CN112364753B - Method for detecting abnormal signal of low-voltage power carrier - Google Patents
Method for detecting abnormal signal of low-voltage power carrier Download PDFInfo
- Publication number
- CN112364753B CN112364753B CN202011237903.3A CN202011237903A CN112364753B CN 112364753 B CN112364753 B CN 112364753B CN 202011237903 A CN202011237903 A CN 202011237903A CN 112364753 B CN112364753 B CN 112364753B
- Authority
- CN
- China
- Prior art keywords
- layer
- signal
- classification model
- signals
- data
- 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
- 230000002159 abnormal effect Effects 0.000 title claims abstract description 21
- 238000000034 method Methods 0.000 title claims abstract description 19
- 238000013145 classification model Methods 0.000 claims abstract description 35
- 238000012549 training Methods 0.000 claims abstract description 29
- 238000004891 communication Methods 0.000 claims abstract description 28
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 20
- 238000012360 testing method Methods 0.000 claims abstract description 11
- 238000012795 verification Methods 0.000 claims abstract description 10
- 238000011176 pooling Methods 0.000 claims description 5
- 238000013461 design Methods 0.000 claims description 3
- 210000002569 neuron Anatomy 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 description 5
- 238000001514 detection method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- 239000000969 carrier Substances 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Cable Transmission Systems, Equalization Of Radio And Reduction Of Echo (AREA)
Abstract
The invention relates to a method for detecting abnormal signals of a low-voltage power carrier wave, which comprises the following steps: collecting signals, namely collecting current and voltage signal original data on a power supply power line of the protected equipment; training a classification model, establishing a training set, a verification set and a test set aiming at collected data, designing a convolutional neural network classification model, inputting the data set into the classification model, and training a signal classification model; inputting signal data to be detected into a trained signal discrimination model to judge the type of the signal; and displaying and alarming the identified communication signal type. The method meets the requirement of identifying the modulation mode of the power carrier communication by identifying the current and voltage signals in various different power carrier communication states.
Description
Technical Field
The invention relates to the technical field of power carrier detection, in particular to a method for detecting abnormal signals of a low-voltage power carrier.
Background
The use of information devices in modern information systems is becoming increasingly common. Electromagnetic waves excited by time-varying current signals generated by information equipment in operation are typical carrier waves, and once sensitive information is carried in the electromagnetic waves, the electromagnetic waves can be intercepted and restored, so that leakage of the sensitive information is caused. In recent years, breakthroughs are made in the technology of intercepting and restoring power line conduction coupling power leakage information, but power lines such as power lines and ground lines which have little attention in the past become potential paths for information leakage.
Preventing information leakage is an important task for enterprises and institutions, and it is required to provide a safe information protection means for enterprises and institutions. In important places such as fixed office environments and machine rooms, hidden danger of information disclosure may exist in a power line for supplying power to information equipment in time.
The existing device can only identify single modulation signals such as BPSK or FSK in a power carrier on a voltage power line, and is difficult to identify and distinguish under the condition that a plurality of communication signals coexist. And the carrier communication signal is difficult to extract, and the recognition rate of the communication signal is low. If deep learning technology can be utilized, by identifying the change of the low-voltage alternating-current power carrier current and voltage signals and identifying FSK, SSB, PSK, BPSK, QPSK, PAM and 16QAM multiple carrier communication signals existing on the power line, the effect of protecting power line information can be better achieved.
Disclosure of Invention
The invention relates to a method for detecting abnormal signals of a low-voltage power carrier wave, which is used for solving the problems in the prior art.
The invention discloses a method for detecting abnormal signals of a low-voltage power carrier, which comprises the following steps: collecting signals, namely collecting current and voltage signal original data on a power supply power line of the protected equipment; training a classification model, establishing a training set, a verification set and a test set aiming at collected data, designing a convolutional neural network classification model, inputting the data set into the classification model, and training a signal classification model; inputting signal data to be detected into a trained signal discrimination model to judge the type of the signal; and displaying and alarming the identified communication signal type.
According to one embodiment of the method for detecting abnormal signals of low-voltage power carriers, FSK, SSB, PSK, BPSK, QPSK, PAM and 16QAM carrier communication signals are respectively acquired when new data are acquired, and voltage and current signals under eight different conditions are respectively acquired under no carrier communication signals, and the categories of the signals are marked.
According to an embodiment of the method for detecting abnormal signals of a low-voltage power carrier wave, in the step of training the classification model, a convolutional neural network signal classification model is designed, and a convolutional layer and a pooling layer of the convolutional neural network are composed of a corresponding number of feature matrixes, wherein each two-dimensional feature comprises a plurality of independent neurons which are not connected with each other.
According to an embodiment of the present invention, the convolutional neural network layer design is divided into eight layers, the number of output data of the first layer is 63×63×16, the number of output data of the second layer is 31×31×16, the number of output data of the third layer is 29×29×32, the number of output data of the fourth layer is 29×29×32, the number of output data of the fifth layer is 14×14×32, the number of output data of the sixth layer is 6272, the number of output data of the seventh layer is 512, the number of output data of the eighth layer is 8, and the number of output data of the fourth layer corresponds to 8 signals to be identified.
According to one embodiment of the method for detecting the abnormal signal of the low-voltage power carrier, the collected current and voltage signals on the low-voltage alternating-current power line are used as the first-layer input of a trained convolutional neural network classification model, higher-level essential characteristics of data are extracted step by step through the classification model, and the signal type is identified and judged.
The method for detecting the abnormal signal of the power carrier wave meets the requirement of identifying the modulation mode of the power carrier wave communication by identifying the current voltage signals under various different power carrier wave communication states.
Drawings
FIG. 1 is a flow chart of the detection of an abnormal signal of a power supply carrier according to the present invention;
FIG. 2 is a schematic diagram showing the constitution of a device for detecting abnormal signal of a power supply carrier according to the present invention;
fig. 3 is a schematic diagram of the components of the power source carrier abnormal signal acquisition module according to the present invention.
Detailed Description
For the purposes of clarity, content, and advantages of the present invention, a detailed description of the embodiments of the present invention will be described in detail below with reference to the drawings and examples.
The invention provides a device for detecting abnormal signals of a low-voltage power carrier, which comprises three parts: the system comprises an acquisition module, a calculation module and an alarm module.
Fig. 1 is a flowchart of detecting an abnormal signal of a low-voltage power carrier according to the present invention, and fig. 2 is a schematic diagram of a device for detecting an abnormal signal of a low-voltage power carrier according to the present invention, as shown in fig. 1 and 2, the operation flow is divided into four steps: the method comprises the steps of collecting data by an acquisition module, training a convolutional neural network by a calculation module to establish a signal discrimination and classification model, judging the type of communication signals by the calculation module through the signal discrimination model, and outputting a judging result by an alarm module when abnormal communication signals exist.
And the acquisition module is used for: the function is to collect the current and voltage signal data on the low-voltage power line. In specific implementation, the acquisition module is composed of a voltage transformer, a current transformer, a filter circuit, a bias circuit and an AD converter. As shown in fig. 3.
The calculation module: the function is to train convolutional neural network, build signal classification model, through the model that trains, can judge the class of signal. In specific implementation, the computing module can adopt a conventional micro industrial control PC.
And an alarm module. The function is to display the signal type result, when the detected communication signal is used for warning, a buzzer can be selected for warning, and a common display is selected for signal type display
The invention provides a method for detecting an abnormal signal of a low-voltage power carrier, which comprises the following steps:
(1) And (3) signal acquisition: a signal acquisition module is arranged on a power supply power line of the protection equipment to acquire current and voltage signal data;
(2) Training a signal classification model: aiming at the collected data, a training set, a verification set and a test set are established, a convolutional neural classification training network is designed, the training data set is input into the training set, a classification model of signals is obtained through training, the verification set is utilized for verification, and finally the test set is utilized for testing;
(3) Judging signal category: and inputting the signal data to be detected into a trained classification model, and judging the signal type.
(4) Alarm display: and displaying the type of the identified communication signal and alarming.
For another specific embodiment, the method for detecting the abnormal signal of the low-voltage power carrier comprises the following 4 steps:
(1) And (3) signal acquisition: installing a signal acquisition module on a power supply power line of the protected equipment, and acquiring current and voltage signal original data;
In the specific implementation, in order to train the signal classification model, voltage and current signals under eight conditions of no carrier communication signal, FSK, SSB, PSK, BPSK, QPSK, PAM and 16QAM carrier communication are acquired respectively and are marked manually to be used as training data.
(2) Training a signal classification model: establishing a training set, a verification set and a test set aiming at the acquired data, inputting the training data set with the label into a training signal classification model by using a convolutional neural network, verifying by using the verification set, and finally testing by using the test set;
the specific implementation method comprises the following steps: the convolution layer and the pooling layer of the convolution neural network are composed of a corresponding number of feature matrices, and each two-dimensional feature comprises a plurality of independent neurons which are not connected with each other. The convolutional neural network layer design is divided into eight layers of network architecture, the number of output data of a first layer is 63 x 16, the number of output data of a second layer is 31 x 16, the number of output data of a third layer is 29 x 32, the number of output data of a fourth layer incomplete connection layer is 29 x 32, the number of output data of a fifth layer is 14 x 32, the number of output data of a sixth layer full connection layer is 6272, the number of output data of a seventh layer full connection layer is 512, the number of output data of an eighth layer full connection layer is 8, and 8 signals to be identified are correspondingly needed.
Firstly, inputting a training data set into a convolutional neural network, and realizing automatic matching of the same characteristics of the same kind of signals by comparing the similarity of characteristic vectors; and fine-tuning the characteristics of the power carrier signal from top to bottom by using the class label information of the training data. And inputting the verification set into the trained convolutional neural network for verification. And finally, testing the convolutional neural network model by using the test set.
(3) Judging signal category: and inputting the signal data to be detected into a trained classification model, and judging the signal type.
In the specific implementation, the current and voltage signals on the low-voltage alternating-current power line collected by the collection module are directly used as the first-layer input of the trained convolutional neural network classification model, the higher-level essential characteristics of the data are extracted step by step through the classification model, after the characteristics are extracted, the similarity of the characteristic vectors is compared, and the matching of the same characteristics of the signals to be detected and the training signals is realized, so that the signal type is identified and judged.
(4) Alarm display: and displaying and alarming the identified communication signal type.
The specific implementation method comprises the following steps: and displaying the signal type of the communication signal judged by the signal type judging module through the alarm display module and giving an alarm.
The invention discloses a method for detecting abnormal signals of a low-voltage power carrier. The device comprises three parts: (1) an acquisition module; (2) a computing module; (3) an alarm module. The operation is divided into four steps: (1) signal acquisition: collecting data by using a collecting module; (2) training a signal classification model: training a neural network by using a computing module to establish a signal classification model; (3) discriminating the signal category: judging the communication signal category by a calculation module through a signal classification model; (4) alarm display: and displaying the detected communication signal by using an alarm module and giving an alarm. The invention is applied to a leakage-proof safety system for the power carrier information, can rapidly and accurately identify various power carrier communication signals, reduces the risk of leakage information of important information equipment through a power line, and improves the safety of the information equipment.
Compared with the existing power carrier communication signal detection method, the method has the advantages that the convolutional neural network signal classification model is utilized, more modulation types of power carrier communication signals can be identified, and the types of the signals can be identified more. Therefore, the invention plays an important role in the application of preventing information leakage on the power line.
The foregoing is merely a preferred embodiment of the present invention, and it should be noted that modifications and variations could be made by those skilled in the art without departing from the technical principles of the present invention, and such modifications and variations should also be regarded as being within the scope of the invention.
Claims (1)
1. The method for detecting the abnormal signal of the power supply carrier is characterized by comprising the following steps of:
collecting signals, namely collecting current and voltage signal original data on a power supply power line of the protected equipment;
Training a classification model, establishing a training set, a verification set and a test set aiming at collected data, designing a convolutional neural network classification model, inputting the data set into the classification model, and training a signal classification model;
inputting signal data to be detected into a trained signal discrimination model to judge the type of the signal;
Displaying and alarming the identified communication signal type;
Wherein,
When training data are collected, FSK, SSB, PSK, BPSK, QPSK, PAM and 16QAM carrier communication signals are collected respectively, and under no carrier communication signals, voltage and current signals under eight different signals are collected, and the types of the signals are marked;
In the step of training the classification model, a convolutional neural network signal classification model is designed, a convolutional layer and a pooling layer of the convolutional neural network are formed by a corresponding number of feature matrixes, and each two-dimensional feature comprises a plurality of independent neurons which are not connected with each other;
The convolutional neural network layer design is divided into eight layers of network architecture, the number of output data of a first layer of convolutional layer is 63 x 16, the number of output data of a second layer of pooling layer is 31 x 16, the number of output data of a third layer of convolutional layer is 29 x 32, the number of output data of a fourth layer of incomplete connection layer is 29 x 32, the number of output data of a fifth layer of pooling layer is 14 x 32, the number of output data of a sixth layer of full connection layer is 6272, the number of output data of a seventh layer of full connection layer is 512, the number of output data of an eighth layer of full connection layer is 8, and 8 signals to be identified are correspondingly needed;
The collected current and voltage signals on the low-voltage alternating-current power line are used as the first-layer input of a trained convolutional neural network classification model, and the higher-level essential characteristics of the data are extracted step by step and layer by layer through the classification model, so that the signal types are identified and judged.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011237903.3A CN112364753B (en) | 2020-11-09 | 2020-11-09 | Method for detecting abnormal signal of low-voltage power carrier |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011237903.3A CN112364753B (en) | 2020-11-09 | 2020-11-09 | Method for detecting abnormal signal of low-voltage power carrier |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112364753A CN112364753A (en) | 2021-02-12 |
| CN112364753B true CN112364753B (en) | 2024-06-28 |
Family
ID=74510171
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011237903.3A Active CN112364753B (en) | 2020-11-09 | 2020-11-09 | Method for detecting abnormal signal of low-voltage power carrier |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112364753B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104935359A (en) * | 2015-05-12 | 2015-09-23 | 国网重庆市电力公司电力科学研究院 | Low voltage power line carrier communication signal modulation mode identification device and system |
| CN108616470A (en) * | 2018-03-26 | 2018-10-02 | 天津大学 | Modulation Signals Recognition method based on convolutional neural networks |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020130768A1 (en) * | 2001-03-14 | 2002-09-19 | Hongyuan Che | Low voltage power line carrier communications at fundamental working frequency |
| DE102004017548B4 (en) * | 2004-04-07 | 2007-12-27 | Medav Gmbh | Method for detecting and analyzing radio signals |
| JP5546989B2 (en) * | 2010-07-30 | 2014-07-09 | 一般財団法人 関西電気保安協会 | Electric equipment abnormality factor estimation device, computer program, electric equipment abnormality factor estimation method, and electric equipment abnormality factor estimation system |
| CN110855591B (en) * | 2019-12-09 | 2021-10-29 | 山东大学 | QAM and PSK signal intra-class modulation classification method based on convolutional neural network structure |
| CN111092836B (en) * | 2019-12-13 | 2022-05-17 | 中国人民解放军空军工程大学 | Signal modulation mode identification method and device |
-
2020
- 2020-11-09 CN CN202011237903.3A patent/CN112364753B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104935359A (en) * | 2015-05-12 | 2015-09-23 | 国网重庆市电力公司电力科学研究院 | Low voltage power line carrier communication signal modulation mode identification device and system |
| CN108616470A (en) * | 2018-03-26 | 2018-10-02 | 天津大学 | Modulation Signals Recognition method based on convolutional neural networks |
Non-Patent Citations (1)
| Title |
|---|
| 电力载波异常信号检测***设计;李雨航;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;C042-2679 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112364753A (en) | 2021-02-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US5406260A (en) | Network security system for detecting removal of electronic equipment | |
| CN102095926B (en) | System for automatically detecting power system configuration | |
| CN203241494U (en) | Direct current system AC series current fault and insulation fault monitoring line-selection apparatus | |
| CN103135038A (en) | Alternative current electric crosstalk and insulation fault monitoring line selection method of direct current system | |
| US4707679A (en) | Magnetic tamper detector | |
| US20070132503A1 (en) | Power patch panel with guided mac capability | |
| CN106133533A (en) | The measurement equipment of at least one physical quantity of electric utility | |
| CN103548304A (en) | System and method of detecting and locating intermittent and other faults | |
| CN106789386A (en) | The method and the error detector for network system of mistake on detection communication bus | |
| CN109827652A (en) | One kind being directed to Fibre Optical Sensor vibration signal recognition and system | |
| CN111797725A (en) | Equipment state detection method, device and system | |
| CN108320235B (en) | Monitoring management system | |
| CN206322201U (en) | Transformer substation grounding wire real-time monitoring system | |
| CN112364753B (en) | Method for detecting abnormal signal of low-voltage power carrier | |
| CN104237716A (en) | Serial connector detection system and method | |
| CN110031719A (en) | A kind of circuit detecting method and equipment based on Intelligent electric leakage monitoring miniature circuit breaker | |
| CN105425056A (en) | Static protection system and method with personnel state identification | |
| CN103413105A (en) | Device for protecting shell of information safety device | |
| CN108731731A (en) | A kind of lighning proof type safety supervision system and lighning proof type safety supervision method | |
| KR102272863B1 (en) | Apparatus and method for detecting abnormal status | |
| CN117294004A (en) | Intelligent identification system and method for relay protection operation mode of transformer substation | |
| CN207200894U (en) | Tooling plate and test system | |
| CN108683639A (en) | A kind of computer network abnormality detection and automatic repair system, method and mobile terminal | |
| CN107067700A (en) | Gather the methods, devices and systems of power information | |
| Sidhu et al. | Arcing fault detection using artificial neural networks |
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 |