CN113269135A - Satellite transponder recognition model and training method and using method thereof - Google Patents
Satellite transponder recognition model and training method and using method thereof Download PDFInfo
- Publication number
- CN113269135A CN113269135A CN202110670593.2A CN202110670593A CN113269135A CN 113269135 A CN113269135 A CN 113269135A CN 202110670593 A CN202110670593 A CN 202110670593A CN 113269135 A CN113269135 A CN 113269135A
- Authority
- CN
- China
- Prior art keywords
- domain waveform
- satellite transponder
- waveform curve
- data
- signal
- 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
- 238000012549 training Methods 0.000 title claims abstract description 56
- 238000000034 method Methods 0.000 title claims abstract description 48
- 238000013528 artificial neural network Methods 0.000 claims abstract description 20
- 238000007781 pre-processing Methods 0.000 claims abstract description 17
- 238000012545 processing Methods 0.000 claims abstract description 6
- 238000013507 mapping Methods 0.000 claims abstract description 5
- 238000011282 treatment Methods 0.000 claims description 4
- 230000004913 activation Effects 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 238000012544 monitoring process Methods 0.000 description 6
- 238000001228 spectrum Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 4
- 238000004140 cleaning Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000000007 visual effect Effects 0.000 description 3
- 238000000605 extraction Methods 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000001154 acute effect Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000010287 polarization Effects 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
Abstract
The application provides a training method of a satellite transponder recognition model, which comprises the following steps: constructing a satellite transponder identification model by using an artificial neural network; constructing a training data set; and training the satellite transponder identification model by using the constructed training data set. The step of constructing the training data set comprises a data classification step and a data preprocessing step. The data classification step includes: establishing a mapping relationship between each signal spectrogram and a corresponding satellite transponder name tag aiming at a plurality of signal spectrograms acquired from a receiver, wherein the signal spectrograms comprise background and time domain waveform curves. The data preprocessing step comprises: and extracting a time domain waveform curve from the signal spectrogram, and processing two-dimensional data of the extracted time domain waveform curve.
Description
Technical Field
The invention relates to an automatic identification and classification auxiliary technology for satellite transponders in human-on-loop, in particular to a satellite transponder identification model and a training method thereof, which are suitable for application scenes of a plurality of satellite transponders, large data quantity of received signal spectrograms and low manual classification and identification efficiency.
Background
At present, most of the conventional satellite transponder identification methods are based on the existing expert information such as public transponder configurations (satellite beacons, transponder initial frequency bands, polarization modes, satellite orbit positions), historical monitoring spectrograms in the past year, and the like, and the identification operation is completed in a mode of manual comparison and characteristic matching so as to identify the satellite transponder of which satellite the signal received by the ground receiver comes from. Therefore, the traditional satellite transponder identification method seriously depends on the existing data such as public transponder configuration, past year historical monitoring spectrograms and the like, but cannot identify the signal spectrogram which lacks prior knowledge, and the identification accuracy depends on the experience and level of workers, so that the method has strong artificial subjectivity, lacks timeliness, has high omission possibility and consumes a large amount of manpower and material resources. And this problem is particularly acute in the case of multiple satellites. Due to the fact that the orbit change and the common rail exist in the operation process of a plurality of satellites, even the satellite from which the signal received by the ground receiver comes cannot be identified manually according to the prior knowledge of the configuration of the satellite transponder, the reference signal of the existing satellite and the like under the situation.
Disclosure of Invention
At least one object of the present application is to provide a satellite transponder recognition model and a training method and a using method thereof, by which a satellite transponder of a satellite from which a received signal is coming can be automatically recognized. The identification operation can be carried out continuously for 24 hours all day long, so that auxiliary identification information is provided for monitoring personnel, the monitoring personnel are helped to quickly find and identify targets, and the workload of the monitoring personnel for identifying and classifying the satellite transponder from a mass signal spectrogram is reduced.
The application provides a training method of a satellite transponder recognition model, which comprises the following steps:
s01: constructing a satellite transponder identification model by using an artificial neural network;
s02: constructing a training data set;
s03: training the satellite transponder recognition model by using the constructed training data set,
wherein the step S02 includes a data classification step and a data preprocessing step,
the data classification step includes: establishing a mapping relation between each signal spectrogram and a corresponding satellite transponder name label aiming at a plurality of signal spectrograms obtained from a receiver, wherein the signal spectrograms comprise background and time domain waveform curves;
the data preprocessing step comprises: extracting time domain waveform curve from the signal spectrogram, and processing two-dimensional data ((x) of the extracted time domain waveform curve1,y1),(x2,y2),…(xn,yn) ) the following treatments were performed:
xi=(xi-xmin)/(xmax-xmin)
yi=(yi-ymin)/(ymax-ymin);
wherein x isiIs the x-axis coordinate, x, of the ith point of the time-domain waveform curveminIs the minimum x-axis coordinate, x, of the time-domain waveform curvemaxIs the maximum x-axis coordinate of the time domain waveform curve,
wherein, yiIs the y-axis coordinate, y, of the ith point of the time-domain waveform curveminIs the minimum y-axis coordinate, y, of the time domain waveform curvemaxIs the maximum y-axis coordinate of the time domain waveform curve,
And the two-dimensional data of each signal spectrogram processed by the data preprocessing step and the satellite transponder name label corresponding to the signal spectrogram form training data, and a training data set is constructed by utilizing a plurality of training data.
In at least one embodiment according to the present application, the step S02 further includes a data cleansing step of cleansing the original data received from the receiver according to the a priori information, and cleansing the data with the error information.
In at least one embodiment according to the present application, in the data preprocessing step, the background and the time-domain waveform curve are distinguished by colors to extract the time-domain waveform curve from the signal spectrogram.
In at least one embodiment according to the present application, wherein in the step S01, the satellite transponder identification model is constructed by using an artificial neural network based on PyTorch framework construction.
In at least one embodiment according to the present application, the satellite transponder identification model is composed of three layers of fully-connected neural networks, an activation function between each layer of fully-connected neural networks is leak _ refill, a last layer is softmax, and a loss function is mselos.
In at least one embodiment consistent with the present application, wherein the signal spectrogram is a time domain oscillogram.
In at least one embodiment according to the present application, wherein the input of the satellite transponder identification model is a training data set, and the output of the satellite transponder identification model is a name tag of the satellite transponder to which the signal belongs.
The application also provides a satellite transponder identification model which is trained by the method.
The application also provides a using method of the satellite transponder identification model, and the identification is carried out by using the satellite transponder identification model.
In at least one embodiment according to the present application, the above-mentioned method of use includes:
s11: acquiring a signal from a receiver, and drawing a signal spectrogram according to original data of the signal;
s12: extracting time domain waveform curve in the signal spectrogram, and storing the time domain waveform curve as two-dimensional data ((x)1,y1),(x2,y2),…(xn,yn));
S13: two-dimensional data ((x) of the extracted time-domain waveform curve1,y1),(x2,y2),…(xn,yn) ) the following treatments were performed:
xi=(xi-xmin)/(xmax-xmin)
yi=(yi-ymin)/(ymax-ymin)
wherein x isiIs the x-axis coordinate, x, of the ith point of the time-domain waveform curveminIs the minimum x-axis coordinate, x, of the time-domain waveform curvemaxThe maximum x-axis coordinate of the time domain waveform curve.
Wherein, yiIs the y-axis coordinate, y, of the ith point of the time-domain waveform curveminIs the minimum y-axis coordinate, y, of the time domain waveform curvemaxThe maximum y-axis coordinate of the time domain waveform curve;
s14: and inputting the two-dimensional data obtained in the step S13 into a satellite transponder identification model, and outputting the name label of the satellite transponder to which the signal spectrogram belongs.
By the satellite transponder identification model provided by the application, the satellite transponder of which satellite the received signal comes from can be automatically identified. According to the technical scheme, the satellite transponder identification problem is converted into the image classification problem, so that the complex multi-parameter problem is visual and simple.
According to the training method for the satellite transponder identification model, a high-quality training data set is obtained by adopting a specific data preprocessing method, so that the high-quality model can be trained, the data volume required to be processed by the model is small, the convergence speed of the model is high, and the identification accuracy of the satellite transponder identification model can be improved.
Drawings
The above features, technical features, advantages and modes of realisation of the present application will be further described in the following detailed description of preferred embodiments in a clearly understandable manner, in conjunction with the accompanying drawings. The drawings are only for purposes of illustrating and explaining the present application and are not to be construed as limiting the scope of the present application. Wherein:
FIG. 1 illustrates a satellite transponder identification model training method according to one embodiment of the present application;
fig. 2 illustrates a satellite transponder identification model using method according to an embodiment of the present application.
Detailed Description
In order to more clearly understand the technical features, objects, and effects of the present application, embodiments of the present application will now be described with reference to the accompanying drawings.
In order to identify the satellite transponder of which satellite the received signal comes from, the prior art mainly adopts a manual comparison and feature matching manner to identify a signal spectrogram. This recognition method is heavily dependent on the experience and level of the monitoring personnel, and the recognition efficiency and accuracy are low.
In order to improve the identification efficiency and the identification accuracy of a satellite transponder to which a signal belongs, the application provides a neural network-based satellite transponder identification model, a training method and a using method thereof. In the technical scheme of the application, an artificial neural network is used for constructing (for example, constructing based on a PyTorch framework) a satellite transponder identification model based on a time domain oscillogram of a received signal, and a training data set constructed by a satellite transponder data set is used for training the satellite transponder identification model. By using the trained satellite transponder identification model, the time domain oscillogram of the received signal can be classified according to which satellite the time domain oscillogram belongs to, so that which satellite the signal comes from can be identified.
Therefore, in the technical scheme provided by the application, the satellite transponder identification problem is converted into the image classification problem, so that the complex multi-parameter problem is visual and simple. The technical scheme provided by the application not only can obviously improve the identification efficiency of the satellite transponder and reduce the omission factor, but also can provide soft-decision pre-classification auxiliary information for the artificial accurate identification of the satellite transponder to which the ground receiving signal belongs. The satellite transponder automatic identification auxiliary technology can relieve the pressure of observers, reduce misjudgment caused by long-time observation fatigue of the observers, and provide an auxiliary effect for subsequent manual accurate identification and judgment.
Ideally, each type of device (e.g., each different satellite transponder) emits a signal having a fine signature that is uniquely tied to the manufacturing imperfections of the device, the characteristics of the various components of the device. By extracting and analyzing the characteristics of the received signals, the transmitting equipment corresponding to the signals can be reversely deduced, so that the mapping relation between the signals and the signal transmitting source equipment is obtained. In the prior art, a signal spectrogram is mainly identified by means of manual comparison and characteristic matching. The deep artificial neural network adopted by the application does not need to design and extract features artificially, and the neural network can extract high-dimensional abstract features which are enough to describe a certain specific feature mode according to original input, so that the satellite transponder of which satellite the received signal comes from can be identified.
The training method for the satellite transponder recognition model provided by the embodiment of the invention comprises the following steps:
s01: method for constructing satellite transponder recognition model by utilizing artificial neural network
An artificial neural network based satellite transponder identification model is constructed based on, for example, a pytorre framework. The trunk of the satellite transponder identification model consists of three layers of fully-connected neural networks, the activation function between each layer of the network is leakage _ refill, and the last layer is a softmax layer. The loss function is mselos. The input of the satellite transponder identification model is a training data set, and the output of the satellite transponder identification model is a name tag of a satellite transponder to which the signal belongs.
S02: constructing a training data set
This step S02 includes data cleansing, data classification, and data preprocessing.
Deep learning is a data-driven representation learning method, and the performance of a neural network model depends on the quality of a training data set, so that the construction of a high-quality training data set is very important. Through the steps of data cleaning, data classification and data preprocessing, a high-quality training data set can be obtained.
Wherein, the data cleaning step comprises: according to prior information (existing satellite transponder configuration, reference signals of existing satellites and the like), data cleaning is carried out on original data (namely a plurality of signal spectrograms, the horizontal axis of the signal spectrograms is frequency, and the vertical axis of the signal spectrograms is amplitude) received from a receiver, data with error information is cleaned, and clean data are obtained.
The step of data classification includes: and effectively classifying the clean data according to the prior information, wherein the classification is based on a typical signal spectrogram pattern corresponding to each satellite transponder. After the data classification step, a mapping relation between each signal spectrogram and a satellite transponder corresponding to the signal spectrogram is established.
The data preprocessing step comprises the following steps: 1) removing redundant information which is useless for improving the performance of the model from a signal spectrogram; 2) and calibrating the frequency spectrum waveform. These two steps will be described in detail below with reference to specific examples.
1) Removing redundant information useless for improving model performance in signal spectrogram
As described above, in the technical solution provided by the present application, the satellite transponder identification problem is converted into an image classification problem. In the prior art, the processing method aiming at the image classification problem or the image identification problem is to analyze and calculate the whole area of an image, so that the problems of high computing resource consumption and long processing time exist.
However, for the satellite transponder identification problem addressed by the present application, the important information in the signal spectrogram (i.e., the time domain waveform map) is not distributed over the entire area of the image. The signal spectrum of the raw data received from the receiver is an image of: which includes a background (e.g., white) and a time-domain waveform curve (e.g., red) that is a different color than the background. The background is redundant information which is useless for improving the performance of the satellite transponder identification model, and if the background data is introduced into the training data in the training data set, the training of the model is interfered and influenced. In fact, important information in a signal spectrogram is sparse, and only a time-domain waveform curve in the signal spectrogram is useful information.
In order to improve the quality of training data, in the technical scheme provided by the application, in the step of data preprocessing, time domain waveform curve extraction is carried out on a signal spectrogram. And removing the background of the signal spectrogram, and only extracting a time domain waveform curve in the signal spectrogram to be used as training data.In particular, useless backgrounds and useful time domain waveform profiles can be distinguished by color. Then, the extracted time domain waveform curve is saved as two-dimensional data ((x)1,y1),(x2,y2),…(xn,yn))。
By taking such a step of removing redundant information, the following technical effects can be achieved: 1. the quality of the training data can be improved. High quality training data can train high quality models. Compared with background information which is useless for identification and classification, the time domain waveform curve (overall trend, number of wave crests, relative size of peak value, relative position relation of wave crests and the like) data is the target to be analyzed and processed. The neural network only carries out learning modeling on data of the time domain waveform curve, and is easier to learn essential characteristics only strongly related to the time domain waveform curve; 2. through the extraction of time domain waveform curve data, the data volume of the model needing to be processed can be reduced by times, only the concerned high-value data waveform is learned, and the convergence speed of the model is high.
2) Spectral waveform calibration
Data transmission abnormity can be caused by network jamming of a signal transceiving data link, and noise, waveform peak relative position and waveform envelope deviation can be introduced by waveform distortion caused by artificial factors. Conventional data processing methods do not take these factors into account. As a result, different data sample instances may be very different for the same class label, and data sample instances belonging to different class labels may be rather similar. This can lead to failure of neural network classifier training to converge, which in turn leads to failure of the classification task.
In order to solve the problem, the technical scheme of the application adopts frequency spectrum waveform calibration, and processes the two-dimensional data of the extracted time domain waveform curve to eliminate the influence of external factors, so that the identification accuracy of the satellite transponder identification model is improved.
Specifically, the following processing is performed on the two-dimensional data of the extracted time-domain waveform curve:
xi=(xi-xmin)/(xmax-xmin)
yi=(yi-ymin)/(ymax-ymin)
wherein x isiIs the x-axis coordinate, x, of the ith point of the time-domain waveform curveminIs the minimum x-axis coordinate, x, of the time-domain waveform curvemaxThe maximum x-axis coordinate of the time domain waveform curve.
Wherein, yiIs the y-axis coordinate, y, of the ith point of the time-domain waveform curve minIs the minimum y-axis coordinate, y, of the time domain waveform curvemaxThe maximum y-axis coordinate of the time domain waveform curve.
The coordinates of the processed data are limited between (0, 1), the overall trend of the waveform data is restored, and the relation between the relative position and the relative peak value is restored. By this way of preprocessing, different sample instances under the same label will have similar waveforms.
The tag information of the data can be encoded by adopting a one-hot encoding mode. This can solve the problem of the classifier not handling the attribute data well, while the features can be extended to a certain extent.
S03: training satellite transponder recognition models using training data sets
After step S02, a training data set including high-quality training data is obtained. The satellite transponder identification model obtained in step S01 is trained using the training data set, and thus a neural network satellite transponder identification model with excellent performance can be obtained.
According to another embodiment of the present application, a satellite transponder identification model is provided, which is trained by the training method of the satellite transponder identification model provided by the present application.
According to another embodiment of the application, a method for using the satellite transponder identification model is also provided. The using method comprises the following steps:
S11: acquiring a signal from a receiver, and drawing a signal spectrogram according to original data of the signal;
s12: extracting a time domain waveform curve in a signal spectrogram, and storing the time domain waveform curve as two-dimensional data((x1,y1),(x2,y2),…(xn,yn));
S13: performing spectrum waveform calibration, and performing the following processing on the two-dimensional data of the extracted time domain waveform curve:
xi=(xi-xmin)/(xmax-xmin)
yi=(yi-ymin)/(ymax-ymin)
wherein x isiIs the x-axis coordinate, x, of the ith point of the time-domain waveform curveminIs the minimum x-axis coordinate, x, of the time-domain waveform curvemaxThe maximum x-axis coordinate of the time domain waveform curve.
Wherein, yiIs the y-axis coordinate, y, of the ith point of the time-domain waveform curveminIs the minimum y-axis coordinate, y, of the time domain waveform curvemaxThe maximum y-axis coordinate of the time domain waveform curve;
s14: inputting the two-dimensional data subjected to the spectrum waveform calibration into a satellite transponder identification model, and outputting a name tag of a satellite transponder to which the signal spectrogram belongs so as to identify the satellite transponder of which satellite the signal spectrogram comes from.
In order to test the recognition efficiency and recognition accuracy of the satellite transponder recognition model provided according to the present application, the inventors tested the satellite transponder recognition model obtained according to the training method of the present application. According to yet another embodiment of the present application, a satellite transponder identification model is trained on a training data set containing 13 satellite transponder data according to the method described above including steps S01-S03. Through tests, the classification accuracy reaches 96%, the prediction classification accuracy is high, and high-quality auxiliary information can be provided for subsequent manual accurate judgment and rechecking. Therefore, in the task of automatically identifying the satellite transponder, the trained model can process a signal spectrogram obtained from a ground receiver in real time and give a prediction classification result. The automatic identification auxiliary technology has the capacity of continuous 24-hour working all-weather and uninterrupted.
In summary, the application provides a satellite transponder identification model based on a neural network, and a training method and a using method thereof. Through the satellite transponder identification model, the satellite transponder of which satellite the received signal comes from can be automatically identified. According to the technical scheme, the satellite transponder identification problem is converted into the image classification problem, so that the complex multi-parameter problem is visual and simple.
According to the training method for the satellite transponder identification model, a high-quality training data set is obtained by adopting a specific data preprocessing method, so that the high-quality model can be trained, the data volume required to be processed by the model is small, the convergence speed of the model is high, and the identification accuracy of the satellite transponder identification model can be improved.
It should be understood that although the present description has been described in terms of various embodiments, not every embodiment includes only a single embodiment, and such description is for clarity purposes only, and those skilled in the art will recognize that the embodiments described herein may be combined as suitable to form other embodiments, as will be appreciated by those skilled in the art.
The above description is only illustrative of the present invention and is not intended to limit the scope of the present invention. Any equivalent alterations, modifications and combinations that may be made by those skilled in the art without departing from the spirit and principles of this application shall fall within the scope of this application.
Claims (10)
1. A training method of a satellite transponder recognition model comprises the following steps:
s01: constructing a satellite transponder identification model by using an artificial neural network;
s02: constructing a training data set;
s03: training the satellite transponder recognition model by using the constructed training data set,
wherein the step S02 includes a data classification step and a data preprocessing step,
the data classification step includes: establishing a mapping relation between each signal spectrogram and a corresponding satellite transponder name label aiming at a plurality of signal spectrograms obtained from a receiver, wherein the signal spectrograms comprise background and time domain waveform curves;
the data preprocessing step comprises: extracting time domain waveform curve from the signal spectrogram, and processing two-dimensional data ((x) of the extracted time domain waveform curve1,y1),(x2,y2),…(xn,yn) ) the following treatments were performed:
xi=(xi-xmin)/(xmax-xmin)
yi=(yi-ymin)/(ymax-ymin);
wherein x isiIs the x-axis coordinate, x, of the ith point of the time-domain waveform curve minIs the minimum x-axis coordinate, x, of the time-domain waveform curvemaxThe maximum x-axis coordinate of the time domain waveform curve.
Wherein, yiIs the y-axis coordinate, y, of the ith point of the time-domain waveform curveminIs the minimum y-axis coordinate, y, of the time domain waveform curvemaxIs the maximum y-axis coordinate of the time domain waveform curve,
and the two-dimensional data of each signal spectrogram processed by the data preprocessing step and the satellite transponder name label corresponding to the signal spectrogram form training data, and a training data set is constructed by utilizing a plurality of training data.
2. The method according to claim 1, wherein the step S02 further comprises a data cleansing step of cleansing data of the original data received from the receiver according to the prior information, and cleansing data with error information.
3. The method of claim 1, wherein in the data preprocessing step, the background and the time-domain waveform curve are distinguished by color to extract the time-domain waveform curve from the signal spectrogram.
4. The method according to claim 1, wherein in the step S01, the satellite transponder identification model is constructed using an artificial neural network based on PyTorch framework construction.
5. The method of claim 4, wherein the satellite transponder identification model is comprised of three layers of fully-connected neural networks, the activation function between each layer of fully-connected neural networks is leak _ relu, the last layer is softmax layer, and the loss function is MSELoss.
6. The method of claim 1, wherein the signal spectrogram is a time domain oscillogram.
7. The method of claim 1, wherein the input of the satellite transponder identification model is a training data set and the output of the satellite transponder identification model is a name tag of the satellite transponder to which the signal belongs.
8. A satellite transponder identification model trained by the method according to any one of claims 1-7.
9. A method of using a satellite transponder identification model for identification using the satellite transponder identification model of claim 8.
10. The method of claim 9, the method comprising:
s11: acquiring a signal from a receiver, and drawing a signal spectrogram according to original data of the signal;
s12: extracting time domain waveform curve in the signal spectrogram, and storing the time domain waveform curve as two-dimensional data ((x)1,y1),(x2,y2),…(xn,yn));
S13: two-dimensional data ((x) of the extracted time-domain waveform curve 1,y1),(x2,y2),…(xn,yn) Subjected to the following treatment:
xi=(xi-xmin)/(xmax-xmin)
yi=(yi-ymin)/(ymax-ymin)
Wherein x isiIs the x-axis coordinate, x, of the ith point of the time-domain waveform curveminIs the minimum x-axis coordinate, x, of the time-domain waveform curvemaxIs the maximum x-axis coordinate of the time domain waveform curve,
wherein, yiIs the y-axis coordinate, y, of the ith point of the time-domain waveform curveminIs the minimum y-axis coordinate, y, of the time domain waveform curvemaxThe maximum y-axis coordinate of the time domain waveform curve;
s14: and inputting the two-dimensional data obtained in the step S13 into a satellite transponder identification model, and outputting the name label of the satellite transponder to which the signal spectrogram belongs.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110670593.2A CN113269135A (en) | 2021-06-17 | 2021-06-17 | Satellite transponder recognition model and training method and using method thereof |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110670593.2A CN113269135A (en) | 2021-06-17 | 2021-06-17 | Satellite transponder recognition model and training method and using method thereof |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113269135A true CN113269135A (en) | 2021-08-17 |
Family
ID=77235162
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110670593.2A Pending CN113269135A (en) | 2021-06-17 | 2021-06-17 | Satellite transponder recognition model and training method and using method thereof |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113269135A (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110705508A (en) * | 2019-10-15 | 2020-01-17 | 中国人民解放军战略支援部队航天工程大学 | Satellite identification method of ISAR image |
CN111562597A (en) * | 2020-06-02 | 2020-08-21 | 南京敏智达科技有限公司 | Beidou satellite navigation interference source identification method based on BP neural network |
CN112083393A (en) * | 2020-10-27 | 2020-12-15 | 西安电子科技大学 | Intermittent sampling forwarding interference identification method based on spectrogram average time characteristic |
CN112184705A (en) * | 2020-10-28 | 2021-01-05 | 成都智数医联科技有限公司 | Human body acupuncture point identification, positioning and application system based on computer vision technology |
-
2021
- 2021-06-17 CN CN202110670593.2A patent/CN113269135A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110705508A (en) * | 2019-10-15 | 2020-01-17 | 中国人民解放军战略支援部队航天工程大学 | Satellite identification method of ISAR image |
CN111562597A (en) * | 2020-06-02 | 2020-08-21 | 南京敏智达科技有限公司 | Beidou satellite navigation interference source identification method based on BP neural network |
CN112083393A (en) * | 2020-10-27 | 2020-12-15 | 西安电子科技大学 | Intermittent sampling forwarding interference identification method based on spectrogram average time characteristic |
CN112184705A (en) * | 2020-10-28 | 2021-01-05 | 成都智数医联科技有限公司 | Human body acupuncture point identification, positioning and application system based on computer vision technology |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109118479B (en) | Capsule network-based insulator defect identification and positioning device and method | |
CN108492272B (en) | Cardiovascular vulnerable plaque identification method and system based on attention model and multitask neural network | |
CN110879989B (en) | Ads-b signal target identification method based on small sample local machine learning model | |
CN109767422B (en) | Pipeline detection and identification method based on deep learning, storage medium and robot | |
CN108564115A (en) | Semi-supervised polarization SAR terrain classification method based on full convolution GAN | |
CN111985499B (en) | High-precision bridge apparent disease identification method based on computer vision | |
CN111091544B (en) | Method for detecting breakage fault of side integrated framework of railway wagon bogie | |
CN104268538A (en) | Online visual inspection method for dot matrix sprayed code characters of beverage cans | |
CN108171119B (en) | SAR image change detection method based on residual error network | |
CN109946304A (en) | Surface defects of parts on-line detecting system and detection method based on characteristic matching | |
CN110647830A (en) | Bearing fault diagnosis method based on convolutional neural network and Gaussian mixture model | |
CN106951863B (en) | Method for detecting change of infrared image of substation equipment based on random forest | |
CN115661649B (en) | BP neural network-based shipborne microwave radar image oil spill detection method and system | |
CN111860277A (en) | Civil aviation aircraft airspeed pipe sleeve safety warning method based on color histogram features | |
CN114170511A (en) | Pavement crack disease identification method based on Cascade RCNN | |
CN106604193A (en) | Defect detection method and system applied to receiver | |
CN106645180A (en) | Method for checking defects of substrate glass, field terminal and server | |
CN114998215A (en) | Method and system for detecting abnormal paint spraying on surface of sewing machine | |
CN105894035B (en) | SAR image classification method based on SAR-SIFT and DBN | |
CN113076920B (en) | Intelligent fault diagnosis method based on asymmetric domain confrontation self-adaptive model | |
CN112257621A (en) | Equipment image identification method for unmanned aerial vehicle inspection | |
CN113269135A (en) | Satellite transponder recognition model and training method and using method thereof | |
CN111325076B (en) | Aviation ground building extraction method based on fusion of U-net and Seg-net networks | |
CN109829887B (en) | Image quality evaluation method based on deep neural network | |
CN110349119B (en) | Pavement disease detection method and device based on edge detection neural network |
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 |