedly connected to the fixed shaft, and the battery pack is arranged in the energy storage box.
[0032] Furthermore, a noise eliminator is fixedly connected to one side of the audio monitoring module, and the noise eliminator includes:
[0033] Condenser microphone, used to convert sound signals into electrical signals;
[0034] an analog-to-digital converter for converting the electrical signal generated by the condenser microphone into a digital signal;
[0035] An audio signal processing chip for processing the digital signal generated by the analog-to-digital converter and outputting a speech enhancement signal;
[0036] A digital-to-analog converter, used for receiving the speech enhancement signal generated by the audio signal processing chip and converting it into an analog signal;
[0037] The condenser microphone, the analog-to-digital converter, the audio signal processing chip, and the digital-to-analog converter are connected in sequence.
[0038] On the other hand, a lightning short-term forecasting device is provided, which is operated based on any one of the lightning short-term forecasting methods described above, and comprises:
[0039] The first acquisition module is used to detect and monitor the lightning strike location, and extract the first lightning strike location when lightning is detected.
[0040] The second acquisition module is used to collect audio within the coverage area in real time. According to the set preset verification range, with the first lightning landing point as the center, the audio data collected by the audio monitoring module within the preset verification range is extracted and processed into the first audio data.
[0041] The verification module is used to verify the first lightning landing point position of the first audio data. When verifying the position deviation, the lightning landing point position is updated to the second lightning landing point position, and the second lightning landing point position is used as the predicted lightning landing point position.
[0042] The output module is used to input the second lightning strike point position into the lightning short-term forecast model and output it as a lightning short-term forecast result.
[0043] Further: the second verification module is used to update the first lightning landing point position to the second lightning landing point position based on the position information of each audio monitoring module and the collected first audio data, and randomly select three audio monitoring modules that receive the first audio data based on the following formula:
[0044] d1=c·ΔT1
[0045] d2=c·ΔT2
[0046] d3=c·ΔT3
[0047] Where c is the speed of sound propagation, which is a constant of 340 m / s; d1, d2, and d3 are the distances between the three audio monitoring modules and the actual lightning strike location; ΔT1, ΔT2, and ΔT3 are the times when the three audio monitoring modules actually receive the audio;
[0048] The radii d1, d2 and d3 calculated by the above formula are used as the circle center, and the circular area is divided with the positions of the three audio monitoring modules as the center. The intersection of the three circular areas is selected as the second lightning landing point.
[0049] The above scheme has the following beneficial effects:
[0050] 1. The present invention first monitors thunder through an audio monitoring module to obtain the position of thunder, and designs a noise eliminator to avoid capturing noise that may exist due to environmental conditions or the like or to reduce unwanted noise that may exist due to environmental conditions or the like in the captured audio; and verifies and updates the lightning strike point position monitored by the detector to obtain a more accurate position, collects and predicts relevant data based on the verified and updated position information, and performs a short-term lightning forecast. Compared with the prior art, the present invention reduces the influence of various reasons on lightning position detection, can improve the accuracy of position recognition, and makes the lightning strike point position used for short-term lightning forecast more accurate, thereby performing a more accurate short-term lightning forecast.
[0051] 2. The present invention can effectively save resources and reduce waste by monitoring whether it is within the last forecasted lightning short-term forecast area, thereby deciding whether to start the audio monitoring module to perform real-time audio monitoring.
[0052] 3. The present invention verifies the first lightning strike point position by combining the first audio data, and compares the audio time actually received by the audio monitoring module with the audio time received under an ideal state, so as to determine whether there is a deviation, thereby further improving the accuracy of the monitoring position.
[0053] 4. The present invention, after having a deviation, randomly selects three audio monitoring modules to calculate the distance between each audio monitoring module and the sound source, and then uses the distance between each audio monitoring module and the sound source as the radius to obtain the area where the sound source may exist, and based on the intersection of the three areas where the sound source may exist, finally obtains a more accurate lightning position, which can effectively reduce the influence of irrelevant factors on position judgment and provide more complete and accurate position information. BRIEF DESCRIPTION OF THE DRAWINGS
[0054] Figure 1 The figure is a flow chart of the method for forecasting short-term and impending lightning in an embodiment of the present invention.
[0055] Figure 2 Schematic diagram of the structure of a lightning short-term forecasting device in an embodiment of the present invention.
[0056] Figure 3 Schematic diagram of three-point positioning of the method for short-term and impending lightning forecasting in an embodiment of the present invention.
[0057] Figure 4 It is a front view of the device for short-term and impending lightning forecasting in an embodiment of the present invention.
[0058] Figure 5 It is a side view of the device for short-term and impending lightning forecasting in an embodiment of the present invention.
[0059] The reference numerals in the drawings of the specification include: a lightning rod 1, a fixed shaft 2, an energy storage box 3, an audio monitoring module 4, a noise eliminator 401, and a solar panel 5. DETAILED DESCRIPTION
[0060] The following is further described in detail through specific implementation methods:
[0061] Embodiment 1:
[0062] Basically as attached Figure 1-Figure 5 A method for short-term lightning forecasting includes the following steps:
[0063] Step 1: Detect lightning and obtain the location of the lightning strike. In this embodiment, lightning detection and monitoring of the lightning strike location are both performed through a three-dimensional lightning detection network. The three-dimensional lightning monitoring network detects lightning activity parameters in real time. The lightning activity parameters include: the time, type, longitude and latitude of the lightning, altitude and peak current intensity. The lightning strike location is obtained based on the longitude and latitude of the detected lightning. The time of the lightning is used for subsequent position verification of the lightning strike location, and the type, altitude and peak current intensity are used to understand the nature and characteristics of lightning activity, reveal the spatial distribution and changes of lightning activity, improve the accuracy of forecasts, and reflect the energy size and destructive power of lightning activity.
[0064] When lightning is detected, the lightning strike point is extracted as the first lightning strike point, that is, the longitude and latitude where the lightning occurred; the first lightning strike point is also used for subsequent position verification of the lightning strike point.
[0065] Step 2: The audio within the coverage area is collected in real time by the audio monitoring module 4. There are many ways to set the audio monitoring module 4, for example, it can be installed on a wall or on a ceiling in the room. In order to facilitate installation, the present embodiment adopts a pole tower installation method. Specifically, a fixing component is used to fix the audio monitoring module 4. The audio monitoring module 4 is arranged on the fixing component. The fixing component includes a fixing shaft 2. A lightning rod 1 is fixedly connected to the top of the fixing shaft 2. A solar panel 5 is fixedly connected to the fixing shaft 2. The solar panel 5 is electrically connected to a battery pack. The solar panel 5 and the battery pack are electrically connected. An energy storage box 3 is also fixedly connected to the fixing shaft 2. The battery pack is arranged in the energy storage box 3.
[0066] Then, according to the preset verification range, the audio data collected by the audio monitoring module 4 within the preset verification range is extracted with the first lightning landing point as the center, and processed into the first audio data.
[0067] Step 3: verifying the first lightning strike location based on the first audio data, and when the position verification deviation is less than the deviation threshold, directly using the first lightning strike location as the predicted lightning strike location;
[0068] When the verification position deviation is greater than or equal to the deviation threshold, the first lightning landing position is updated to the second lightning landing position based on the position information of each audio monitoring module 4 and the collected first audio data, and the second lightning landing position is used as the predicted lightning landing position.
[0069] The step of verifying the first lightning strike point based on the first audio data includes the following steps:
[0070] The time when the audio monitoring module 4 is expected to receive the first audio data is calculated by the following formula:
[0071] Δt=d / c
[0072] Wherein, Δt is the time when the audio monitoring module 4 is expected to receive the first audio data, d is the distance between the audio monitoring module 4 that receives the first audio data and the first lightning strike point, and c is the sound propagation speed, which is a constant of 340 m / s;
[0073] The time when the audio monitoring module 4 is expected to receive the first audio data is obtained by calculation, and the time when the audio monitoring module 4 actually receives the first audio data is compared with the time when the audio monitoring module 4 expects to receive the first audio data. If the time difference between the two is less than the threshold, it is determined that the position verification deviation is less than the deviation threshold. If the time difference between the two is greater than or equal to the threshold, it is determined that the position verification deviation is greater than or equal to the deviation threshold. Assuming the threshold is 0.2s, the time when the audio monitoring module 4 actually receives the first audio data is 5s, and the audio monitoring module 4 estimates that the time of the first audio data received is 4.9s using this formula. The difference between the two is 0.1s, and the difference is less than the threshold, then it is determined that the position verification deviation is less than the deviation threshold. If it is determined that the position verification deviation is greater than or equal to the deviation threshold, the same applies.
[0074] If it is determined that the position verification deviation is less than the deviation threshold, the position is directly input into the lightning short-term forecast model and output as the lightning short-term forecast result. If it is determined that the position verification deviation is greater than or equal to the deviation threshold, based on the position information of each audio monitoring module 4 and the collected first audio data, the first lightning landing point position is updated to the second lightning landing point position, including the following steps:
[0075] Three audio monitoring modules 4 that receive the first audio data are randomly selected based on the following formula:
[0076] d1=c·ΔT1
[0077] d2=c·ΔT2
[0078] d3=c·ΔT3
[0079] Wherein, c is the sound propagation speed, which is a constant of 340 m / s, d1, d2 and d3 are the distances between the three audio monitoring modules 4 and the actual lightning landing point, and ΔT1, ΔT2 and ΔT3 are the times when the three audio monitoring modules 4 actually receive the audio;
[0080] Combined with Figure 3 As shown, d1, d2 and d3 calculated by the above formula are used as the radius, and the positions of the three audio monitoring modules 4 are used as the center of the circle to divide the circular area, and the intersection of the three circular areas is selected as the second lightning landing point. At this time, the second landing point is a more accurate lightning landing point.
[0081] Step 4: Obtain prediction-related data with the predicted lightning strike point as the center, and input the prediction-related data into the lightning short-term forecast model, and output the lightning short-term forecast result through the lightning short-term forecast model.
[0082] The establishment of the lightning short-term forecast model is to use satellite, radar and lightning observation data to perform lightning nowcasting and establish a multi-source lightning prediction model; the steps of constructing the multi-source lightning prediction model of this embodiment are as follows:
[0083] Data collection step, collect data related to the prediction, such as radar data: radar can provide lightning information, including the time, longitude and latitude of lightning, and is an important data source for lightning forecast models; satellite data: satellite images can provide cloud characteristics and changes; ground observation data: including temperature, humidity and other parameters measured by meteorological stations; lightning detection data: real-time monitoring of lightning activities, which can indicate the intensity, height, type and distribution of lightning;
[0084] Feature extraction step, extracting features related to lightning activity from the collected data, such as cloud vertical structure, cloud top temperature, vorticity, horizontal wind speed, etc., and using statistical methods and machine learning techniques, such as feature selection algorithms, to determine which features have the most predictive power for lightning forecasting;
[0085] Model selection and establishment steps: statistical models: such as logistic regression, linear regression, etc., can be modeled based on historical data; machine learning models: such as support vector machines, random forests, neural networks, etc., these models can more complexly capture the nonlinear relationship between features; hybrid models: combine multiple models to improve prediction accuracy and stability, select a suitable model from them and establish the model;
[0086] Model training step: Use historical data sets to train the model. Usually, the data sets are divided into training sets and test sets to verify the generalization ability of the model. Cross-validation and other techniques are used to adjust the model parameters to prevent overfitting or underfitting and use related techniques to adjust the model parameters.
[0087] The performance optimization evaluation step uses relevant indicators, such as prediction accuracy, recall rate, ROC curve, etc. to evaluate model performance, and optimizes the model based on the evaluation results, which may include improving feature selection, adjusting model parameters, etc.
[0088] Embodiment 2:
[0089] As attached Figure 2 As shown: a lightning short-term forecasting device, which is based on the lightning short-term forecasting method of Example 1, includes a first acquisition module, a second acquisition module, a verification module and an output module, and the first acquisition module, the second acquisition module, the verification module and the output module interact and communicate with each other.
[0090] The first acquisition module is used to detect and monitor the lightning strike location. The lightning detection and monitoring of the lightning strike location are both performed through a three-dimensional lightning detection network. When lightning is detected, the first lightning strike location is extracted. The acquisition module is mainly used to collect the approximate lightning strike location;
[0091] The second acquisition module collects the audio within the coverage area in real time through the audio monitoring module 4. In order to facilitate installation, this embodiment adopts a pole tower installation method. Then, according to the preset verification range, the first lightning strike point is taken as the center, and the audio data within the preset verification range is extracted and processed into the first audio data. The acquisition module is mainly used to collect audio data;
[0092] A verification module, used to verify the first lightning strike location, perform position verification on the first lightning strike location based on the first audio data, and directly use the first lightning strike location as a predicted lightning strike location when a position verification deviation is less than a deviation threshold;
[0093] When the verification position deviation is greater than or equal to the deviation threshold, the first lightning landing position is updated to the second lightning landing position based on the position information of each audio monitoring module 4 and the collected first audio data, and the second lightning landing position is used as the predicted lightning landing position;
[0094] Specifically, based on the position information of each audio monitoring module 4 and the collected first audio data, the first lightning landing position is updated to the second lightning landing position, and at least three audio monitoring modules 4 that receive the first audio data are randomly selected based on the following formula:
[0095] d1=c·ΔT1
[0096] d2=c·ΔT2
[0097] d3=c·ΔT3
[0098] Where ΔT1, ΔT2 and ΔT3 are the times when the three audio monitoring modules 4 actually receive the audio, d1, d2 and d3 are the distances between the three audio monitoring modules 4 and the actual lightning strike point, and c is the sound propagation speed, which is a constant of 340 m / s.
[0099] Circular areas are divided with d1, d2 and d3 as radii and the positions of the three audio monitoring modules 4 as centers, and the intersection of the three circular areas is selected as the second lightning landing point.
[0100] The verification module verifies the lightning strike location obtained through audio data and the initial lightning strike location, and can determine whether the obtained location is accurate;
[0101] The output module is used to obtain prediction-related data with the predicted lightning strike point as the center, and input the prediction-related data into the lightning short-term forecast model. The lightning short-term forecast model outputs the lightning short-term forecast result to achieve the final lightning short-term forecast.
[0102] Embodiment 3:
[0103] As attached Figure 4 and attached Figure 5 As shown: the noise eliminator 401 used in Example 1 is because the audio monitoring module 4 is installed outdoors and operates in thunderstorm weather, so there may be noise interference. In order to reduce noise interference and improve the audio clarity collected by the audio monitoring module 4, this embodiment is fixedly connected to one side of the audio monitoring module 4. The noise eliminator 401 includes:
[0104] Condenser microphone, used to convert sound signals into electrical signals;
[0105] an analog-to-digital converter for converting the electrical signal generated by the condenser microphone into a digital signal;
[0106] An audio signal processing chip for processing the digital signal generated by the analog-to-digital converter and outputting a speech enhancement signal;
[0107] A digital-to-analog converter, used for receiving the speech enhancement signal generated by the audio signal processing chip and converting it into an analog signal;
[0108] The condenser microphone, the analog-to-digital converter, the audio signal processing chip, and the digital-to-analog converter are connected in sequence.
[0109] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.
[0110] The above is only an embodiment of the present invention. The common sense such as the known specific structure and characteristics in the scheme is not described in detail here. The ordinary technicians in the relevant field know all the common technical knowledge in the technical field of the invention before the application date or priority date, can know all the existing technologies in the field, and have the ability to apply the conventional experimental means before that date. The ordinary technicians in the relevant field can improve and implement this scheme in combination with their own abilities under the enlightenment given by this application. Some typical known structures or known methods should not become obstacles for ordinary technicians in the relevant field to implement this application. It should be pointed out that for those skilled in the art, without departing from the structure of the present invention, several deformations and improvements can be made, which should also be regarded as the protection scope of the present invention, which will not affect the effect of the implementation of the present invention and the practicality of the patent. The protection scope required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.
Claims
1. A method for short-term lightning forecasting, characterized in that: The following steps are involved: Step 1: Detect lightning and obtain the lightning strike location. When lightning is detected, extract the lightning strike location as the first lightning strike location. Step 2: The audio within the coverage area is collected in real time by the set audio monitoring module, and according to the set preset verification range, the audio data collected by the audio monitoring module within the preset verification range is extracted with the first lightning strike point as the center, and processed into the first audio data; Step 3: verifying the first lightning strike location based on the first audio data, and when the position verification deviation is less than the deviation threshold, directly using the first lightning strike location as the predicted lightning strike location; When the verification position deviation is greater than or equal to the deviation threshold, the first lightning landing point position is updated to the second lightning landing point position based on the position information of each audio monitoring module and the collected first audio data, and the second lightning landing point position is used as the predicted lightning landing point position; Step 4: Obtain prediction-related data with the predicted lightning strike point as the center, and input the prediction-related data into the lightning short-term forecast model, and output the lightning short-term forecast result through the lightning short-term forecast model.
2. The method for short-term and impending lightning forecasting according to claim 1, characterized in that: In step one, lightning detection and monitoring of lightning strike locations are both performed through a three-dimensional lightning detection network; the three-dimensional lightning monitoring network detects lightning activity parameters in real time, and lightning activity parameters include: the time, type, longitude and latitude, height, and peak current intensity of lightning occurrence; the lightning strike location is obtained based on the longitude and latitude of the lightning occurrence.
3. The method for short-term and impending lightning forecasting according to claim 1, characterized in that: In step 2, the audio monitoring module is based on whether it is located in the short-term lightning forecast area at the previous moment; if it is located in the short-term lightning forecast area, the corresponding audio monitoring module is started and real-time audio collection is performed.
4. The method for short-term and impending lightning forecasting according to claim 1, characterized in that: Verifying the first lightning strike point based on the first audio data includes the following steps: The time when the audio monitoring module is expected to receive the first audio data is calculated by the following formula: Δt=d / c Wherein, Δt is the time when the audio monitoring module is expected to receive the first audio data, d is the distance between the audio monitoring module that receives the first audio data and the first lightning strike point, and c is the sound propagation speed, which is a constant of 340 m / s; Compare the time when the audio monitoring module actually receives the first audio data with the time when the audio monitoring module expects to receive the first audio data. If the time difference between the two is less than the threshold, then it is determined that the position verification deviation is less than the deviation threshold; if the time difference between the two is greater than or equal to the threshold, then it is determined that the position verification deviation is greater than or equal to the deviation threshold.
5. The method for short-term and impending lightning forecasting according to claim 1, characterized in that: Based on the position information of each audio monitoring module and the collected first audio data, updating the first lightning landing point position to the second lightning landing point position includes the following steps: randomly selecting three audio monitoring modules that receive the first audio data, based on the following formula: d1=c·ΔT1 d2=c·ΔT2 d3=c·ΔT3 Where c is the sound propagation speed, which is a constant of 340 m / s; d1, d2, and d3 are the distances between the three audio monitoring modules and the actual lightning strike location; ΔT1, ΔT2, and ΔT3 are the times when the three audio monitoring modules actually receive the audio; The radii d1, d2 and d3 calculated by the above formula are used as the circle center, and the circular area is divided with the positions of the three audio monitoring modules as the center. The intersection of the three circular areas is selected as the second lightning landing point.
6. The method for short-term and impending lightning forecasting according to claim 1, characterized in that: In step 4, a lightning short-term forecast model is established based on satellite, radar and lightning observation data. Establishing a lightning short-term forecast model includes the following steps: Data collection steps: Collect satellite, radar and lightning observation data; Feature extraction step: extracting features related to lightning activity from the collected satellite, radar and lightning observation data; Model selection and establishment steps: Select and establish a lightning short-term forecast model; Model training steps: Use the collected satellite, radar and lightning observation data to establish a data set, use the data set to train the lightning short-term forecast model and optimize the parameters of the lightning short-term forecast model; Evaluation step: Evaluate the performance of the trained lightning short-term forecast model, and put the lightning short-term forecast model that passes the performance evaluation into use.
7. The method for short-term and impending lightning forecasting according to claim 1, characterized in that: The audio monitoring module is arranged on a fixed component, which includes a fixed shaft, a lightning rod is fixedly connected to the top of the fixed shaft, a solar panel is fixedly connected to the fixed shaft, the solar panel is electrically connected to a battery pack, the solar panel and the battery pack are electrically connected, and an energy storage box is also fixedly connected to the fixed shaft, and the battery pack is arranged in the energy storage box.
8. The method for short-term and impending lightning forecasting according to claim 7, characterized in that: A noise eliminator is fixedly connected to one side of the audio monitoring module. The noise eliminator includes: Condenser microphone, used to convert sound signals into electrical signals; an analog-to-digital converter for converting the electrical signal generated by the condenser microphone into a digital signal; An audio signal processing chip for processing the digital signal generated by the analog-to-digital converter and outputting a speech enhancement signal; A digital-to-analog converter, used for receiving the speech enhancement signal generated by the audio signal processing chip and converting it into an analog signal; The condenser microphone, the analog-to-digital converter, the audio signal processing chip, and the digital-to-analog converter are connected in sequence.
9. A lightning short-term forecasting device, operating based on the lightning short-term forecasting method according to any one of claims 1 to 8, characterized in that: include: The first acquisition module is used to detect and monitor the lightning strike location, and extract the first lightning strike location when lightning is detected. The second acquisition module is used to collect audio within the coverage area in real time. According to the set preset verification range, with the first lightning landing point as the center, the audio data collected by the audio monitoring module within the preset verification range is extracted and processed into the first audio data. The verification module is used to verify the first lightning landing point position of the second audio data. When verifying the position deviation, the lightning landing point position is updated to the second lightning landing point position, and the second lightning landing point position is used as the predicted lightning landing point position. The output module is used to input the second lightning strike point position into the lightning short-term forecast model and output it as a lightning short-term forecast result.
10. The lightning short-term forecasting device according to claim 9, characterized in that: The second verification module is used to update the first lightning landing point position to the second lightning landing point position based on the position information of each audio monitoring module and the collected first audio data, and randomly select three audio monitoring modules that receive the first audio data based on the following formula: d1=c·ΔT1 d2=c·ΔT2 d3=c·ΔT3 Where c is the sound propagation speed, which is a constant of 340 m / s; d1, d2, and d3 are the distances between the three audio monitoring modules and the actual lightning strike location; ΔT1, ΔT2, and ΔT3 are the times when the three audio monitoring modules actually receive the audio; Using d1, d2 and d3 calculated by the above formula as the radius and the three audio monitoring module positions as the center, the circular area is divided, and the intersection of the three circular areas is selected as the second lightning landing point.
