CN113408805A - Lightning ground flashover identification method, device, equipment and readable storage medium - Google Patents
Lightning ground flashover identification method, device, equipment and readable storage medium Download PDFInfo
- Publication number
- CN113408805A CN113408805A CN202110705947.2A CN202110705947A CN113408805A CN 113408805 A CN113408805 A CN 113408805A CN 202110705947 A CN202110705947 A CN 202110705947A CN 113408805 A CN113408805 A CN 113408805A
- Authority
- CN
- China
- Prior art keywords
- lightning
- waveform
- data
- ground
- identified
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 56
- 238000012549 training Methods 0.000 claims abstract description 22
- 230000000694 effects Effects 0.000 claims abstract description 16
- 238000010801 machine learning Methods 0.000 claims abstract description 15
- 238000012545 processing Methods 0.000 claims description 25
- 238000004590 computer program Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000006243 chemical reaction Methods 0.000 claims description 7
- 238000003384 imaging method Methods 0.000 claims description 4
- 238000007796 conventional method Methods 0.000 abstract description 3
- 238000013527 convolutional neural network Methods 0.000 description 8
- 238000001514 detection method Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 7
- 230000005670 electromagnetic radiation Effects 0.000 description 3
- 238000012216 screening Methods 0.000 description 3
- 238000013461 design Methods 0.000 description 2
- 238000010998 test method Methods 0.000 description 2
- 230000002411 adverse Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004883 computer application Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 238000012850 discrimination method Methods 0.000 description 1
- 230000005674 electromagnetic induction Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 238000000611 regression analysis Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Tourism & Hospitality (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Economics (AREA)
- Biomedical Technology (AREA)
- Development Economics (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Business, Economics & Management (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Marketing (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Educational Administration (AREA)
- Primary Health Care (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention discloses a lightning ground flash identification method, which does not directly adopt a typical fixed peak and trough characteristic identification model to match lightning waveforms, but converts the waveform identification problem into an image classification target problem to realize, trains a built machine learning model according to a large number of lightning ground lightning magnetic wave waveform pictures and non-lightning ground lightning magnetic wave pictures as training samples to obtain a model capable of identifying lightning ground lightning magnetic waves, identifies the integral form of the lightning ground lightning magnetic waves through the model, identifies the waveform characteristics of all positions of electromagnetic waves, has a larger waveform range and more comprehensive characteristics compared with the conventional method in which identification is carried out in a fixed threshold matching mode on key positions such as peaks, troughs and the like of the waveforms, and can improve the identification capability of the lightning electromagnetic waves, thereby improving the identification accuracy of the lightning ground lightning magnetic waves. The invention also discloses a lightning ground flashover recognition device, equipment and a readable storage medium, and the device and the equipment have corresponding technical effects.
Description
Technical Field
The invention relates to the technical field of lightning electromagnetic wave detection and identification, in particular to a lightning ground flashover identification method, device, equipment and readable storage medium.
Background
Thunderstorm weather is one of natural phenomena, and in recent years, disasters caused by thunder and lightning frequently occur and tend to rise rapidly. The lightning detection and early warning become an important technical means for preventing and treating lightning disasters.
The electromagnetic radiation field generated by lightning activity, especially the low frequency/very low frequency band electromagnetic radiation field with concentrated main energy, can propagate hundreds of kilometers or more along the surface of the earth. The existing lightning detection method and apparatus usually detect the electromagnetic induction signal corresponding to the low/very low frequency electromagnetic radiation wave generated by the lightning activity. Lightning is classified into major categories, mainly including cloud flashover and ground flashover, wherein the ground flashover causes more harm. The identification criteria for distinguishing between cloud flashes and ground flashes that are presently disclosed are: a time discrimination method, namely detecting the time from a reference zero point to a peak point of a main peak of an input signal, wherein a cloud lightning detection signal has short rise time; a bipolar test method, i.e. comparing with the first peak, if the peak value of the subsequent opposite polarity peak is larger, it is determined as cloud flash; the other method is a homopolar test method, namely, a subsequent peak with the same polarity as the first peak exists, and the peak value of the subsequent peak is larger, so that the cloud flash is judged. If the input induced signal does not satisfy the above determination, it will be assumed to be a lightning signal. However, even if the above criteria are used in combination, there is still an erroneous determination that the lightning signal is being made.
In summary, how to improve the accuracy of lightning magnetic wave identification of the lightning ground and reduce the probability of wrong determination is a technical problem that needs to be solved urgently by technical personnel in the field at present.
Disclosure of Invention
The invention aims to provide a lightning ground lightning identification method, a lightning ground lightning identification device, lightning ground lightning identification equipment and a readable storage medium, which can improve the lightning ground lightning magnetic wave identification accuracy and reduce the error judgment probability.
In order to solve the technical problems, the invention provides the following technical scheme:
a lightning strike identification method, comprising:
determining electromagnetic wave voltage data to be identified; the voltage data of the electromagnetic waves to be identified comprises low-frequency or very low-frequency band data;
carrying out graphical processing on the voltage data of the electromagnetic waves to be identified to obtain a waveform image;
calling a pre-trained lightning ground lightning magnetic wave recognition model to perform electromagnetic wave form recognition on the waveform image to obtain the lightning ground lightning prediction probability; the lightning magnetic wave identification model is a machine learning model obtained by performing waveform identification training according to a lightning magnetic wave waveform picture and a non-lightning magnetic wave waveform picture;
and generating a lightning ground flashover recognition result according to the lightning ground flashover prediction probability.
Optionally, the step of performing graphical processing on the voltage data of the electromagnetic wave to be identified to obtain a waveform image includes:
dividing the voltage value of each point in the electromagnetic wave voltage data to be identified by the peak voltage value respectively, and taking the quotient value corresponding to each point obtained by calculation as a standard value;
and drawing the waveform by taking the time corresponding to each point as a horizontal axis and the standard value as a vertical axis to generate a waveform image.
Optionally, after the waveform drawing is performed with the time corresponding to each point as a horizontal axis and the standard value as a vertical axis, the method further includes:
and converting the waveform image into a square image with a set value as a side length.
Optionally, the determining the voltage data of the electromagnetic wave to be identified includes:
receiving the voltage value recording data of electromagnetic waves in thunderstorm activities;
determining a peak point of a wave crest in a low-frequency or very low-frequency band range in the electromagnetic wave voltage value record data;
and intercepting data in a front preset range and a rear preset range as the electromagnetic wave voltage data to be identified by taking the peak point of the peak in the electromagnetic wave voltage value recording data as a central point.
A lightning strike identification device comprising:
the data determination unit is used for determining the voltage data of the electromagnetic waves to be identified; the voltage data of the electromagnetic waves to be identified comprises low-frequency or very low-frequency band data;
the imaging processing unit is used for carrying out imaging processing on the voltage data of the electromagnetic waves to be identified to obtain a waveform image;
the form recognition unit is used for calling a pre-trained lightning ground lightning magnetic wave recognition model to perform electromagnetic wave form recognition on the waveform image to obtain the lightning ground lightning prediction probability; the lightning magnetic wave identification model is a machine learning model obtained by performing waveform identification training according to a lightning magnetic wave waveform picture and a non-lightning magnetic wave waveform picture;
and the result generation unit is used for generating a lightning ground flashover recognition result according to the lightning ground flashover prediction probability.
Optionally, the graphic processing unit includes:
the standard value conversion subunit is used for dividing the voltage value of each point in the electromagnetic wave voltage data to be identified by the peak voltage value respectively, and taking the quotient value corresponding to each point obtained by calculation as a standard value;
and the waveform drawing subunit is used for drawing the waveform by taking the time corresponding to each point as a horizontal axis and the standard value as a vertical axis to generate a waveform image.
Optionally, the graphic processing unit further includes: and the shape conversion subunit is connected with the waveform drawing subunit and is used for converting the waveform image into a square image with a set value as side length.
Optionally, the data determination unit includes:
the data receiving subunit is used for receiving the electromagnetic wave voltage value recording data in the thunderstorm activity;
a peak point determining subunit, configured to determine a peak point of a low-frequency or very low-frequency band range in the electromagnetic wave voltage value record data;
and the data intercepting subunit is used for intercepting data in a front preset range and a rear preset range as the electromagnetic wave voltage data to be identified by taking the peak point of the wave crest as a central point in the electromagnetic wave voltage value recording data.
A computer device, comprising:
a memory for storing a computer program;
a processor for implementing the steps of the lightning strike identification method when executing the computer program.
A readable storage medium having stored thereon a computer program which, when being executed by a processor, carries out the steps of the lightning strike identification method as described above.
The method provided by the embodiment of the invention does not directly adopt a typical fixed peak and trough characteristic identification model to match the lightning waveform, but converts the waveform identification problem into the image classification target problem to realize, according to a large number of lightning magnetic wave waveform pictures of lightning grounds and non-lightning magnetic wave pictures of lightning grounds as training samples, training the built machine learning model to finally form a model capable of identifying lightning magnetic waves of the lightning ground, the overall morphology of the lightning magnetic wave of the lightning is identified by the lightning magnetic wave identification model of the lightning, the waveform characteristics of all the positions of the electromagnetic wave are identified, compared with the conventional method in which the waveform characteristics are identified by matching key positions such as wave crests, wave troughs and the like of the waveform with fixed thresholds, the method has the advantages of large waveform range and more characteristics, the lightning electromagnetic wave identification capability can be improved, and therefore the lightning electromagnetic wave identification accuracy of the lightning ground is improved.
Correspondingly, the embodiment of the invention also provides a lightning ground flashover recognition device, equipment and a readable storage medium corresponding to the lightning ground flashover recognition method, which have the technical effects and are not described herein again.
Drawings
In order to more clearly illustrate the embodiments of the present invention or technical solutions in related arts, the drawings used in the description of the embodiments or related arts will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flow chart of an embodiment of a lightning ground flashover identification method according to the present invention;
FIG. 2 is a schematic structural diagram of a CNN model for identifying lightning magnetic waves of lightning ground in the embodiment of the invention;
FIG. 3 is a schematic structural diagram of a lightning and ground lightning identification device according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of a computer device according to an embodiment of the present invention.
Detailed Description
The core of the invention is to provide a lightning ground lightning identification method, which can improve the lightning ground lightning magnetic wave identification accuracy and reduce the error judgment probability.
In order that those skilled in the art will better understand the disclosure, the invention will be described in further detail with reference to the accompanying drawings and specific embodiments. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
A large number of researches of the applicant based on the existing lightning magnetic wave identification method of the lightning ground show that the existing lightning magnetic wave identification method of the lightning ground adopts a typical lightning model to match and identify with detection signals, however, the lightning magnetic wave of the lightning ground is delayed or deformed along the waveform which is easily caused by mountains, rivers, vegetation, high buildings and the like when being propagated on the ground surface, and when the deformed waveform exceeds any characteristic threshold value in the typical lightning model of the lightning ground, the electromagnetic wave can not be identified, so that the lightning ground is missed to be detected.
In view of the above problem, the applicant proposes a lightning ground flashover identification method, which can accurately identify whether the lightning ground flashover is the ground flashover, please refer to fig. 1, where fig. 1 is a flowchart of a lightning ground flashover identification method in an embodiment of the present invention, and the method includes the following steps:
s101, determining electromagnetic wave voltage data to be identified;
the electromagnetic wave voltage data to be identified refers to data to be identified, wherein the electromagnetic wave voltage value change is recorded. Because the characteristics of the lightning ground flashover are mainly concentrated on the low frequency or very low frequency band (30kHz to 100kHz), it needs to be ensured that the voltage data of the electromagnetic wave to be identified needs to contain the low frequency or very low frequency band data, and the embodiment is not limited to whether the voltage data contains other frequency bands.
In order to reduce adverse effects of the waveform data irrelevant to the ground flash identification features on identification workload and identification accuracy as much as possible, and meanwhile, to keep waveform change forms and reduce man-made waveform screening workload, the embodiment provides an implementation manner for determining electromagnetic wave voltage data to be identified, which can directly screen and reject electromagnetic wave voltage values in the recorded thunderstorm activity process, so that other irrelevant interference can be eliminated, and the identification accuracy is improved.
Specifically, the implementation manner of determining the voltage data of the electromagnetic wave to be identified is as follows:
(1) receiving the voltage value recording data of electromagnetic waves in thunderstorm activities;
(2) determining a peak point of a wave crest in a low-frequency or very low-frequency band range in the recorded data of the voltage value of the electromagnetic wave;
(3) and intercepting data in a front preset range and a rear preset range as electromagnetic wave voltage data to be identified by taking a peak point of a peak in the electromagnetic wave voltage value recording data as a central point.
In the mode, the electromagnetic wave voltage value induced by the antenna in the thunderstorm activity process (one thunderstorm process may contain multiple lightning) is completely recorded by directly acquiring the electromagnetic wave waveform recording device, peak points meeting the range of a low frequency/very low frequency band (30 kHz-100 kHz) in the waveform recording are selected, waveforms in a preset range (such as 500 us) before and after each peak point is intercepted, the values can be adjusted according to actual use requirements, specific values are set, and are not limited in the embodiment) are taken as electromagnetic wave voltage data to be identified, the manual workload is reduced through automatic identification and screening of equipment, meanwhile, irrelevant voltage values are removed, waveform change data are selected, interference data can be reduced, and the identification accuracy is improved. In this embodiment, only the implementation manner of determining the voltage data of the electromagnetic wave to be identified is taken as an example, and other implementation processes (for example, manually screening out the waveform change data including the lightning process as the voltage data of the electromagnetic wave to be identified, or directly taking the waveform change data including one or more lightning processes as the voltage data of the electromagnetic wave to be identified, etc.) may refer to the implementation process described in this embodiment, and are not described herein again.
S102, performing graphical processing on the voltage data of the electromagnetic waves to be identified to obtain a waveform image;
in order to improve the identification capability, a typical fixed identification model is not directly adopted to match lightning waveforms, the waveform identification problem is converted into an image classification target problem to be realized, and the voltage data of electromagnetic waves to be identified is subjected to graphical processing, so that the overall identification of the lightning magnetic wave form on the lightning ground is realized on a waveform image, the key positions such as wave crests and wave troughs also comprise the characteristic matching of other positions, the identified characteristics are more comprehensive, and the identification accuracy is improved.
The specific implementation process of performing the graphical processing on the electromagnetic wave voltage data to be identified is not limited in this embodiment, and other data waveform processing manners may be referred to, and are not limited herein. In order to ensure the lightning magnetic wave waveform identification effect of the lightning ground, a graphical processing means is introduced, and the graphical processing means specifically comprises the following steps:
(1) dividing the voltage value of each point in the electromagnetic wave voltage data to be identified by the peak voltage value respectively, and taking the quotient value of each point obtained by calculation as a standard value;
wherein, each point refers to each data point in the electromagnetic wave voltage data to be identified.
Assume the value L of the peak PFThe value corresponding to the arbitrary point x of the waveform is LxThen the peak values are all set to 1 and the value of any point x is set to LF/Lx. By subjecting the voltage data to waveform normalization processing, the fluctuation range can be increased, thereby facilitating identification of waveform characteristics.
It should be noted that, in the graphical processing method described in this embodiment, only one peak in a set of electromagnetic wave voltage data to be identified needs to be ensured, and on the premise that each peak data segment in the data is respectively intercepted in determining the electromagnetic wave voltage data to be identified, that is, only one peak is in the determined electromagnetic wave voltage data to be identified, or the method of performing the graphical processing described in this embodiment may advance to intercept the peak data, which is not limited herein.
(2) And drawing the waveform by taking the time corresponding to each point as a horizontal axis and the standard value as a vertical axis to generate a waveform image.
After the standard values corresponding to the respective points are obtained, the standard values are converted into waveform images with the horizontal axis as time and the vertical axis as standard values.
In addition, in this embodiment, since the waveform image is subjected to the form recognition by calling the machine learning model, in order to ensure uniform recognition of the machine learning model and improve recognition accuracy, the shape and the side length of the obtained waveform image may be further set, for example, the obtained waveform image is uniformly converted into a square, and the side length of the pixel of the picture may be set according to the calculation capability of the model training machine, for example, to 64 × 64 pixel picture. Since different machine learning types have different requirements on image formats, the shape and size adjustment described above is only taken as an example in this embodiment, and other image adjustments may be set according to actual situations.
In this embodiment, only the above-mentioned waveform-pattern conversion implementation process is taken as an example for description, and other implementation manners, such as directly performing waveform drawing on the voltage values corresponding to each point, can refer to the description of this embodiment, and are not described herein again.
S103, calling a pre-trained lightning ground lightning magnetic wave recognition model to perform electromagnetic wave form recognition on the waveform image to obtain the lightning ground lightning prediction probability;
the lightning magnetic wave recognition model called in the embodiment is a machine learning model obtained by performing waveform recognition training according to a lightning magnetic wave waveform picture and a non-lightning magnetic wave waveform picture, the type of machine learning is not limited in the embodiment, for example, the type of machine learning may be a neural network, a decision tree, regression analysis, and the like.
In this embodiment, a model structure of the lightning magnetic wave identification model is not limited, and may be set according to the requirement of the actual image identification precision, where an architecture design of the CNN convolutional neural network is introduced, as shown in fig. 2, a structural diagram of the lightning magnetic wave identification CNN model is shown, where the structural diagram includes a convolutional layer, a pooling layer, a convolutional layer, and 2 full-connection layers, and each layer of parameters refers to fig. 2.
In calling the lightning magnetic wave recognition model to perform electromagnetic wave form recognition on the waveform image, the called lightning magnetic wave recognition model is a model obtained by performing waveform recognition pre-training according to the lightning magnetic wave waveform picture and the non-lightning magnetic wave waveform picture, and the realization process of model training is not limited in the embodiment, and can refer to the realization of the related technology, and is not repeated herein.
And S104, generating a lightning ground flashover recognition result according to the lightning ground flashover prediction probability.
After the waveform image is input into the lightning ground lightning magnetic wave recognition model, the lightning ground lightning magnetic wave recognition model can output the lightning ground lightning prediction probability of the image, such as 65%, the probability indicating that the image is a ground lightning is 65%, a lightning ground lightning recognition result is generated according to the lightning ground lightning prediction probability, the probability can be directly used as the recognition result, and further judgment can be performed according to the probability, such as if the prediction value is greater than 50%, the image is identified as a lightning ground lightning, otherwise, the image is identified as a non-lightning ground lightning, such as if the lightning ground lightning prediction probability is 65%, the lightning ground lightning recognition result is greater than 50%, and the image is the lightning ground lightning recognition result of the lightning ground lightning.
Based on the above description, the lightning ground flashover identification method provided by the embodiment of the invention does not directly adopt a typical fixed peak and trough feature identification model to match the lightning waveform, but converts the waveform identification problem into the image classification target problem to realize, according to a large number of lightning magnetic wave waveform pictures of lightning grounds and non-lightning magnetic wave pictures of lightning grounds as training samples, training the built machine learning model to finally form a model capable of identifying lightning magnetic waves of the lightning ground, the overall morphology of the lightning magnetic wave of the lightning is identified by the lightning magnetic wave identification model of the lightning, the waveform characteristics of all the positions of the electromagnetic wave are identified, compared with the conventional method in which the waveform characteristics are identified by matching key positions such as wave crests, wave troughs and the like of the waveform with fixed thresholds, the method has the advantages of large waveform range and more characteristics, the lightning electromagnetic wave identification capability can be improved, and therefore the lightning electromagnetic wave identification accuracy of the lightning ground is improved.
In the above embodiment, the training process of the lightning magnetic wave identification model in the lightning ground is not limited, and for further understanding, the model training method for the CNN convolutional neural network is provided in this embodiment, and the training processes of other types of machine learning models can refer to the description of this embodiment, and are not described herein again.
(1) An electromagnetic wave waveform recording device is adopted to completely record the voltage value of the electromagnetic wave induced by an antenna in the thunderstorm activity process (one thunderstorm process may contain multiple lightning);
(2) selecting peak points of wave peaks which accord with the range of 30 kHz-100 kHz in the waveform record, intercepting the waveform of 500us before and after each peak point of wave peaks as a sample, and recording the detailed time of the peak points of wave peaks;
(3) each waveform sample is imaged. In order to guarantee the efficiency of training and the recognition effect of lightning magnetic wave waveforms of the lightning ground, the following steps are mainly carried out:
(3.1) normalizing the waveform, assuming the value L of the peak PFThe value corresponding to the arbitrary point x of the waveform is LxThen the peak values are all set to 1 and the value of any point x is set to LF/Lx；
(3.2) converting the waveform into a waveform picture with the horizontal axis as time and the vertical axis as electromagnetic wave induction voltage;
(3.3) uniformly converting the pictures into squares, wherein the side length of pixels of the pictures can be set according to the calculation capacity of a model training machine, and the pictures can be recommended to be set to be 64 x 64 pixels;
(4) comparing data of other lightning ground flashover detection equipment according to the peak time of each waveform sample in the step (2), if three or more lightning ground flashover detection equipment can be matched to acquire a certain peak value and the difference of the peak time is within 0.1ms, marking the sample as a lightning ground flashover sample (which can be marked as 1), otherwise, turning to the step (5);
(5) manually judging whether the waveform sample is a lightning ground lightning sample, if so, marking the waveform sample as the lightning ground lightning sample (which can be marked as 1), and otherwise, marking the waveform sample as a non-lightning ground lightning sample (which can be marked as 0);
(6) and inputting the marked waveform sample into a pre-established CNN convolutional neural network for training, calculating the target identification accuracy of the sample test set, and stopping training if the accuracy is more than 95%.
Through actual measurement, the model obtained through the training in the above manner better identifies the waveform form of the lightning magnetic wave of the lightning ground, and is closer to the manual judgment result.
Corresponding to the above method embodiment, the embodiment of the present invention further provides a lightning ground flashover recognition apparatus, and the lightning ground flashover recognition apparatus described below and the lightning ground flashover recognition method described above may be referred to in correspondence with each other.
Referring to fig. 3, the apparatus includes the following modules:
the data determination unit 110 is mainly used for determining the voltage data of the electromagnetic waves to be identified; the voltage data of the electromagnetic waves to be identified comprises low-frequency or very low-frequency band data;
the graphical processing unit 120 is mainly used for performing graphical processing on the voltage data of the electromagnetic waves to be identified to obtain a waveform image;
the form recognition unit 130 is mainly used for calling a pre-trained lightning ground lightning magnetic wave recognition model to perform electromagnetic wave form recognition on the waveform image to obtain the lightning ground lightning prediction probability; the lightning magnetic wave identification model is a machine learning model obtained by performing waveform identification training according to a lightning magnetic wave waveform picture and a non-lightning magnetic wave waveform picture;
the result generating unit 140 is mainly used for generating the lightning strike identification result according to the lightning strike prediction probability.
In one embodiment of the present invention, a graphic processing unit includes:
the standard value conversion subunit is used for dividing the voltage value of each point in the electromagnetic wave voltage data to be identified by the peak voltage value respectively, and taking the quotient value corresponding to each point obtained by calculation as a standard value;
and the waveform drawing subunit is used for drawing the waveform by taking the time corresponding to each point as a horizontal axis and the standard value as a vertical axis to generate a waveform image.
In an embodiment of the present invention, the graphic processing unit further includes: and the shape conversion subunit is connected with the waveform drawing subunit and is used for converting the waveform image into a square image with the side length of a set value.
In one embodiment of the present invention, the data determination unit includes:
the data receiving subunit is used for receiving the electromagnetic wave voltage value recording data in the thunderstorm activity;
the peak point determining subunit is used for determining a peak point of a low-frequency or very low-frequency range in the electromagnetic wave voltage value recording data;
and the data intercepting subunit is used for intercepting data in a front preset range and a rear preset range as the electromagnetic wave voltage data to be identified by taking the peak point of the wave crest as a central point in the electromagnetic wave voltage value recording data.
Corresponding to the above method embodiment, the embodiment of the present invention further provides a computer device, and a computer device described below and a lightning ground flashover identification method described above may be referred to correspondingly.
The computer device includes:
a memory for storing a computer program;
a processor for implementing the steps of the lightning strike identification method of the above-described method embodiments when executing the computer program.
Specifically, referring to fig. 4, a specific structural diagram of a computer device provided in this embodiment is a schematic diagram, where the computer device may generate a relatively large difference due to different configurations or performances, and may include one or more processors (CPUs) 322 (e.g., one or more processors) and a memory 332, where the memory 332 stores one or more computer applications 342 or data 344. Memory 332 may be, among other things, transient or persistent storage. The program stored in memory 332 may include one or more modules (not shown), each of which may include a sequence of instructions operating on a data processing device. Still further, the central processor 322 may be configured to communicate with the memory 332 to execute a series of instruction operations in the memory 332 on the computer device 301.
The computer device 301 may also include one or more power supplies 326, one or more wired or wireless network interfaces 350, one or more input-output interfaces 358, and/or one or more operating systems 341.
The steps in the lightning ground flashover recognition method described above may be implemented by the structure of the computer device provided in the present embodiment.
Corresponding to the above method embodiment, the embodiment of the present invention further provides a readable storage medium, and a readable storage medium described below and a lightning ground flashover identification method described above may be referred to in correspondence.
A readable storage medium, on which a computer program is stored, which computer program, when being executed by a processor, carries out the steps of the lightning strike identification method of the above-mentioned method embodiments.
The readable storage medium may be a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and various other readable storage media capable of storing program codes.
Those of skill would further appreciate that the various illustrative components and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative components and steps have been described above generally in terms of their functionality in order to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
Claims (10)
1. A lightning ground flash identification method is characterized by comprising the following steps:
determining electromagnetic wave voltage data to be identified; the voltage data of the electromagnetic waves to be identified comprises low-frequency or very low-frequency band data;
carrying out graphical processing on the voltage data of the electromagnetic waves to be identified to obtain a waveform image;
calling a pre-trained lightning ground lightning magnetic wave recognition model to perform electromagnetic wave form recognition on the waveform image to obtain the lightning ground lightning prediction probability; the lightning magnetic wave identification model is a machine learning model obtained by performing waveform identification training according to a lightning magnetic wave waveform picture and a non-lightning magnetic wave waveform picture;
and generating a lightning ground flashover recognition result according to the lightning ground flashover prediction probability.
2. The lightning ground flash identification method of claim 1, wherein the step of performing graphical processing on the voltage data of the electromagnetic waves to be identified to obtain a waveform image comprises the steps of:
dividing the voltage value of each point in the electromagnetic wave voltage data to be identified by the peak voltage value respectively, and taking the quotient value corresponding to each point obtained by calculation as a standard value;
and drawing the waveform by taking the time corresponding to each point as a horizontal axis and the standard value as a vertical axis to generate a waveform image.
3. The lightning strike identification method according to claim 2, wherein after the waveform drawing is performed with the time corresponding to each point as a horizontal axis and the standard value as a vertical axis, the method further comprises:
and converting the waveform image into a square image with a set value as a side length.
4. The lightning strike identification method of claim 1, wherein the determining electromagnetic wave voltage data to be identified comprises:
receiving the voltage value recording data of electromagnetic waves in thunderstorm activities;
determining a peak point of a wave crest in a low-frequency or very low-frequency band range in the electromagnetic wave voltage value record data;
and intercepting data in a front preset range and a rear preset range as the electromagnetic wave voltage data to be identified by taking the peak point of the peak in the electromagnetic wave voltage value recording data as a central point.
5. A lightning strike identification device, comprising:
the data determination unit is used for determining the voltage data of the electromagnetic waves to be identified; the voltage data of the electromagnetic waves to be identified comprises low-frequency or very low-frequency band data;
the imaging processing unit is used for carrying out imaging processing on the voltage data of the electromagnetic waves to be identified to obtain a waveform image;
the form recognition unit is used for calling a pre-trained lightning ground lightning magnetic wave recognition model to perform electromagnetic wave form recognition on the waveform image to obtain the lightning ground lightning prediction probability; the lightning magnetic wave identification model is a machine learning model obtained by performing waveform identification training according to a lightning magnetic wave waveform picture and a non-lightning magnetic wave waveform picture;
and the result generation unit is used for generating a lightning ground flashover recognition result according to the lightning ground flashover prediction probability.
6. The lightning strike identification device of claim 5, wherein the graphical processing unit comprises:
the standard value conversion subunit is used for dividing the voltage value of each point in the electromagnetic wave voltage data to be identified by the peak voltage value respectively, and taking the quotient value corresponding to each point obtained by calculation as a standard value;
and the waveform drawing subunit is used for drawing the waveform by taking the time corresponding to each point as a horizontal axis and the standard value as a vertical axis to generate a waveform image.
7. The lightning strike identification device of claim 6, wherein the graphical processing unit further comprises: and the shape conversion subunit is connected with the waveform drawing subunit and is used for converting the waveform image into a square image with a set value as side length.
8. The lightning strike identification device of claim 5, wherein the data determination unit comprises:
the data receiving subunit is used for receiving the electromagnetic wave voltage value recording data in the thunderstorm activity;
a peak point determining subunit, configured to determine a peak point of a low-frequency or very low-frequency band range in the electromagnetic wave voltage value record data;
and the data intercepting subunit is used for intercepting data in a front preset range and a rear preset range as the electromagnetic wave voltage data to be identified by taking the peak point of the wave crest as a central point in the electromagnetic wave voltage value recording data.
9. A computer device, comprising:
a memory for storing a computer program;
processor for implementing the steps of the lightning strike identification method according to any one of claims 1 to 4 when executing the computer program.
10. A readable storage medium, characterized in that the readable storage medium has stored thereon a computer program which, when being executed by a processor, carries out the steps of the lightning strike identification method according to any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110705947.2A CN113408805A (en) | 2021-06-24 | 2021-06-24 | Lightning ground flashover identification method, device, equipment and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110705947.2A CN113408805A (en) | 2021-06-24 | 2021-06-24 | Lightning ground flashover identification method, device, equipment and readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113408805A true CN113408805A (en) | 2021-09-17 |
Family
ID=77683064
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110705947.2A Pending CN113408805A (en) | 2021-06-24 | 2021-06-24 | Lightning ground flashover identification method, device, equipment and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113408805A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115204222A (en) * | 2022-06-30 | 2022-10-18 | 宁波麦思捷科技有限公司 | Thunder and lightning prediction method, device and equipment based on synchronous compression wavelet transform |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101539599A (en) * | 2009-04-14 | 2009-09-23 | 国网电力科学研究院 | Digital lightning detection method and device thereof |
| CN109190667A (en) * | 2018-07-31 | 2019-01-11 | 中国电子科技集团公司第二十九研究所 | A kind of Object Threat Evaluation method, model and model building method based on electronic reconnaissance signal |
| CN110135654A (en) * | 2019-05-24 | 2019-08-16 | 北京百度网讯科技有限公司 | Method and apparatus for predicting strong convective weather |
| CN110808068A (en) * | 2019-11-11 | 2020-02-18 | 北京软通智城科技有限公司 | Sound detection method, device, equipment and storage medium |
| CN111881723A (en) * | 2020-06-09 | 2020-11-03 | 南瑞集团有限公司 | Automatic identification method and system for lightning ground flashback waveform and electronic equipment |
| CN112727710A (en) * | 2020-12-15 | 2021-04-30 | 北京天泽智云科技有限公司 | Wind field thunderbolt density statistical method and system based on audio signals |
| CN112986698A (en) * | 2020-10-22 | 2021-06-18 | 南京信息工程大学 | Three-dimensional lightning positioning method |
-
2021
- 2021-06-24 CN CN202110705947.2A patent/CN113408805A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101539599A (en) * | 2009-04-14 | 2009-09-23 | 国网电力科学研究院 | Digital lightning detection method and device thereof |
| CN109190667A (en) * | 2018-07-31 | 2019-01-11 | 中国电子科技集团公司第二十九研究所 | A kind of Object Threat Evaluation method, model and model building method based on electronic reconnaissance signal |
| CN110135654A (en) * | 2019-05-24 | 2019-08-16 | 北京百度网讯科技有限公司 | Method and apparatus for predicting strong convective weather |
| CN110808068A (en) * | 2019-11-11 | 2020-02-18 | 北京软通智城科技有限公司 | Sound detection method, device, equipment and storage medium |
| CN111881723A (en) * | 2020-06-09 | 2020-11-03 | 南瑞集团有限公司 | Automatic identification method and system for lightning ground flashback waveform and electronic equipment |
| CN112986698A (en) * | 2020-10-22 | 2021-06-18 | 南京信息工程大学 | Three-dimensional lightning positioning method |
| CN112727710A (en) * | 2020-12-15 | 2021-04-30 | 北京天泽智云科技有限公司 | Wind field thunderbolt density statistical method and system based on audio signals |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115204222A (en) * | 2022-06-30 | 2022-10-18 | 宁波麦思捷科技有限公司 | Thunder and lightning prediction method, device and equipment based on synchronous compression wavelet transform |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10269138B2 (en) | UAV inspection method for power line based on human visual system | |
| CN109782274B (en) | Water damage identification method based on time-frequency statistical characteristics of ground penetrating radar signals | |
| CN108489996B (en) | Insulator defect detection method and system and terminal equipment | |
| CN106778734A (en) | A kind of insulator based on rarefaction representation falls to go here and there defect inspection method | |
| CN109376740A (en) | A kind of water gauge reading detection method based on video | |
| WO2018000252A1 (en) | Oceanic background modelling and restraining method and system for high-resolution remote sensing oceanic image | |
| CN106156758B (en) | A kind of tidal saltmarsh method in SAR seashore image | |
| CN110705619B (en) | Mist concentration grade discriminating method and device | |
| CN115841488B (en) | PCB hole inspection method based on computer vision | |
| CN116776279A (en) | Multi-mode data collaborative power transmission line flash explosion early warning abnormal target detection method | |
| CN110321890B (en) | Digital instrument identification method of power inspection robot | |
| CN113408805A (en) | Lightning ground flashover identification method, device, equipment and readable storage medium | |
| CN114694130A (en) | Method and device for detecting telegraph poles and pole numbers along railway based on deep learning | |
| CN113506290A (en) | Method and device for detecting defects of line insulator | |
| CN117853942A (en) | Cloud and fog identification method, cloud and fog identification device and cloud and fog identification system | |
| CN117197700A (en) | Intelligent unmanned inspection contact net defect identification system | |
| CN113052234A (en) | Jade classification method based on image features and deep learning technology | |
| CN116403085A (en) | Multi-pointer instrument identification method and device for transformer substation based on multitasking | |
| CN113487550B (en) | Target detection method and device based on improved activation function | |
| CN113643234B (en) | Composite insulator damage detection method, terminal equipment and readable storage medium | |
| CN109886133B (en) | Ship detection method and system based on optical remote sensing image | |
| CN112964938B (en) | Lightning single-station positioning method, device and system based on artificial intelligence | |
| CN110781758A (en) | Dynamic video monitoring method and device for abnormal pantograph structure | |
| CN114882389A (en) | Helmet wearing detection method based on extrusion excitation residual error network improved YOLOv3 | |
| CN115294141B (en) | Deep sea fishing net detection method based on sonar image |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210917 |