CN112633320B - Radar radiation source data cleaning method based on phase image coefficient and DBSCAN - Google Patents
Radar radiation source data cleaning method based on phase image coefficient and DBSCAN Download PDFInfo
- Publication number
- CN112633320B CN112633320B CN202011348266.7A CN202011348266A CN112633320B CN 112633320 B CN112633320 B CN 112633320B CN 202011348266 A CN202011348266 A CN 202011348266A CN 112633320 B CN112633320 B CN 112633320B
- Authority
- CN
- China
- Prior art keywords
- radiation source
- sample
- radar radiation
- data set
- samples
- 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.)
- Active
Links
- 230000005855 radiation Effects 0.000 title claims abstract description 113
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000004140 cleaning Methods 0.000 title claims abstract description 26
- 239000013598 vector Substances 0.000 claims abstract description 26
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 16
- 238000007781 pre-processing Methods 0.000 claims abstract description 4
- 238000005070 sampling Methods 0.000 claims description 34
- 238000012545 processing Methods 0.000 claims description 12
- 238000010606 normalization Methods 0.000 claims description 8
- 238000000605 extraction Methods 0.000 claims description 5
- 238000009499 grossing Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000002596 correlated effect Effects 0.000 claims description 3
- 230000000875 corresponding effect Effects 0.000 claims description 3
- 239000000284 extract Substances 0.000 abstract description 2
- 238000004088 simulation Methods 0.000 description 7
- 238000012360 testing method Methods 0.000 description 7
- 230000007547 defect Effects 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- Computer Networks & Wireless Communication (AREA)
- Remote Sensing (AREA)
- Probability & Statistics with Applications (AREA)
- Quality & Reliability (AREA)
- Radar, Positioning & Navigation (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a radar radiation source data cleaning method based on a phase coefficient and DBSCAN, which mainly solves the problem that the existing cleaning method cannot clean the data of radar radiation source data. The method comprises the following implementation steps: the method comprises the following steps of (1) obtaining a radar radiation source signal sample; (2) preprocessing samples in the data set; (3) Calculating a phase coefficient of each sample in the radar radiation source data set; (4) generating a feature vector; (5) Clustering a radar radiation source signal data set by using a DBSCAN algorithm; and (6) cleaning the noise sample. The invention extracts the phase image coefficient characteristics of the radar radiation source signal and divides the noise signal and the effective radar pulse signal by using a DBSCAN clustering method, so that the invention can obtain a radar radiation source data set with fewer noise samples and higher data quality.
Description
Technical Field
The invention belongs to the technical field of communication, and further relates to a radar radiation source data cleaning method Based on a Noise-containing data Spatial Clustering algorithm DBSCAN (sensitivity-Based Spatial Clustering of Applications with Noise) in the technical field of radar signal processing. The invention can be used for cleaning noise data in radiation source signals received by a radar in electronic information reconnaissance, electronic support and threat warning systems.
Background
With the rapid development of radar technology, the electromagnetic environment is more and more complex, the number of radiation source signals intercepted by electronic reconnaissance equipment is increased continuously, and noise data and effective data are mixed together, so that the difficulty in acquiring battlefield situation information is increased. On the other hand, in the field of radar countermeasure, many researchers introduce methods in the fields of artificial intelligence, data mining and the like, and the requirements of high-performance data-driven algorithms on data quality are higher. Therefore, the method has important significance for cleaning the data of the original radar radiation source signal and improving the data quality. At present, the proposed data cleaning method is mostly used for data of types such as two-dimensional tables, time series, images and the like, and is mainly applied to the fields of medical treatment, energy, retail sale, automobiles, finance and the like. However, the existing data cleaning method is not suitable for radar radiation source data.
Shanghai Re-Sr information technology Co., ltd discloses a sample data cleaning method and system in the patent document applied by Shanghai Re-Sr information technology Co., ltd. (patent application No. 201910239561.X, application publication No. CN 111651433A). The method comprises the specific steps of (1) acquiring a multi-dimensional test feature vector of each test picture in a test picture set according to a neural network model; (2) Acquiring a multi-dimensional reference characteristic vector in the selected typical picture; (3) Generating a positive sample test picture set and a negative sample test picture set according to a K nearest neighbor algorithm, a multi-dimensional test feature vector and a multi-dimensional reference feature vector, and training to obtain a fine-grained two-classifier; (4) Performing class prediction on the picture data to be cleaned according to the fine-grained secondary classifier, and acquiring the confidence coefficient of the class prediction of each picture data to be cleaned; (5) And cleaning the sample data according to a preset confidence interval and the confidence of the class prediction of each piece of picture data to be cleaned. The method can obtain a better positive sample test picture set and a sample picture set, and realizes automatic data cleaning. However, the method still has the defects that only image data can be processed, while radar radiation source data is a time domain sequence with a specific inter-pulse modulation mode and an intra-pulse modulation mode, and the method cannot clean the data.
Disclosure of Invention
The invention aims to provide a radar radiation source data cleaning method based on a phase contrast coefficient and DBSCAN (direct space division multiple access) to overcome the defects in the prior art, and the problem that the existing data cleaning method cannot process radar radiation source data is solved.
The technical idea for realizing the purpose of the invention is as follows: the method extracts the phase image coefficient of the radar radiation source signal as the characteristic, utilizes the DBSCAN clustering algorithm to cluster and divide data, detects and eliminates noise data, and solves the problem that the prior art cannot clean the data of the radar radiation source signal. Firstly extracting the envelope of a radiation source signal, then obtaining a rectangular phase coefficient and a triangular phase coefficient of the envelope as features to form a feature vector, finally performing cluster division on feature data by using a DBSCAN algorithm, distinguishing a noise signal and a pulse signal, and rejecting the noise signal.
The method comprises the following specific steps:
(1) Obtaining radar radiation source signal samples:
(1a) Converting the radar high-frequency pulse signal received by the radar receiver into an intermediate-frequency signal by using a low-pass filter;
(1b) Collecting at least 500 samples from the intermediate frequency signal by adopting a sampling frequency not lower than 500Hz to form a radar radiation source signal data set;
(2) Pre-processing samples in the dataset:
(2a) Extracting an envelope value of each sample in a radar radiation source signal data set by utilizing a normalized Shannon energy envelope extraction algorithm;
(2b) Carrying out normalization processing on the envelope value of each sample in the radar radiation source signal data set by using a min-max normalization method;
(3) Calculating the phase coefficient of each sample in the radar radiation source data set:
(3a) Calculating a rectangular phase coefficient of each sample in a radar radiation source signal data set by using a rectangular phase coefficient formula;
(3b) Calculating a triangular phase coefficient of each sample in the radar radiation source signal data set by using a triangular phase coefficient formula;
(4) Generating a feature vector:
connecting the rectangular phase coefficient of each sample in the radar radiation source signal data set with the triangular phase coefficient of each sample in an end-to-end manner to generate a sample characteristic vector;
(5) Clustering the radar radiation source signal data set by using a DBSCAN algorithm:
(5a) Randomly selecting an unprocessed sample from a radar radiation source signal data set as a current processing sample, calculating the Euclidean distance between the feature vector of the current processing sample and the feature vector of each sample in the radar radiation source data set, selecting all samples with Euclidean distances smaller than a neighborhood radius epsilon from the samples, counting the number alpha of the samples, and executing a step (5 b), wherein the size of the neighborhood radius epsilon is positively correlated with the total number of the samples in the radar radiation source data set;
(5b) Judging whether alpha is larger than or equal to a neighborhood parameter MinPts, if so, executing the step (5 c), otherwise, executing the step (5 i), wherein the size of the neighborhood parameter MinPts is in negative correlation with a neighborhood radius epsilon;
(5c) Putting the current processed sample and all selected samples with Euclidean distance smaller than the neighborhood radius epsilon into an empty set M, marking all samples in the set M as unprocessed samples, and then executing the step (5 d);
(5d) Taking an unprocessed sample from the M set as an operated sample lambda, calculating Euclidean distances between the characteristic vector of the operated sample lambda and the characteristic vector of each sample in the radar radiation source signal data set respectively, selecting all samples with Euclidean distances smaller than the neighborhood radius epsilon from the characteristic vectors, counting the number beta of the samples, marking the operated sample lambda as a processed sample, and then executing the step (5 e);
(5e) Judging whether beta is greater than or equal to MinPts, if so, executing the step (5 f), otherwise, executing the step (5 g);
(5f) Putting all samples with Euclidean distance smaller than the radius epsilon of the neighborhood into an M set, marking the samples as unprocessed samples, and then executing the step (5 g);
(5g) Judging whether all unprocessed samples in the M sets are selected, if so, executing the step (5 h), and otherwise, executing the step (5 d);
(5h) Step (5 i) is executed after all processed samples in the M sets form a cluster;
(5i) Judging whether all the radar radiation source data sets are processed samples, if so, executing the step (6) after obtaining a cluster corresponding to each sample, otherwise, executing the step (5 a);
(6) Cleaning a noise sample:
(6a) Randomly extracting 20 samples from each cluster, respectively finding out samples with the rectangular phase coefficient larger than 0.9, and counting the number of the samples;
(6b) And finding out the cluster with the maximum number of samples with the rectangular phase coefficient larger than 0.9 from the 20 samples, and deleting all samples in the cluster from the radar radiation source signal data set to obtain the cleaned radar radiation source signal data set.
Compared with the prior art, the invention has the following advantages:
firstly, the computed phase image coefficient of each sample in the radar radiation source data set is used as the characteristic value of the sample, the characteristic value can effectively reflect the difference between a noise sample and a pulse sample, only two characteristics of the rectangular and triangular phase image coefficients of each sample need to be computed, and the computation is simple, so that the noise data in the radar radiation source data set can be effectively and quickly eliminated.
Secondly, because the invention utilizes the DBSCAN algorithm to cluster the radar radiation source signal data set, the noise data in the radar radiation source data set can be effectively removed, and the problem that the data cleaning of the radar radiation source data can not be carried out in the prior art is solved, so that the invention can obtain the radar radiation source data set with fewer noise samples and higher data quality.
Drawings
FIG. 1 is a flow chart of the present invention.
Detailed Description
The following describes the specific implementation steps of the present invention with reference to fig. 1.
Step 1, obtaining a radar radiation source signal sample.
And utilizing a low-pass filter to convert the radar high-frequency pulse signal received by the radar receiver into an intermediate-frequency signal.
And at least 500 samples are collected from the intermediate frequency signals by adopting a sampling frequency not lower than 500Hz to form a radar radiation source signal data set.
And 2, preprocessing the samples in the data set.
And extracting the envelope value of each sample in the radar radiation source signal data set by utilizing a normalized Shannon energy envelope extraction algorithm.
The normalized shannon energy envelope extraction algorithm comprises the following specific steps:
step 1, normalizing each sampling point of each sample in a radar radiation source signal data set according to the following formula:
wherein,
a normalized value, x, representing the ith sample point in the jth sample in the radar radiation source signal data set j (i) Representing the amplitude value, x, of the ith sample in the jth sample in the radar radiation source signal data set j And the sequence is composed of amplitude values of all sampling points in the jth sample in the radar radiation source signal data set, max (·) represents the operation of solving the maximum value, and | is | represents the operation of taking the absolute value.
Step 2, calculating the shannon energy of each sampling point of each sample in the radar radiation source signal data set according to the following formula:
wherein E is j (i) The shannon energy of the ith sampling point in the jth sample in the radar radiation source signal data set is represented, and log (-) represents the logarithmic operation with the base 10 as the base.
And 3, performing windowing smoothing treatment on the shannon energy of each sampling point of each sample in the radar radiation source signal data set according to the following formula:
wherein,
representing the Shannon energy after the ith sampling point in the jth sample in the radar radiation source signal data set is smoothed, N representing the number of sampling points in the windowing smoothing processing window, the value is 200, and sigma represents the summation operation.
And 4, calculating an envelope value of each sampling point of each sample in the radar radiation source signal data set according to the following formula:
wherein, P j (i) Representing the envelope value of the ith sample point in the jth sample in the radar radiation source signal data set,
and the sequence of the jth sample in the radar radiation source signal data set, which consists of the smoothed shannon energy of all sampling points, is represented, mean (-) represents the averaging operation, and S (-) represents the standard deviation operation.
And (4) carrying out normalization processing on the envelope value of each sample in the radar radiation source signal data set by using a min-max normalization method.
The min-max normalization is as follows:
wherein,
a normalized value, P, representing the envelope of the ith sample point in the jth sample in the radar radiation source signal data set j And a sequence consisting of envelope values of all sampling points in the jth sample in the radar radiation source signal data set is represented, and min (-) represents the minimum value calculation operation.
And 3, calculating the phase image coefficient of each sample in the radar radiation source data set.
Calculating the rectangular phase coefficient of each sample in the radar radiation source signal data set by using the following rectangular phase coefficient formula:
wherein, C j Representing the rectangular phase coefficient of the jth sample in the radar radiation source signal data set, M representing the number of sampling points of the sample in the radar radiation source signal data set, S j (k) Represents a sequence formed by normalized envelope values of all sampling points of the jth sample in the radar radiation source signal data set, U (k) represents a rectangular reference sequence with the number of sampling points being M and each sampling value being 1,
indicating a square root operation.
Calculating the triangular phase coefficient of each sample in the radar radiation source signal data set by using the following triangular phase coefficient formula:
wherein, I j And T (n) represents a triangular reference sequence with the number of the sampling points M.
And 4, generating a feature vector.
And connecting the rectangular phase coefficient of each sample in the radar radiation source signal data set with the triangular phase coefficient thereof end to generate the sample characteristic vector.
And 5, clustering the radar radiation source signal data set by using a DBSCAN algorithm.
Step 1, randomly selecting an unprocessed sample from a radar radiation source signal data set as a current processing sample, calculating Euclidean distances between feature vectors of the current processing sample and feature vectors of each sample in the radar radiation source data set, selecting all samples with Euclidean distances smaller than a neighborhood radius epsilon from the samples, counting the number alpha of the samples, and executing step 2, wherein the size of the neighborhood radius epsilon is positively correlated with the total number of the samples in the radar radiation source data set.
And 2, judging whether alpha is larger than or equal to a neighborhood parameter MinPts, if so, executing the step 3, otherwise, executing the ninth step, wherein the size of the neighborhood parameter MinPts is in negative correlation with the neighborhood radius epsilon.
And 3, putting the currently processed sample and all selected samples with Euclidean distances smaller than the neighborhood radius epsilon into an empty set M, marking all samples in the set M as unprocessed samples, and then executing the 4 th step.
And 4, taking an unprocessed sample from the M set as an operated sample lambda, calculating Euclidean distances between the characteristic vector of the operated sample lambda and the characteristic vector of each sample in the radar radiation source signal data set, selecting all samples with Euclidean distances smaller than the neighborhood radius epsilon, counting the number beta of the samples, and executing the 5 th step after marking the operated sample lambda as a processed sample.
And 5, judging whether the beta is greater than or equal to MinPts, if so, executing the step 6, and otherwise, executing the step 7.
And 6, putting all samples with Euclidean distances smaller than the radius epsilon of the neighborhood into an M set, marking the samples as unprocessed samples, and then executing the 7 th step.
And 7, judging whether all unprocessed samples in the M set are selected, if so, executing the step 8, and otherwise, executing the step 4.
And 8, forming all processed samples in the M set into a cluster, and then executing the 9 th step.
And 9, judging whether all the radar radiation source data sets are processed samples, if so, executing the step 6 after obtaining the clustering cluster corresponding to each sample, and otherwise, executing the step 1.
And 6, cleaning the noise sample.
Randomly extracting 20 samples from each cluster, finding out samples with the rectangular phase coefficient larger than 0.9, and counting the number of the samples.
And finding out the cluster with the maximum number of samples with the rectangular phase coefficient larger than 0.9 from the 20 samples, and deleting all samples in the cluster from the radar radiation source signal data set to obtain the cleaned radar radiation source signal data set.
The effect of the present invention is further explained by combining the simulation experiment as follows:
1. simulation conditions are as follows:
the hardware platform of the simulation experiment of the invention is as follows: the processor is Intel (R) Core i5-8300H, the main frequency is 2.30GHZ and the internal memory is 8GB.
The software platform of the simulation experiment of the invention is as follows: WINDOWS 10 operating System, MATLAB R2018a.
The data used in the simulation experiment of the invention is a radar radiation source data set formed by 5000 samples collected by a radar simulator, the data set comprises two types of samples of noise signals and pulse signals, different labels are manually marked on the different types of samples, wherein the number of the labeled noise label samples is 3796, and the number of the labeled pulse label samples is 1204.
2. Simulation content and result analysis:
the simulation experiment of the invention is to adopt the method of the invention to carry out data cleaning on 5000 samples in a data set, delete noise signal samples from the data set and keep pulse signal samples. And counting the number of correctly deleted and incorrectly deleted samples by taking the manually marked label as a reference, dividing the number of correctly deleted samples by the total number of samples of 5000 to obtain the correct rate of data cleaning by the method, and drawing all calculation results into a table 1.
TABLE 1 statistical table of marking conditions of measured data samples
| Total number of samples | Correct number of samples to delete | Number of erroneous erasure samples | Accuracy (%) |
| 5000 | 4990 | 10 | 99.8％ |
As can be seen from Table 1, the accuracy of the method for cleaning the data of the radar radiation source can reach 99.8%, and the level of manual cleaning is basically reached. Therefore, the invention can effectively clean and remove the noise sample and improve the data quality.
Claims (3)
1. A radar radiation source data cleaning method based on a phase image coefficient and DBSCAN is characterized in that the phase image coefficient of each sample in a radar radiation source data set is calculated to serve as a characteristic, a DBSCAN algorithm is used for clustering a radar radiation source signal data set, and noise data are detected and removed, and the method comprises the following steps:
(1) Obtaining radar radiation source signal samples:
(1a) Converting the radar high-frequency pulse signal received by the radar receiver into an intermediate-frequency signal by using a low-pass filter;
(1b) Collecting at least 500 samples from the intermediate frequency signals by adopting a sampling frequency not lower than 500Hz to form a radar radiation source signal data set;
(2) Pre-processing samples in the dataset:
(2a) Extracting an envelope value of each sample in a radar radiation source signal data set by utilizing a normalized Shannon energy envelope extraction algorithm;
the normalized shannon energy envelope extraction algorithm comprises the following specific steps:
firstly, normalizing each sampling point of each sample in a radar radiation source signal data set according to the following formula:
wherein,
indicating the normalized value, x, of the ith sampling point in the jth sample in the radar radiation source signal data set j (i) An amplitude value, x, representing the ith sample point in the jth sample in the radar radiation source signal data set j Representing a sequence consisting of amplitude values of all sampling points in the jth sample in the radar radiation source signal data set, wherein max (·) represents the operation of solving the maximum value, and | is | represents the operation of taking an absolute value;
secondly, calculating the shannon energy of each sampling point of each sample in the radar radiation source signal data set according to the following formula:
wherein E is j (i) Representing the Shannon energy of the ith sampling point in the jth sample in the signal data set of the radar radiation source, and log (-) representing the logarithmic operation with 10 as the base;
thirdly, windowing and smoothing the shannon energy of each sampling point of each sample in the radar radiation source signal data set according to the following formula:
wherein,
representing the Shannon energy after the ith sampling point in the jth sample in the radar radiation source signal data set is smoothed, N represents the number of sampling points in a windowing smoothing processing window, the value is 200, and sigma represents the summation operation;
fourthly, calculating an envelope value of each sampling point of each sample in the radar radiation source signal data set according to the following formula:
wherein, P j (i) Representing the envelope value of the ith sample point in the jth sample in the radar radiation source signal data set,
representing a sequence consisting of smoothed shannon energy of all sampling points in the jth sample in the radar radiation source signal data set, mean (-) representing the operation of taking the mean value, and S (-) representing the operation of taking the standard deviation;
(2b) Carrying out normalization processing on the envelope value of each sample in the radar radiation source signal data set by using a min-max normalization method;
the min-max normalization is as follows:
wherein,
a normalized value, P, representing the envelope of the ith sample point in the jth sample in the radar radiation source signal data set j Representing a sequence consisting of envelope values of all sampling points in the jth sample in the radar radiation source signal data set, wherein min (-) represents the minimum value calculation operation;
(3) Calculating a phase coefficient of each sample in the radar radiation source data set:
(3a) Calculating the rectangular phase coefficient of each sample in the radar radiation source signal data set by using a rectangular phase coefficient formula;
(3b) Calculating the triangular phase coefficient of each sample in the radar radiation source signal data set by using a triangular phase coefficient formula;
(4) Generating a feature vector:
connecting the rectangular phase coefficient of each sample in the radar radiation source signal data set with the triangular phase coefficient of each sample in an end-to-end manner to generate a sample characteristic vector;
(5) Clustering the radar radiation source signal data set by using a DBSCAN algorithm:
(5a) Randomly selecting an unprocessed sample from a radar radiation source signal data set as a current processing sample, calculating the Euclidean distance between the feature vector of the current processing sample and the feature vector of each sample in the radar radiation source data set, selecting all samples with Euclidean distances smaller than a neighborhood radius epsilon from the samples, counting the number alpha of the samples, and executing a step (5 b), wherein the size of the neighborhood radius epsilon is positively correlated with the total number of the samples in the radar radiation source data set;
(5b) Judging whether alpha is larger than or equal to a neighborhood parameter MinPts, if so, executing the step (5 c), otherwise, executing the step (5 i), wherein the size of the neighborhood parameter MinPts is in negative correlation with a neighborhood radius epsilon;
(5c) Putting the current processed sample and all selected samples with Euclidean distances smaller than the neighborhood radius epsilon into an empty set M, and executing the step (5 d) after all samples in the set M are marked as unprocessed samples;
(5d) Taking an unprocessed sample from the M set as an operated sample lambda, calculating Euclidean distances between the characteristic vector of the operated sample lambda and the characteristic vector of each sample in the radar radiation source signal data set respectively, selecting all samples with Euclidean distances smaller than the neighborhood radius epsilon from the characteristic vectors, counting the number beta of the samples, marking the operated sample lambda as a processed sample, and then executing the step (5 e);
(5e) Judging whether beta is greater than or equal to MinPts, if so, executing the step (5 f), otherwise, executing the step (5 g);
(5f) Putting all samples with Euclidean distance smaller than the radius epsilon of the neighborhood into an M set, marking the samples as unprocessed samples, and then executing the step (5 g);
(5g) Judging whether all unprocessed samples in the M set are selected, if so, executing the step (5 h), otherwise, executing the step (5 d);
(5h) Step (5 i) is executed after all processed samples in the M sets form a cluster;
(5i) Judging whether all the radar radiation source data sets are processed samples, if so, executing the step (6) after obtaining a cluster corresponding to each sample, otherwise, executing the step (5 a);
(6) Cleaning a noise sample:
(6a) Randomly extracting 20 samples from each cluster, respectively finding out samples of which the rectangular phase coefficient is greater than 0.9, and counting the number of the samples;
(6b) And finding out the cluster with the maximum number of samples with the rectangular phase coefficient larger than 0.9 from the 20 samples, and deleting all samples in the cluster from the radar radiation source signal data set to obtain the cleaned radar radiation source signal data set.
2. The method for cleaning radar radiation source data based on phase contrast ratio and DBSCAN of claim 1, wherein the rectangular phase contrast ratio formula in step (3 a) is as follows:
wherein, C j Representing the rectangular phase coefficient of the jth sample in the radar radiation source signal data set, M representing the number of sampling points of the sample in the radar radiation source signal data set, S j (k) Representing a sequence formed by normalized envelope values of all sampling points of the jth sample in a radar radiation source signal data set, U (k) representing a rectangular reference sequence with the number of the sampling points being M and each sampling value being 1,
indicating a square root operation.
3. The method for cleaning radar radiation source data based on the phase coefficient and the DBSCAN of claim 2, wherein the triangular phase coefficient formula in the step (3 b) is as follows:
wherein, I j And T (n) represents a triangular reference sequence with the number of the sampling points M.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011348266.7A CN112633320B (en) | 2020-11-26 | 2020-11-26 | Radar radiation source data cleaning method based on phase image coefficient and DBSCAN |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011348266.7A CN112633320B (en) | 2020-11-26 | 2020-11-26 | Radar radiation source data cleaning method based on phase image coefficient and DBSCAN |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112633320A CN112633320A (en) | 2021-04-09 |
| CN112633320B true CN112633320B (en) | 2023-04-07 |
Family
ID=75304157
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011348266.7A Active CN112633320B (en) | 2020-11-26 | 2020-11-26 | Radar radiation source data cleaning method based on phase image coefficient and DBSCAN |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112633320B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113128598B (en) * | 2021-04-22 | 2024-04-09 | 深信服科技股份有限公司 | Sensing data detection method, device, equipment and readable storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102135979A (en) * | 2010-12-08 | 2011-07-27 | 华为技术有限公司 | Data cleaning method and device |
| CN110928862A (en) * | 2019-10-23 | 2020-03-27 | 深圳市华讯方舟太赫兹科技有限公司 | Data cleaning method, data cleaning apparatus, and computer storage medium |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050028046A1 (en) * | 2003-07-31 | 2005-02-03 | International Business Machines Corporation | Alert flags for data cleaning and data analysis |
| CN108090412A (en) * | 2017-11-17 | 2018-05-29 | 西北工业大学 | A kind of radar emission source category recognition methods based on deep learning |
| CN110083728B (en) * | 2019-04-03 | 2021-08-20 | 上海铼锶信息技术有限公司 | Method, device and system for optimizing automatic picture data cleaning quality |
| CN110263230B (en) * | 2019-04-25 | 2021-04-06 | 北京科技大学 | Data cleaning method and device based on density clustering |
-
2020
- 2020-11-26 CN CN202011348266.7A patent/CN112633320B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102135979A (en) * | 2010-12-08 | 2011-07-27 | 华为技术有限公司 | Data cleaning method and device |
| CN110928862A (en) * | 2019-10-23 | 2020-03-27 | 深圳市华讯方舟太赫兹科技有限公司 | Data cleaning method, data cleaning apparatus, and computer storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112633320A (en) | 2021-04-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107220606B (en) | Radar radiation source signal identification method based on one-dimensional convolutional neural network | |
| CN110927706B (en) | Convolutional neural network-based radar interference detection and identification method | |
| CN108492272B (en) | Cardiovascular vulnerable plaque identification method and system based on attention model and multitask neural network | |
| CN107563433B (en) | Infrared small target detection method based on convolutional neural network | |
| CN114595732B (en) | Radar radiation source sorting method based on depth clustering | |
| CN108376403B (en) | Grid colony image segmentation method based on Hough circle transformation | |
| CN111832462B (en) | Frequency hopping signal detection and parameter estimation method based on deep neural network | |
| CN111428631B (en) | Visual identification and sorting method for unmanned aerial vehicle flight control signals | |
| CN112560803A (en) | Radar signal modulation identification method based on time-frequency analysis and machine learning | |
| CN105117736B (en) | Classification of Polarimetric SAR Image method based on sparse depth heap stack network | |
| CN110956632B (en) | Method and device for automatically detecting pectoralis major region in molybdenum target image | |
| CN103824302B (en) | The SAR image change detection merged based on direction wave area image | |
| CN108171119B (en) | SAR image change detection method based on residual error network | |
| Suryavamsi et al. | Comparative analysis of various enhancement methods for astrocytoma MRI images | |
| CN112633320B (en) | Radar radiation source data cleaning method based on phase image coefficient and DBSCAN | |
| CN110751667B (en) | Method for detecting infrared dim and small targets under complex background based on human visual system | |
| Verma et al. | Edge detection using adaptive thresholding and ant colony optimization | |
| CN115935144A (en) | Denoising and reconstructing method for operation and maintenance data | |
| CN114863189A (en) | Intelligent image identification method based on big data | |
| CN113313179A (en) | Noise image classification method based on l2p norm robust least square method | |
| Neelakanteswara et al. | Computer based advanced approach for MRI image classification using neural network with the texture features extracted | |
| CN116079713A (en) | Multi-drill-arm cooperative control method, system, equipment and medium for drilling and anchoring robot | |
| CN112329796B (en) | Infrared imaging cloud detection method and device based on visual saliency | |
| CN114627164A (en) | Image local structure difference evaluation method and system | |
| Wu et al. | Time-frequency parameter estimation method of frequency hopping signal based on morphology method under low SNR |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |