CN115840875A - Millimeter wave radar abnormal signal detection method and system based on analog transducer - Google Patents
Millimeter wave radar abnormal signal detection method and system based on analog transducer Download PDFInfo
- Publication number
- CN115840875A CN115840875A CN202211405034.XA CN202211405034A CN115840875A CN 115840875 A CN115840875 A CN 115840875A CN 202211405034 A CN202211405034 A CN 202211405034A CN 115840875 A CN115840875 A CN 115840875A
- Authority
- CN
- China
- Prior art keywords
- millimeter wave
- wave radar
- data
- correlation
- training
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 51
- 230000002159 abnormal effect Effects 0.000 title claims abstract description 38
- 238000012549 training Methods 0.000 claims abstract description 30
- 238000012544 monitoring process Methods 0.000 claims abstract description 13
- 238000004458 analytical method Methods 0.000 claims abstract description 10
- 230000007246 mechanism Effects 0.000 claims abstract description 7
- 230000005856 abnormality Effects 0.000 claims abstract description 6
- 230000000903 blocking effect Effects 0.000 claims abstract description 5
- 238000010219 correlation analysis Methods 0.000 claims abstract description 5
- 230000000007 visual effect Effects 0.000 claims abstract description 3
- 238000000034 method Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 6
- 230000007123 defense Effects 0.000 claims description 6
- 238000002372 labelling Methods 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 claims description 5
- 238000004519 manufacturing process Methods 0.000 claims description 5
- 230000008569 process Effects 0.000 claims description 5
- 102100034112 Alkyldihydroxyacetonephosphate synthase, peroxisomal Human genes 0.000 claims description 3
- 101000799143 Homo sapiens Alkyldihydroxyacetonephosphate synthase, peroxisomal Proteins 0.000 claims description 3
- 238000000848 angular dependent Auger electron spectroscopy Methods 0.000 claims description 3
- 238000010801 machine learning Methods 0.000 claims description 3
- 230000008447 perception Effects 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 3
- 238000009826 distribution Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000004069 differentiation Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000013016 learning Effects 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000035045 associative learning Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
Images
Landscapes
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention belongs to the technical field of vehicle radar safety, and relates to a millimeter wave radar abnormal signal detection method and system based on an analog transducer, wherein correlation differences are calculated through an Attention mechanism analog-Attention, and prior correlation and sequence correlation are modeled in a unified manner; calculating symmetrical KL distances of prior association and sequence association of each layer to serve as a judgment basis of subsequent anomaly detection; distinguishing an abnormality judgment curve based on correlation analysis through visual analysis under different abnormality criteria; and (3) taking each sample as a group of related data, inputting the group of related data into an analog Transformer for training, and obtaining the normal working condition data monitoring model. The anomalyTransformer-based millimeter wave radar abnormal signal detection method is used for detecting abnormal data generated by attack on a millimeter wave radar, and blocking and defending are achieved.
Description
Technical Field
The invention belongs to the technical field of vehicle radar safety, and particularly relates to a millimeter wave radar abnormal signal detection method and system based on an analog transducer.
Background
The millimeter wave radar is a sensor widely used in a driving assisting system and automatic driving, and is mainly applied to functions of automatic vehicle following, emergency braking, collision early warning, blind spot monitoring at medium and long distances, lane changing assistance and the like. Along with the increasing popularization of intelligent automobiles, the development of information technology is faster and faster, the data sampling rate of a chip is increased rapidly, and the malicious attack mode aiming at the millimeter wave radar is changed. For example, by means of a high-sampling-rate radio frequency storage and forwarding technology, an electromagnetic wave detection signal transmitted by a millimeter wave radar can be copied and stored, and a false echo signal with extremely high similarity corresponding to the detection signal is rapidly forged, which is commonly called dense false target interference; even, echo signals corresponding to detection signals of actually detected detection targets can be directly collected, and then the collected echo signals are replayed at other time and transmitted to the vehicle, so that replay attack is realized; the attacks cause great property loss, and the interference prevention method for the millimeter wave radar provided in the related art is difficult to avoid the malicious attack, and how to avoid the malicious attack is a technical problem to be solved urgently.
From the nature of the time series, each time point can be represented by its association with the entire series, i.e. as an associated weight distribution in the time dimension. Compared with the characteristics at the point level, the association relationship implies the pattern information of the sequence, such as period, trend and the like, and is more information-content.
Meanwhile, compared with normal points, it is difficult for abnormal points to establish a strong correlation with the whole sequence dominated by the normal mode, and they tend to pay more attention to the neighboring regions (due to continuity). Thus, this difference in correlation with the global sequence, proximity priors, provides a natural, strongly discriminatory criterion for anomaly detection.
Therefore, a method and a system for detecting abnormal signals of millimeter wave radar are needed.
Disclosure of Invention
In order to solve the above technical problem, the present invention provides a method for detecting abnormal signals of millimeter wave radar based on analog transform, which includes the following steps:
step 1: calculating correlation difference through an Attention mechanism Anomaly-Attention, and uniformly modeling prior correlation and sequence correlation;
step 2: calculating symmetrical KL distances of prior association and sequence association of each layer to serve as a judgment basis of subsequent anomaly detection;
and step 3: distinguishing an abnormality judgment curve based on correlation analysis through visual analysis under different abnormality criteria;
and 4, step 4: and (3) taking each sample as a group of related data, inputting the group of related data into an analog Transformer for training, and obtaining the normal working condition data monitoring model.
Further, the correlation formula of Attention mechanism Anomaly-Attention is:
Initialization：
Prior-Association：
Series-Association：
Reconstruction：
further, the calculation formula of the symmetric KL distance is as follows:
further, the training method comprises the following steps:
step 1: manufacturing a training sample, building a TI vehicle-mounted millimeter wave radar data acquisition platform, and actually measuring a normal working condition data sample in an outdoor environment;
step 2: training an analog transform model training set, and introducing training set samples into the model in a vector form;
and step 3: and verifying the anomally Transformer model test set, and obtaining an anomally Transformer millimeter wave radar abnormal signal detection model by artificially manufacturing interference data and inputting the data into the anomally Transformer model.
On the other hand, the millimeter wave radar abnormal signal detection system based on the analog transducer comprises a target result receiving module, a target reliability defining, modeling and analyzing module, a state monitoring and fault management module, an execution controller, a CAN network, a data labeling platform and an AI training platform;
the target result receiving module receives the image data of the camera and the signal data sent by the millimeter wave radar, preprocesses the image data and the signal data, prepares for the subsequent calculation of the reliability of the target result, and uploads the target result to the data annotation platform;
the target credibility definition, modeling and analysis module analyzes the received result through a machine learning algorithm, judges whether the result is fraudulent data, and realizes detection and defense of ADAS perception data;
the state monitoring and fault management module monitors and manages the sensing data and the reliability result in real time to realize real-time monitoring of fraud identification;
the execution controller is used for generating a defense instruction and blocking fraudulent attacks;
the CAN network adopts a star structure, and all the CAN networks are connected through a central gateway;
the data labeling platform performs pseudo label labeling through a generation model and provides a data basis for incremental training;
and the AI training platform performs incremental training on the reliability analysis model and optimizes model parameters.
Furthermore, the operating frequency range of the millimeter wave radar is 30GHz to 300GHz, and the millimeter wave radar is used for detecting targets, measuring speed, measuring distance and measuring directions.
Further, the millimeter wave radar performs signal windowing by using a hanning window in the detection process, and the formula is as follows:
further, detection of the millimeter wave radar also involves frequency domain conversion, and the frequency domain conversion formula is as follows:
has the advantages that:
the millimeter wave radar abnormal signal detection method based on the anomallyTransformer is used for detecting abnormal data generated by attack on the millimeter wave radar, and blocking and defending are achieved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:
FIG. 1 is a diagram of a millimeter wave radar abnormal signal detection method based on an analog transducer;
FIG. 2 is a prior correlation and sequence correlation diagram of the data;
FIG. 3 is an additional correlation difference loss plot;
FIG. 4 is a graph of anomaly evaluation based on correlation analysis;
FIG. 5 is a graph of a prior correlation anomaly evaluation plot;
FIG. 6 is a flow chart of an anomally transducer-based millimeter wave radar abnormal signal detection system;
FIG. 7 is a normal millimeter wave radar data state diagram;
FIG. 8 is a flow chart of the main detection of data acquisition;
FIG. 9 is a target echo spectrum.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. 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.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.
An analog-Association unit is redesigned on the basis of an analog transform device and is used for learning potential time sequence Association from multi-level depth features, as shown in fig. 2, the analog-Association simultaneously models Prior Association (i.e. Prior of more Attention to neighboring regions) and sequence Association (i.e. dependency mined from data) of the data.
Meanwhile, a minimum and maximum strategy (Minimax) is adopted to further increase the gap of the correlation difference between the abnormal point and the normal point, so that the abnormal point can be detected more easily.
In order to calculate the correlation difference, a brand-new Attention mechanism, analog-Attention, is provided for uniformly modeling prior correlation and sequence correlation.
Initialization：
/>
Prior-Association：
Series-Association：
Reconstruction：
1. The a priori correlation is used to indicate a priori that each time instant is more concerned about its neighborhood due to the continuity of the time series. Is represented by a gaussian kernel function that learns the scale parameters. The kernel function is centered on the index of the corresponding time point, and due to the unimodal distribution characteristic of the gaussian distribution, the learned weight can be naturally concentrated in the domain of the corresponding time point. Meanwhile, the adaptive scale parameter can help the priori correlation to dynamically adapt to different timing modes.
2. Sequence correlation is used to represent the dependency mined directly from sequence data. The calculation method is similar to the calculation method of the attention matrix of the standard anomally Transformer, and the weight distribution of each row in the attention matrix corresponds to the sequence association of a time point. Meanwhile, in order to better complete the sequence reconstruction task, the model can actively mine reasonable time sequence dependence.
Through the design, the model can respectively capture the prior relevance and the sequence relevance, and compared with the past model which contains more abundant information based on the relevance representation.
The correlation difference is defined as the difference between the two, and is used as the judgment basis of the subsequent abnormal detection, and the correlation difference is calculated by the prior correlation of each layer and the symmetrical KL distance of the sequence correlation:
in addition to the reconstruction error widely used by the unsupervised task, an additional correlation difference loss is introduced to increase the difference between the normal point and the abnormal point, as shown in fig. 3, due to the unimodal characteristic of gaussian distribution in the prior correlation, the added correlation difference loss drives the sequence correlation to focus more on the non-adjacent region, so that the reconstruction of the abnormal point is harder, and the discrimination between the normal point and the abnormal point is easier.
In experiments, it is found that directly minimizing the correlation difference can make the learnable scale parameter in the prior distribution sharply smaller, resulting in model degradation. A Minimax (Minimax) strategy is therefore used for better control of the process of associative learning.
Minimize Phase：
Maximize Phase：
1. The fixed sequence correlation at the minimization stage is adopted, so that the prior correlation is approximate, and the prior correlation can be suitable for different time sequence modes;
2. in the maximization stage, the prior correlation is fixed, the sequence correlation is optimized to maximize the difference between the correlations, and the process can enable the sequence to focus more on non-adjacent and global points, so that the reconstruction of abnormal points is more difficult;
3. finally, the normalized correlation difference is combined with the reconstruction error to define a new anomaly detection criterion:
as shown in fig. 4, by visualizing the analysis of the abnormal judgment curves under different abnormal criteria, it can be found that the abnormal judgment curves based on the correlation analysis have more accurate distinguishability.
And performing parameter visualization in prior association on different abnormal categories. As shown in fig. 5, the outlier is often small compared to other points in the sequence, which represents a weak association with non-adjacent parts, proving that the outlier is difficult to construct a strong association a priori with the whole sequence.
As shown in fig. 6, a target result receiving module receives image data of a camera and signal data sent by a millimeter wave radar, performs preprocessing, prepares for subsequent calculation of reliability of a target result, and uploads the target result to a data annotation platform; the target credibility definition, modeling and analysis module analyzes the received result through a machine learning algorithm, judges whether the result is fraudulent data, and realizes detection and defense of ADAS perception data; the state monitoring and fault management module monitors and manages the sensing data and the reliability result in real time to realize real-time monitoring of fraud identification; the execution controller is used for generating a defense instruction and blocking fraudulent attacks; the CAN network adopts a star structure, and all the CAN networks are connected through a central gateway; the data labeling platform performs pseudo label labeling through a generation model and provides a data basis for incremental training; and the AI training platform performs incremental training on the reliability analysis model and optimizes model parameters.
The working frequency range of the millimeter wave radar is 30GHz to 300GHz, and the millimeter wave radar is used for detecting a target, measuring speed, measuring distance and measuring direction. When a normal target is present, the data state is as shown in FIG. 7.
And performing overall sequence construction on the normal radar signals at each time point through an analog Transformer to construct an association relation, namely representing the association weight distribution in the time dimension. This association implies pattern information of the sequence, such as period, trend, etc., compared to the features at the point level. Meanwhile, compared with the normal point, the abnormal point is difficult to construct a strong association relation with the whole sequence dominated by the normal mode. Therefore, the correlation difference between the overall sequence and the adjacent prior provides a strong distinguishing criterion for anomaly detection.
The analog Transformer model realizes the time sequence abnormity detection based on Association differentiation (Association differentiation). It contains the analysis-Attention mechanism for modeling the two forms of association separately, while the minimal maximum (Minimax) association learning strategy further increases the difference between the normal and abnormal points.
The millimeter wave radar uses a Hanning window to perform signal windowing in the detection process, and the formula is as follows:
the detection of the millimeter wave radar also involves frequency domain conversion, which has the formula:
the vehicle-mounted millimeter wave radar range Doppler image comprises a data situation under a normal working condition, meanwhile, data under the normal working condition and global data have high relevance, when fraudulent data appear, the relevance between a time point when the fraudulent data appear and the data in the above is greatly reduced, each sample is regarded as a group of related data and is input into an analog Transformer for training, and then a normal working condition data monitoring model is obtained.
The target detection model has high detection precision and strong environmental adaptability.
Based on the millimeter wave radar platform fraud detection model training, the target detection work in the actual measurement environment road is mainly carried out.
Data acquisition is all TI vehicle-mounted millimeter wave radar platform, the main detection flow chart is shown in figure 8,
the method comprises the following specific steps:
the first step is as follows: and (5) making a training sample. Firstly, a TI vehicle-mounted millimeter wave radar data acquisition platform is built, and normal working condition data samples are actually measured in an outdoor environment.
The second step is that: and (3) training an anomally Transformer model training set, and introducing training set samples into the model in a vector mode.
The third step: and verifying an anomally transform model test set. And artificially manufacturing interference data and inputting the interference data into an anomally transducer model to obtain the anomally transducer millimeter wave radar abnormal signal detection model.
As shown in fig. 9, based on the target echo spectrum image as the detection basis, the number Np of the protection window units is set to 8, the number Nd of the detection window units is set to 32, and the detection threshold amplitude is set to 1200 units, so as to ensure a proper false alarm probability.
The detection effect is as follows:
| number of samples | Number of successful samples detected | Probability of detection | |
| |
100 | 95 | 95％ |
| |
100 | 100 | 100％ |
Finally, it should be noted that: although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art will understand that various changes, modifications and substitutions can be made without departing from the spirit and scope of the invention as defined by the appended claims. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (8)
1. A millimeter wave radar abnormal signal detection method based on an analog transducer is characterized by comprising the following steps:
step S1: calculating correlation difference through Attention mechanism analysis-Attention, and uniformly modeling prior correlation and sequence correlation;
step S2: calculating symmetrical KL distances of prior association and sequence association of each layer to serve as a judgment basis of subsequent anomaly detection;
and step S3: distinguishing an abnormality judgment curve based on correlation analysis through visual analysis under different abnormality criteria;
and step S4: and (3) taking each sample as a group of related data, inputting the group of related data into an analog Transformer for training, and obtaining the normal working condition data monitoring model.
2. The method of claim 1, wherein the Attention mechanism anomally-Attention correlation formula is as follows:
3. the anomally transducer-based millimeter wave radar abnormal signal detection method according to claim 1, wherein a symmetric KL distance is calculated by the following formula:
4. the anomally transducer-based millimeter wave radar abnormal signal detection method according to claim 1, wherein the training method comprises the following steps:
step 1: manufacturing a training sample, building a TI vehicle-mounted millimeter wave radar data acquisition platform, and actually measuring a normal working condition data sample in an outdoor environment;
step 2: training an analog transform model training set, and introducing training set samples into the model in a vector form;
and step 3: and verifying the anomally Transformer model test set, artificially manufacturing interference data, and inputting the data into an anomally Transformer model to obtain an anomally Transformer millimeter wave radar abnormal signal detection model.
5. The anomally transducer-based millimeter wave radar abnormal signal detection system according to any one of claims 1 to 4, comprising a target result receiving module, a target reliability defining, modeling and analyzing module, a state monitoring and fault management module, an execution controller, a CAN network, a data labeling platform and an AI training platform;
the target result receiving module receives the image data of the camera and the signal data sent by the millimeter wave radar, preprocesses the image data and the signal data, prepares for the calculation of the reliability of the target result, and uploads the target result to the data annotation platform;
the target credibility definition, modeling and analysis module analyzes the received result through a machine learning algorithm, judges whether the result is fraudulent data, and realizes detection and defense of ADAS perception data;
the state monitoring and fault management module monitors and manages the sensing data and the reliability result in real time to realize real-time monitoring of fraud identification;
the execution controller is used for generating a defense instruction and blocking fraudulent attacks;
the CAN networks adopt a star structure, and are connected through a central gateway;
the data annotation platform performs pseudo label annotation through a generation model and provides a data basis for incremental training;
and the AI training platform performs incremental training on the reliability analysis model and optimizes model parameters.
6. The anomally transducer-based millimeter wave radar abnormal signal detection system as claimed in claim 5, wherein the operating frequency range of the millimeter wave radar is 30GHz to 300GHz, and the millimeter wave radar is used for detecting a target, measuring speed, measuring distance and measuring orientation.
7. The anomally transducer-based millimeter wave radar abnormal signal detection system of claim 6, wherein the millimeter wave radar performs signal windowing by using a hanning window in a detection process, and a formula is as follows:
8. the anomally transducer-based millimeter wave radar abnormal signal detection system as recited in claim 7, wherein the detection of the millimeter wave radar further involves a frequency domain conversion, and the frequency domain conversion formula is:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211405034.XA CN115840875A (en) | 2022-11-10 | 2022-11-10 | Millimeter wave radar abnormal signal detection method and system based on analog transducer |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211405034.XA CN115840875A (en) | 2022-11-10 | 2022-11-10 | Millimeter wave radar abnormal signal detection method and system based on analog transducer |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115840875A true CN115840875A (en) | 2023-03-24 |
Family
ID=85577070
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211405034.XA Pending CN115840875A (en) | 2022-11-10 | 2022-11-10 | Millimeter wave radar abnormal signal detection method and system based on analog transducer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115840875A (en) |
Citations (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110288005A (en) * | 2019-05-30 | 2019-09-27 | 南京理工大学 | Millimeter wave detector New Jamming Effects Evaluation Method based on SVM |
| CN110764064A (en) * | 2019-11-08 | 2020-02-07 | 哈尔滨工业大学 | Radar interference signal identification method based on deep convolutional neural network integration |
| CN111638488A (en) * | 2020-04-10 | 2020-09-08 | 西安电子科技大学 | Radar interference signal identification method based on LSTM network |
| CN112560596A (en) * | 2020-12-01 | 2021-03-26 | 中国航天科工集团第二研究院 | Radar interference category identification method and system |
| CN112904282A (en) * | 2021-01-20 | 2021-06-04 | 北京理工大学 | Radar interference signal identification method based on PWVD and convolutional neural network |
| WO2021189642A1 (en) * | 2020-03-25 | 2021-09-30 | 平安科技（深圳）有限公司 | Method and device for signal processing, computer device, and storage medium |
| CN113486898A (en) * | 2021-07-08 | 2021-10-08 | 西安电子科技大学 | Radar signal RD image interference identification method and system based on improved ShuffleNet |
| CN113486961A (en) * | 2021-07-12 | 2021-10-08 | 安徽耀峰雷达科技有限公司 | Radar RD image target detection method and system based on deep learning under low signal-to-noise ratio and computer equipment |
| CN113887356A (en) * | 2021-09-18 | 2022-01-04 | 四川启睿克科技有限公司 | Radar automatic identification method and device for low-altitude small target |
| CN113962151A (en) * | 2021-10-21 | 2022-01-21 | 中国人民解放军空军航空大学 | Intelligent distance decoy identification method based on deep convolution transfer learning |
| CN114217284A (en) * | 2021-12-03 | 2022-03-22 | 珠海安自达科技有限公司 | Radar moving target detection and interference suppression method and system based on characteristics |
| CN114386521A (en) * | 2022-01-14 | 2022-04-22 | 湖南师范大学 | Method, system, device and storage medium for detecting abnormality of time-series data |
| CN114429156A (en) * | 2022-01-21 | 2022-05-03 | 西安电子科技大学 | Radar interference multi-domain feature countermeasure learning and detection identification method |
| CN114550460A (en) * | 2022-04-25 | 2022-05-27 | 中国科学院自动化研究所 | Rail transit anomaly detection method and device and storage medium |
| CN114580461A (en) * | 2022-01-20 | 2022-06-03 | 西安电子科技大学 | Training method of radar interference recognition network and radar interference recognition method |
| WO2022142120A1 (en) * | 2020-12-31 | 2022-07-07 | 平安科技（深圳）有限公司 | Data detection method and apparatus based on artificial intelligence, and server and storage medium |
| CN114722857A (en) * | 2022-03-03 | 2022-07-08 | 西安电子科技大学 | HRRP (high resolution repeat) enhancement preprocessing method based on attention mechanism under interference condition |
| CN114761821A (en) * | 2019-11-28 | 2022-07-15 | 海拉有限双合股份公司 | Method for detecting interference in a radar system |
| CN114791593A (en) * | 2022-03-29 | 2022-07-26 | 北京扬铭科技发展有限责任公司 | Artificial intelligence signal analysis and recognition system |
| CN114818828A (en) * | 2022-05-18 | 2022-07-29 | 电子科技大学 | Training method of radar interference perception model and radar interference signal identification method |
| CN114895263A (en) * | 2022-05-26 | 2022-08-12 | 西安电子科技大学 | Radar active interference signal identification method based on deep migration learning |
-
2022
- 2022-11-10 CN CN202211405034.XA patent/CN115840875A/en active Pending
Patent Citations (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110288005A (en) * | 2019-05-30 | 2019-09-27 | 南京理工大学 | Millimeter wave detector New Jamming Effects Evaluation Method based on SVM |
| CN110764064A (en) * | 2019-11-08 | 2020-02-07 | 哈尔滨工业大学 | Radar interference signal identification method based on deep convolutional neural network integration |
| CN114761821A (en) * | 2019-11-28 | 2022-07-15 | 海拉有限双合股份公司 | Method for detecting interference in a radar system |
| WO2021189642A1 (en) * | 2020-03-25 | 2021-09-30 | 平安科技（深圳）有限公司 | Method and device for signal processing, computer device, and storage medium |
| CN111638488A (en) * | 2020-04-10 | 2020-09-08 | 西安电子科技大学 | Radar interference signal identification method based on LSTM network |
| CN112560596A (en) * | 2020-12-01 | 2021-03-26 | 中国航天科工集团第二研究院 | Radar interference category identification method and system |
| WO2022142120A1 (en) * | 2020-12-31 | 2022-07-07 | 平安科技（深圳）有限公司 | Data detection method and apparatus based on artificial intelligence, and server and storage medium |
| CN112904282A (en) * | 2021-01-20 | 2021-06-04 | 北京理工大学 | Radar interference signal identification method based on PWVD and convolutional neural network |
| CN113486898A (en) * | 2021-07-08 | 2021-10-08 | 西安电子科技大学 | Radar signal RD image interference identification method and system based on improved ShuffleNet |
| CN113486961A (en) * | 2021-07-12 | 2021-10-08 | 安徽耀峰雷达科技有限公司 | Radar RD image target detection method and system based on deep learning under low signal-to-noise ratio and computer equipment |
| CN113887356A (en) * | 2021-09-18 | 2022-01-04 | 四川启睿克科技有限公司 | Radar automatic identification method and device for low-altitude small target |
| CN113962151A (en) * | 2021-10-21 | 2022-01-21 | 中国人民解放军空军航空大学 | Intelligent distance decoy identification method based on deep convolution transfer learning |
| CN114217284A (en) * | 2021-12-03 | 2022-03-22 | 珠海安自达科技有限公司 | Radar moving target detection and interference suppression method and system based on characteristics |
| CN114386521A (en) * | 2022-01-14 | 2022-04-22 | 湖南师范大学 | Method, system, device and storage medium for detecting abnormality of time-series data |
| CN114580461A (en) * | 2022-01-20 | 2022-06-03 | 西安电子科技大学 | Training method of radar interference recognition network and radar interference recognition method |
| CN114429156A (en) * | 2022-01-21 | 2022-05-03 | 西安电子科技大学 | Radar interference multi-domain feature countermeasure learning and detection identification method |
| CN114722857A (en) * | 2022-03-03 | 2022-07-08 | 西安电子科技大学 | HRRP (high resolution repeat) enhancement preprocessing method based on attention mechanism under interference condition |
| CN114791593A (en) * | 2022-03-29 | 2022-07-26 | 北京扬铭科技发展有限责任公司 | Artificial intelligence signal analysis and recognition system |
| CN114550460A (en) * | 2022-04-25 | 2022-05-27 | 中国科学院自动化研究所 | Rail transit anomaly detection method and device and storage medium |
| CN114818828A (en) * | 2022-05-18 | 2022-07-29 | 电子科技大学 | Training method of radar interference perception model and radar interference signal identification method |
| CN114895263A (en) * | 2022-05-26 | 2022-08-12 | 西安电子科技大学 | Radar active interference signal identification method based on deep migration learning |
Non-Patent Citations (2)
| Title |
|---|
| JIEHUI XU: "ANOMALY TRANSFORMER: TIME SERIES ANOMALY DETECTION WITH ASSOCIATION DISCREPANCY", 《ICLR 2022》, 29 June 2022 (2022-06-29), pages 1 - 20 * |
| 杨杰, 陆正刚, 郭英凯: "基于数据融合和多重智能模型的目标识别和跟踪", 高技术通讯, no. 04, 28 April 1999 (1999-04-28) * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109212521B (en) | Target tracking method based on fusion of forward-looking camera and millimeter wave radar | |
| CN1940591B (en) | System and method of target tracking using sensor fusion | |
| JP3186401B2 (en) | Vehicle distance data processing device | |
| CN111899568B (en) | Bridge anti-collision early warning system, method and device and storage medium | |
| CN110163075A (en) | A kind of multi-information fusion method for diagnosing faults based on Weight Training | |
| CN111562570A (en) | Vehicle sensing method for automatic driving based on millimeter wave radar | |
| CN113156417B (en) | Anti-unmanned aerial vehicle detection system, method and radar equipment | |
| Akowuah et al. | Physical invariant based attack detection for autonomous vehicles: Survey, vision, and challenges | |
| CN111830501B (en) | HRRP history feature assisted signal fuzzy data association method and system | |
| Fujita et al. | An approach for intelligent evaluation of the state of complex autonomous objects based on the wavelet analysis | |
| CN109263649B (en) | Vehicle, object recognition method and object recognition system thereof in automatic driving mode | |
| CN110398735B (en) | Multi-radar-based perception data processing method and system | |
| CN117630860A (en) | Gesture recognition method of millimeter wave radar | |
| CN115932765B (en) | Radar failure automatic detection system and method based on multi-source data analysis | |
| CN115840875A (en) | Millimeter wave radar abnormal signal detection method and system based on analog transducer | |
| Ren et al. | Improved shape-based distance method for correlation analysis of multi-radar data fusion in self-driving vehicle | |
| CN116112872A (en) | Mining equipment personnel early warning method and device | |
| CN113866754B (en) | Moving target track association method based on Gaussian distribution wave gate | |
| Ning et al. | Radar Fault Detection via Camera-Radar Branches Learning Network | |
| CN114898202A (en) | Underwater video target scale space discriminant tracking system and method based on multi-model fusion | |
| Rykunov et al. | Multi-radar fusion for failure-tolerant vulnerable road users classification | |
| CN110412518A (en) | A kind of anti-hacker's interference unit attack device of intelligent automobile millimetre-wave radar | |
| Chen et al. | A Cluster-Based Weighted Feature Similarity Moving Target Tracking Algorithm for Automotive FMCW Radar | |
| CN117111019B (en) | Target tracking and monitoring method and system based on radar detection | |
| CN117113259B (en) | Coal mine state data processing method and system for predicting potential safety hazards |
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 |