A lightning positioning method, device and storage medium

Abstract

The present invention relates to the field of lightning detection, and discloses a lightning positioning method, device and storage medium, the method comprising the steps of: laying multiple DAS hosts equipped with Beidou synchronization system in a predetermined area through optical fiber; using Beidou synchronization system to complete time synchronization of DAS hosts; using each DAS host to collect lightning sound waves, and recording the determined timestamp when the lightning sound waves arrive at each DAS host; transmitting the signal data with Beidou synchronization timestamp collected by each DAS host to a central processor or a data center, calculating the arrival time difference of the same lightning event signal between different DAS hosts through TDOA positioning algorithm; calculating the precise coordinates of lightning in three-dimensional space according to the time difference, and realizing lightning positioning. The present invention has the beneficial effects of being able to be directly applied and having high lightning positioning accuracy.

Description

Technical Field

[0001] The present invention relates to the field of lightning detection, and in particular to a lightning locating method, device and storage medium. Background Art

[0002] The hazards of lightning to cities mainly include the following aspects: threatening personal safety, causing damage to buildings, electrical equipment, public service facilities and economic losses. Therefore, it is very important to count the distribution areas of lightning in cities.

[0003] The existing lightning location methods mainly include the following:

[0004] A lightning positioning method, system, device and storage medium include the following steps:

[0005] Acquire sound data and electromagnetic wave data to be identified; determine target thunder information including at least one target thunder signal in the sound data using the sound signal and a pre-trained first model;

[0006] The target thunder signal, the electromagnetic wave signal and the pre-trained second model are used to determine the target lightning information including the target electromagnetic wave signal corresponding to the target thunder signal in the electromagnetic wave data; the target position information of the lightning is determined according to the target thunder information and the target lightning information to achieve lightning positioning. In the above manner, the voiceprint recognition technology is combined with the neural network algorithm to establish a high-precision thunder recognition model and an acoustic-electric matching model to improve the thunder signal recognition accuracy, thereby achieving high-precision positioning of lightning by a single detection station. Due to the high-precision recognition of thunder, the lightning signal contained in the thunder signal can also be extracted for comprehensive lightning monitoring.

[0007] However, the above method still has some problems in application implementation. For example, the performance of the pre-trained first model and the second model largely determines the accuracy of positioning. If the data set used for model training is significantly different from the thunder and electromagnetic wave characteristics in the actual application scenario, it may lead to weak generalization ability of the model and fail to accurately identify and match the actual monitored signals. In addition, the complexity of model updating and maintenance may also become a problem in practical applications. Summary of the invention

[0008] The purpose of the present invention is to provide a lightning location method, device and storage medium to solve the problem that the prior art cannot be directly applied and the lightning location is inaccurate.

[0009] A lightning location method comprises the following steps:

[0010] S1: Multiple DAS hosts equipped with Beidou synchronization system are laid in the predetermined area through optical fiber;

[0011] S2: Use the BeiDou synchronization system to complete the time synchronization of the DAS host;

[0012] S3: using each DAS host to collect lightning sound waves, and recording the confirmed timestamp when the lightning sound wave signal arrives at each DAS host;

[0013] S4: The signal data with Beidou synchronization timestamp collected by each DAS host is transmitted to the central processor or data center, and the arrival time difference of the same lightning event signal between different DAS hosts is calculated through the TDOA positioning algorithm;

[0014] S5: Calculate the precise coordinates of lightning in three-dimensional space based on the time difference to locate the lightning.

[0015] A storage medium stores instructions and data for implementing a lightning locating method.

[0016] A lightning locating device comprises: a processor and the storage medium; the processor loads and executes instructions and data in the storage medium to implement a lightning locating method.

[0017] The beneficial effects provided by the present invention are:

[0018] (1) High-precision Beidou time synchronization: The Beidou satellite navigation system is used to achieve microsecond-level time synchronization between distributed optical fiber acoustic sensor hosts, ensuring the accuracy of recording the moment when lightning events occur, and providing a solid time reference for positioning calculations.

[0019] (2) Distributed fiber optic sensing network: The DAS network is constructed by utilizing the fiber optic resources in the existing composite overhead cables. Each access point is equipped with a fiber optic acoustic emission sensor host to achieve wide-area, multi-angle, and multi-level lightning acoustic wave monitoring, providing rich multi-point observation data to improve positioning accuracy.

[0020] (3) Direct sensing of lightning sound waves by optical fiber: Optical fiber is used as a direct sensing medium to capture the vibration signal caused by lightning sound waves by detecting the change in light phase. It has the advantages of anti-electromagnetic interference, corrosion resistance, and long-distance transmission. It is especially suitable for lightning monitoring in complex environments.

[0021] (4) Real-time lightning geographic distribution map: Realize real-time visualization of positioning results, generate lightning geographic distribution map, and intuitively display the spatial distribution, which helps to quickly assess lightning disaster risks and provide decision support for emergency response and resource scheduling.

[0022] (5) Deep learning-assisted noise discrimination and positioning algorithm integration: Integrate deep learning models to effectively distinguish lightning acoustic signals from various environmental noises (such as cars, excavators, motors, etc.), enhance the system's anti-interference ability in complex environments, and ensure the accuracy of lightning event detection and positioning.

[0023] (6) It explicitly proposes the use of nonlinear optimization methods (such as the least squares method) for positioning calculations, elaborates on the positioning calculation process (calculating distance using time difference, constructing a nonlinear equation group, and optimizing using the least squares method), and provides a specific implementation path for the positioning algorithm.

[0024] (7) Systematic Solution: Provides a complete, systematic lightning monitoring and positioning solution that covers hardware (distributed fiber optic sensor network, Beidou synchronization system) and software (positioning algorithm, real-time geographic distribution map generation). It is highly integrated and practical, and presents a more comprehensive, systematic and technologically advanced lightning monitoring system compared to existing technologies. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is a schematic flow chart of the method of the present invention;

[0026] Figure 2 It is a working schematic diagram of the hardware device of the present invention. DETAILED DESCRIPTION

[0027] To make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0028] Before formally describing the present invention, the scheme of the present invention is first generally described for easy understanding.

[0029] Please refer to Figure 1 , Figure 1 is a schematic diagram of the process of the present invention;

[0030] The present invention provides a lightning locating method, comprising the following steps:

[0031] S1: Multiple DAS hosts equipped with Beidou synchronization system are laid in the predetermined area through optical fiber;

[0032] It should be noted that the present invention first relates to a lightning monitoring network constructed using distributed acoustic fiber sensing technology (DAS), in which there are multiple DAS hosts equipped with Beidou synchronization system, which are laid in a predetermined area through optical fibers to capture vibration signals generated by lightning events.

[0033] S2: Use the BeiDou synchronization system to complete the time synchronization of the DAS host;

[0034] It should be noted that all DAS hosts use the Beidou satellite navigation system for high-precision time synchronization, ensuring that the arrival time of the lightning event signal recorded by each host has an absolute time reference with extremely small errors.

[0035] S3: using each DAS host to collect lightning sound waves and record the confirmed timestamp when the sound wave signal arrives at each DAS host;

[0036] In the present invention, when a lightning event occurs, the sound waves generated by the lightning propagate along the optical fiber, and the DAS host detects these vibration signals through the phase changes of the coherent Rayleigh scattered light and records the exact timestamps of the signals arriving at their respective locations.

[0037] It should be noted that the specific process of using the DAS host to adopt the lightning sound wave in step S3 is as follows:

[0038] S31, obtaining the original sound wave data collected by the DAS host, and preprocessing it to obtain preprocessed data;

[0039] Step S31 mainly performs necessary preprocessing (such as noise reduction, standardization, etc.) on the original sound wave data collected by the DAS system and converts it into an input format acceptable to the lightning recognition model.

[0040] S32, inputting the preprocessed data into a pre-trained lightning recognition model to obtain the probability that the sound wave signal is a lightning signal;

[0041] It should be noted that the process of pre-training the lightning recognition model in step S32 is as follows:

[0042] S321, collecting lightning signals of different lightning acoustic signals, environmental noise, and mixed environmental noise;

[0043] As an embodiment, the present invention collects a large number of audio samples including thunder and lightning sound signals, various environmental noises (such as car engine sounds, excavator operation sounds, motor running sounds, etc.) and mixed scenes (thunder and lightning sounds coexist with environmental noise).

[0044] Furthermore, the present invention also performs data annotation, where samples are marked by professionals or automated algorithms to clearly identify thunder and lightning sounds, different types of environmental noise, and their start and end times.

[0045] S322, performing time-frequency analysis on the signal in step S321 to obtain time-frequency features;

[0046] It should be noted that the time-frequency characteristics of the audio samples are extracted by applying time-frequency analysis techniques (such as short-time Fourier transform, Mel frequency cepstrum coefficients, etc.).

[0047] S323, combining the time-frequency features with the voiceprint features to construct a training set and a test set;

[0048] As an embodiment, the present invention also combines voiceprint features (such as pitch, loudness, spectrum, cepstrum, etc.) and environmental features (such as wind speed, humidity, air pressure, etc.) to further enrich the feature dimensions and improve the resolution ability of the model.

[0049] S324: train the lightning recognition model using the training set, and test the lightning recognition model using the test set to obtain a pre-trained lightning recognition model.

[0050] It should be noted that the lightning recognition model of the present invention selects a deep learning architecture suitable for audio classification tasks, such as convolutional neural network (CNN), recurrent neural network (RNN), long short-term memory network (LSTM), Transformer, etc., or their variants and combinations.

[0051] By setting reasonable parameters such as the number of network layers, number of nodes, activation function, regularization, etc., a deep learning model with efficient recognition and generalization capabilities is constructed.

[0052] The training process of the lightning recognition model is as follows:

[0053] (1) Use labeled data sets to train deep learning models, and use loss functions such as cross entropy and mean square error to measure the gap between model predictions and actual labels.

[0054] (2) Use optimization algorithms (such as Adam, SGD, etc.) to update model parameters until the performance of the model on the validation set reaches the predetermined standard or converges.

[0055] Performance evaluation:

[0056] (1) An independent test set is used to evaluate the performance of the model in the task of distinguishing lightning sound signals from various environmental noises. The indicators include accuracy, recall rate, F1 score, confusion matrix, etc.

[0057] (2) Analyze the performance of the model under different signal-to-noise ratio conditions to ensure that it can still maintain good resolution in complex noise environments.

[0058] S33. When the probability that the sound wave signal is a lightning signal is greater than a preset value, the sound wave signal at this time is used as a vibration signal of a lightning vibration wave.

[0059] It should be noted that in the present invention, the pre-processed sound wave data is sent to the trained lightning recognition model, and the model outputs the probability distribution of the lightning sound signal and various environmental noises.

[0060] Set the threshold strategy. When the model determines that the probability of a lightning sound signal exceeds the threshold, a lightning event is considered to be detected, otherwise it is considered to be environmental noise.

[0061] In addition, the present invention is linked with the positioning algorithm:

[0062] For data identified as lightning acoustic signals, the positioning algorithm is triggered for further processing to calculate the location of the lightning event.

[0063] Data identified as environmental noise is ignored or recorded as interference to avoid misleading the positioning results.

[0064] S4: The signal data with Beidou synchronization timestamp collected by each DAS host is transmitted to the central processor or data center, and the arrival time difference of the same lightning event signal between different DAS hosts is calculated through the TDOA positioning algorithm;

[0065] S5: Calculate the precise coordinates of lightning in three-dimensional space based on the time difference to locate the lightning.

[0066] It should be noted that step S5 is specifically as follows:

[0067] S51, correcting the speed of sound according to the temperature;

[0068] As an embodiment, the propagation speed of thunder waves in the air is indeed affected by a variety of meteorological conditions, especially temperature, humidity and wind speed. Among them, the influence of temperature is the most significant. As the temperature rises, the propagation speed of sound in the air will also increase. This is because high temperature makes the movement of air molecules more active, thereby improving the efficiency of sound waves propagating in the gas medium. The effect of humidity on the speed of sound is relatively small, but in a very humid environment, the speed of sound may be slightly increased due to the presence of moisture. As for wind speed, it mainly indirectly affects the propagation of sound waves through the refraction and scattering effects of sound waves caused by uneven wind in the troposphere, especially in the case of large-scale or long-distance propagation. More obvious, but it is usually not a big impact on short-distance sound speed measurement. To obtain the exact propagation speed of thunder waves in the air at a specific location and a specific time in real time, it is necessary to consider the local meteorological conditions at the time and calculate them in combination with relevant physical models.

[0069] If an exact value is required, the relationship between the speed of sound and air temperature and humidity can be approximately expressed using an empirical formula. For more precise positioning applications, it is indeed necessary to perform a correction calculation on the speed of sound.

[0070] Here is a simple empirical formula for estimating the relationship between the speed of sound (v) and temperature (T, in degrees Celsius) at standard atmospheric pressure near sea level:

[0071] v(T)=331.3+0.606*T (unit: m / s)

[0072] However, the speed of sound in the actual environment is affected by many factors, and more accurate calculations of the speed of sound may require more complex models, such as the "Urick" speed of sound formula used by the National Oceanic and Atmospheric Administration (NOAA):

[0073] v(T, Salt, P)=1449.4+4.6*T-0.55*T 2 +0.006*T 3 +(1.34*Salt-0.007*Salt 2 )*(1+0.01*P)

[0074] In this formula:

[0075] 1) v is the speed of sound (unit: m / s)

[0076] 2) T is the water temperature (for the speed of sound in water) or the air temperature (for the speed of sound in air), unit: °C

[0077] 3) Salt is the salinity of seawater (for the calculation of the speed of sound in water, the unit is ppt, i.e. grams per thousand); for the calculation of the speed of sound in air, the influence of salinity is generally not considered.

[0078] 4) P is pressure (for the calculation of the speed of sound in water, the unit is: dbar; for the calculation of the speed of sound in air, it can be regarded as a function of the altitude, because the higher the altitude, the lower the air pressure)

[0079] When salinity and pressure are not considered, the above formula can be simplified to: v = 1449.4 + 4.6*T-0.55*T 2 +0.006*T 3 In the present invention, the above formula is used to perform sound velocity correction.

[0080] S52, using the corrected temperature to obtain the distance of the lightning sound wave transmission;

[0081] It should be noted that the formula for the propagation distance of lightning sound waves is:

[0082] d i =v*(t i -t0)

[0083] Among them, d i is the distance from the lightning event to the i-th distributed optical fiber sensing point, v is the propagation speed of the lightning wave in the air, and t i is the time when the sound wave reaches the i-th distributed optical fiber sensing point, and t0 is the actual time when the lightning event occurs.

[0084] Under BeiDou synchronization, all distributed optical fiber sensing points have the same starting time, so t0 can be eliminated in the calculation.

[0085] S53. According to the transmission distance of the lightning sound wave, the precise coordinates of the lightning in the three-dimensional space are obtained by using a nonlinear optimization method.

[0086] As an example, the distance formula between two points in three-dimensional space is:

[0087]

[0088] Where (x, y, z) is the location of the lightning event, (x i ,y i ,z i ) is the lightning acoustic signal sensing position of the i-th sensing point of the distributed optical fiber sensor.

[0089] In order to solve the position vector d of the lightning event i , nonlinear optimization methods can be used to solve this set of equations.

[0090] A commonly used method is a variant of the Least Squares Method, especially when the system of equations is overdetermined (that is, the number of equations is greater than the number of unknowns). In this case, we do not have an exact solution, but can find a solution that is closest to all the equations.

[0091] As an embodiment, step S53 is specifically as follows:

[0092] S531, construct the objective function as follows:

[0093]

[0094] in: i is the number of the DAS host; x is i ,y i 、z i is the three-dimensional space coordinate of the i-th DAS host;

[0095] S532. Optimize the objective function S using a nonlinear optimization method to obtain the three-dimensional spatial coordinates (x, y, z) of the lightning after optimization.

[0096] It should be noted that in order to solve this set of overdetermined nonlinear equations to obtain the lightning event position vector (x, y, z), a nonlinear optimization method, especially the least squares method, is used, with the goal of minimizing the objective function S.

[0097] Nonlinear optimization algorithms such as least squares method, Levenberg-Marquardt algorithm, Newton method, quasi-Newton method, etc. can be used to minimize the objective function S.

[0098] Implementation of nonlinear optimization methods: In practical applications, you can use ready-made mathematical optimization libraries (such as SciPy's optimize module) to implement this step. For example, in Python, you can use the scipy.optimize.least_squares function to solve this problem.

[0099] Solution:

[0100] Call the optimization function and pass in the objective function, initial estimate and other parameters to get the optimized lightning event position (x, y, z). The specific nonlinear algorithm optimization process is as follows:

[0101] 1. Data Collection

[0102] Obtain the arrival timestamps of lightning signals from all distributed optical fiber sensors;

[0103] Record the temperature of the environment and calculate the speed of the thunder wave; v = 1449.4 + 4.6*T - 0.55*T 2 +0.006*T 3 ;

[0104] The distributed acoustic sensor obtains the position coordinates of the optical fiber lightning sensing point (x i ,y i , z i );

[0105] 2. Establish the system of equations

[0106] Calculate d i : The distance from the lightning event to the i-th distributed optical fiber sensing point; d i =v*(t i -t0);

[0107] Constructing nonlinear systems of equations;

[0108] 3. Initialization

[0109] Give an initial guess for the location vector (x, y, z) of the lightning event;

[0110] 4. Nonlinear optimization calculation of accurate lightning location

[0111] Select the least squares method, or the Levenberg-Marquardt algorithm, Newton's method, quasi-Newton's method, etc.; iterative solution: calculate the objective function S and the gradient (sum of squared errors), update iterations (x, y, z), and check the stopping condition (specified accuracy, or an upper limit on the number of iterations, etc.) until the stopping condition is met.

[0112] 5. Output results

[0113] The optimal solution (x, y, z) is obtained, which is the three-dimensional spatial position of the lightning event.

[0114] Finally, the present invention also integrates the calculated locations of all lightning events into a map to form a geographical distribution map of real-time or historical lightning activities for easy observation and analysis.

[0115] See also Figure 2 , Figure 2 4 is a schematic diagram of the working of the hardware device of an embodiment of the present invention, wherein the hardware device specifically comprises: a lightning locating device 401, a processor 402 and a storage medium 403.

[0116] A lightning locating device 401: The lightning locating device 401 implements the lightning locating method.

[0117] Processor 402: The processor 402 loads and executes the instructions and data in the storage medium 403 to implement the lightning locating method.

[0118] Storage medium 403: The storage medium 403 stores instructions and data; the storage medium 403 is used to implement the lightning locating method.

[0119] The beneficial effects of the present invention are:

[0120] (1) High-precision Beidou time synchronization: The Beidou satellite navigation system is used to achieve microsecond-level time synchronization between distributed optical fiber acoustic sensor hosts, ensuring the accuracy of recording the moment when lightning events occur, and providing a solid time reference for positioning calculations.

[0121] (2) Distributed fiber optic sensing network: The DAS network is constructed by utilizing the fiber optic resources in the existing composite overhead cables. Each access point is equipped with a fiber optic acoustic emission sensor host to achieve wide-area, multi-angle, and multi-level lightning acoustic wave monitoring, providing rich multi-point observation data to improve positioning accuracy.

[0122] (3) Direct sensing of lightning sound waves by optical fiber: Optical fiber is used as a direct sensing medium to capture the vibration signal caused by lightning sound waves by detecting the change in light phase. It has the advantages of anti-electromagnetic interference, corrosion resistance, and long-distance transmission. It is especially suitable for lightning monitoring in complex environments.

[0123] (4) Real-time lightning geographic distribution map: Realize real-time visualization of positioning results, generate lightning geographic distribution map, and intuitively display the spatial distribution, which helps to quickly assess lightning disaster risks and provide decision support for emergency response and resource scheduling.

[0124] (5) Deep learning-assisted noise discrimination and positioning algorithm integration: Integrate deep learning models to effectively distinguish lightning acoustic signals from various environmental noises (such as cars, excavators, motors, etc.), enhance the system's anti-interference ability in complex environments, and ensure the accuracy of lightning event detection and positioning.

[0125] (6) It explicitly proposes the use of nonlinear optimization methods (such as the least squares method) for positioning calculations, elaborates on the positioning calculation process (calculating distance using time difference, constructing a nonlinear equation group, and optimizing using the least squares method), and provides a specific implementation path for the positioning algorithm.

[0126] (7) Systematic Solution: Provides a complete, systematic lightning monitoring and positioning solution that covers hardware (distributed fiber optic sensor network, Beidou synchronization system) and software (positioning algorithm, real-time geographic distribution map generation). It is highly integrated and practical, and presents a more comprehensive, systematic and technologically advanced lightning monitoring system compared to existing technologies.

[0127] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A lightning location method, characterized in that: The method comprises the following steps: S1: Lay multiple DAS hosts equipped with Beidou synchronization system in the predetermined area through optical fiber; S2: Use the BeiDou synchronization system to complete the time synchronization of the DAS host; S3: using each DAS host to collect lightning sound waves, and recording the confirmed timestamp when the lightning sound wave signal arrives at each DAS host; S4: The signal data with Beidou synchronization timestamp collected by each DAS host is transmitted to the central processor or data center, and the arrival time difference of the same lightning event signal between different DAS hosts is calculated through the TDOA positioning algorithm; S5: Calculate the precise coordinates of lightning in three-dimensional space based on the time difference to achieve lightning location; Step S5 is specifically as follows: S51, correcting the speed of sound according to the temperature; S52, using the corrected temperature to obtain the distance of the lightning sound wave transmission; S53, according to the distance of the lightning sound wave transmission, using a nonlinear optimization method to obtain the precise coordinates of the lightning in the three-dimensional space; The distance of the lightning sound wave transmission in step S52 is as follows: d i =v*(t i -t0) Among them, d i is the distance from the lightning event to the i-th distributed DAS host; t i is the time when the sound wave reaches the i-th distributed optical fiber sensing point; t0 is the actual time when the lightning event occurs; Step S53 is specifically as follows: S531, construct the objective function as follows: in: i is the number of the DAS host; x is i ,y i 、z i is the three-dimensional space coordinate of the i-th DAS host; S532. Optimize the objective function S using a nonlinear optimization method to obtain the three-dimensional spatial coordinates (x, y, z) of the lightning after optimization.

2. A lightning location method according to claim 1, characterized in that: In step S51, the formula for correcting the sound velocity is as follows: v=1449.4+4.6*T-0.55*T 2 +0.006*T 3 Where T is temperature.

3. A lightning location method according to claim 1, characterized in that: The nonlinear optimization methods include: least square method, Newton method, quasi-Newton method and LM method.

4. A lightning location method according to claim 1, characterized in that: The specific process of using the DAS host to adopt the thunder and lightning sound waves in step S3 is as follows: S31, obtaining the original sound wave data collected by the DAS host, and preprocessing it to obtain preprocessed data; S32, inputting the preprocessed data into a pre-trained lightning recognition model to obtain the probability that the sound wave signal is a lightning signal; S33. When the probability that the sound wave signal is a lightning signal is greater than a preset value, the sound wave signal at this time is used as a vibration signal of a lightning vibration wave.

5. A lightning location method according to claim 4, characterized in that: The process of pre-training the lightning recognition model in step S32 is as follows: S321, collecting lightning signals of different lightning acoustic signals, environmental noise, and mixed environmental noise; S322, performing time-frequency analysis on the signal in step S321 to obtain time-frequency features; S323, combining the time-frequency features with the voiceprint features to construct a training set and a test set; S324: train the lightning recognition model using the training set, and test the lightning recognition model using the test set to obtain a pre-trained lightning recognition model.

6. A storage medium, characterized in that: The storage medium stores instructions and data for implementing a lightning locating method as described in any one of claims 1 to 5.

7. A lightning locating device, characterized in that: include: Processor and storage medium; the processor loads and executes instructions and data in the storage medium to implement a lightning locating method as described in any one of claims 1 to 5.

Images