CN116502071A - Key signal detection system and method - Google Patents
Key signal detection system and method Download PDFInfo
- Publication number
- CN116502071A CN116502071A CN202310752613.XA CN202310752613A CN116502071A CN 116502071 A CN116502071 A CN 116502071A CN 202310752613 A CN202310752613 A CN 202310752613A CN 116502071 A CN116502071 A CN 116502071A
- Authority
- CN
- China
- Prior art keywords
- satellite
- satellite signal
- signal
- key
- signals
- 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.)
- Granted
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 92
- 238000000034 method Methods 0.000 title claims abstract description 18
- 239000011159 matrix material Substances 0.000 claims abstract description 57
- 238000012549 training Methods 0.000 claims abstract description 37
- 230000008034 disappearance Effects 0.000 claims abstract description 36
- 230000002787 reinforcement Effects 0.000 claims abstract description 30
- 238000004458 analytical method Methods 0.000 claims abstract description 23
- 239000000284 extract Substances 0.000 claims abstract description 8
- 230000009471 action Effects 0.000 claims description 40
- 239000013598 vector Substances 0.000 claims description 22
- 230000006870 function Effects 0.000 claims description 17
- 238000010276 construction Methods 0.000 claims description 14
- 238000012937 correction Methods 0.000 claims description 11
- 238000012795 verification Methods 0.000 claims description 8
- 238000000605 extraction Methods 0.000 claims description 6
- 238000007667 floating Methods 0.000 claims description 5
- 230000008569 process Effects 0.000 claims description 5
- 238000004422 calculation algorithm Methods 0.000 abstract description 7
- 238000004891 communication Methods 0.000 abstract description 6
- 239000004973 liquid crystal related substance Substances 0.000 description 14
- 238000010586 diagram Methods 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 4
- 238000003066 decision tree Methods 0.000 description 3
- 238000002372 labelling Methods 0.000 description 3
- 238000005291 chaos (dynamical) Methods 0.000 description 2
- 239000003795 chemical substances by application Substances 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000001944 accentuation Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000010287 polarization Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/14—Relay systems
- H04B7/15—Active relay systems
- H04B7/185—Space-based or airborne stations; Stations for satellite systems
- H04B7/1851—Systems using a satellite or space-based relay
- H04B7/18519—Operations control, administration or maintenance
-
- 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
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- 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
- G06N3/092—Reinforcement learning
-
- 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
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Astronomy & Astrophysics (AREA)
- Computational Linguistics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Aviation & Aerospace Engineering (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Radio Relay Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of satellite communication systems, and discloses a key signal detection system and a method, wherein the system comprises a satellite signal acquisition module; a satellite signal characteristic analysis module; a key signal detection module; the satellite signal acquisition module acquires historical satellite signals and satellite signals to be detected; the satellite signal characteristic analysis module extracts a satellite signal characteristic matrix; the key signal detection module performs deep reinforcement learning model training by using a satellite signal feature matrix of the satellite signal to be detected, and performs decision classification tree training by using a satellite signal feature matrix of the historical satellite signal so as to detect whether the satellite signal has the appearance or disappearance of the key signal. According to the invention, the satellite signal feature matrix is used for carrying out deep reinforcement learning model training and decision classification tree training, so that the occurrence or disappearance of the important signals is identified, and the technical problems of high requirements on hardware and algorithm, large error, low reasoning speed and the like in the existing important signal detection are solved.
Description
Technical Field
The invention relates to the technical field of satellite communication, in particular to a key signal detection system and a method.
Background
Satellite communication plays a vital role in the modern communication field, and it can realize global coverage, high-speed transmission, and reliable communication services.
For signal detection and identification, the prior art comprises a key signal detection system based on a chaos theory, a key signal detection system based on digital signal processing and a key signal detection system based on machine learning. However, the above technology has certain drawbacks, such as that the key signal detection system based on digital signal processing has high requirements on hardware and algorithm, requires a large amount of calculation and storage resources, and may have problems of misjudgment and missed judgment. The key signal detection system based on the chaos theory is sensitive to initial conditions, and the problem of large recognition errors may exist. Currently, there are systems that detect signals using deep reinforcement learning. The method uses a deep reinforcement learning method to classify the time-frequency diagram of the signal so as to identify the target signal, but uses the image as data can cause the model reasoning speed to be too slow, so that the target signal cannot be found timely.
Therefore, how to reduce the resource requirement of key signal detection, reduce the detection error, and improve the detection speed is a technical problem that needs to be solved.
Disclosure of Invention
The invention mainly aims to provide a key signal detection system and a key signal detection method, and aims to solve the technical problems that the current key signal detection has higher requirements on hardware and algorithm, larger error, low reasoning speed and the like.
In order to achieve the above object, the present invention provides a key signal detection system, including:
a satellite signal acquisition module;
a satellite signal characteristic analysis module;
a key signal detection module;
the satellite signal acquisition module acquires satellite signals, wherein the satellite signals comprise historical satellite signals and satellite signals to be detected;
the satellite signal characteristic analysis module extracts a historical satellite signal and a satellite signal characteristic matrix of a satellite signal to be detected;
the key signal detection module performs deep reinforcement learning model training by using a satellite signal feature matrix of a satellite signal to be detected, and performs decision classification tree training by using a satellite signal feature matrix of a historical satellite signal so as to detect whether the satellite signal has the appearance or disappearance of the key signal.
Optionally, the satellite signal feature analysis module has:
an extraction unit;
a coding unit;
the extraction unit extracts satellite signal characteristics of historical satellite signals and satellite signals to be detected;
the encoding unit encodes non-floating point characteristics in the extracted satellite signal characteristics.
Optionally, the satellite signal feature analysis module further includes:
the satellite signal feature set construction module;
the satellite signal feature matrix construction module;
the satellite signal feature set construction module constructs a satellite signal feature set corresponding to each time feature according to the extracted satellite signal features;
the satellite signal feature matrix construction module constructs a satellite signal feature matrix comprising each satellite signal feature group according to the constructed satellite signal feature groups, and the satellite signal feature groups in the satellite signal feature matrix are ordered according to the corresponding time features.
Optionally, the key signal detection module includes:
a deep reinforcement learning unit;
the deep reinforcement learning unit performs deep reinforcement learning model training according to the acquired satellite signal feature matrixes of the satellite signals to be detected, takes each satellite signal feature group in the satellite signal feature matrixes as a state input, and acquires action execution vector groups of the satellite signal feature matrixes;
the motion execution vector group comprises motion execution vectors corresponding to each satellite signal feature group, and the motion execution vectors comprise motion action1 and non-motion action2.
Optionally, the loss function trained by the deep reinforcement learning model is specifically:
wherein, the liquid crystal display device comprises a liquid crystal display device,and->For network parameters in the training process, +.>For discounts factor->In order to observe the bonus function,representing the current state and the current action, < >>For the next state, ++>For optimal action, Q is the Q function.
Optionally, the satellite signal feature matrix of the historical satellite signal has a labeling field, where the labeling field characterizes whether the satellite signal feature matrix of the current historical satellite signal has an action of alternately vanishing the key signal.
Optionally, the key signal detection module further includes:
a decision classification tree unit;
the decision classification tree unit performs decision classification tree training by using the satellite signal feature matrix of the historical satellite signals, and judges whether the satellite signal feature matrix of the satellite signals to be detected has actions of alternating key signals appearing and disappearing by using the trained decision classification tree.
Optionally, the key signal detection module further includes:
a key signal appearance and disappearance judging unit;
the key signal appearance and disappearance judging unit is used for judging whether the key signal appears or disappears when the satellite signal feature matrix of the satellite signal to be detected has the action of alternately appearing and disappearing the key signal: judging whether the motion execution vector of the satellite signal feature group has motion action1 or not, when the motion execution vector has motion action1, the key signal appears in the first motion action1 and disappears in the last motion action 1;
the key signal appearance and disappearance judging unit is used for judging whether the key signal appearance and disappearance are alternated when the satellite signal feature matrix of the satellite signal to be detected does not have the action of the key signal appearance and disappearance alternation: judging whether the previous action of the first action1 is action2 or not, if so, generating an important signal; if not, the action execution vector of the satellite signal feature set has action2, and the key signal disappears.
Optionally, the key signal detection system further comprises:
a manual checking module;
and the manual verification module generates correction data according to the detection result of the key signal detection module, and performs decision classification tree training in the key signal detection module by using the correction data.
In addition, in order to achieve the above object, the present invention also provides a key signal detection method, which includes the steps of:
acquiring satellite signals, wherein the satellite signals comprise historical satellite signals and satellite signals to be detected;
extracting a satellite signal characteristic matrix of a historical satellite signal and a satellite signal to be detected;
and training a deep reinforcement learning model by using a satellite signal characteristic matrix of the satellite signal to be detected, and training a decision classification tree by using a satellite signal characteristic matrix of the historical satellite signal to detect whether the satellite signal has the occurrence or the disappearance of an important signal.
The embodiment of the invention provides a key signal detection system and a method, wherein the system comprises a satellite signal acquisition module; a satellite signal characteristic analysis module; a key signal detection module; the satellite signal acquisition module acquires satellite signals, wherein the satellite signals comprise historical satellite signals and satellite signals to be detected; the satellite signal characteristic analysis module extracts a historical satellite signal and a satellite signal characteristic matrix of the satellite signal to be detected; the key signal detection module performs deep reinforcement learning model training by using a satellite signal feature matrix of the satellite signal to be detected, and performs decision classification tree training by using a satellite signal feature matrix of the historical satellite signal so as to detect whether the satellite signal has the appearance or disappearance of the key signal. According to the invention, the deep reinforcement learning model training and the decision classification tree training are carried out through the extracted satellite signal feature matrixes of the satellite signals to be detected and the historical satellite signals, and the occurrence or disappearance of the important signals is identified, so that the technical problems of high requirements on hardware and algorithms, large errors, low reasoning speed and the like in the existing key signal detection are solved.
Drawings
FIG. 1 is a schematic diagram of a first embodiment of a key signal detection system according to the present invention;
FIG. 2 is a schematic diagram of a second embodiment of a key signal detection system according to the present invention;
fig. 3 is a schematic flow chart of a first embodiment of a key signal detection method according to the present invention;
fig. 4 is a schematic flow chart of a second embodiment of a key signal detection method provided by the present invention.
Reference numerals:
10-a satellite signal acquisition module; 20-a satellite signal characteristic analysis module; 30-an important signal detection module; and 40, a manual verification module.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
An embodiment of the present invention provides a key signal detection system, referring to fig. 1, fig. 1 is a schematic diagram of a first embodiment of the key signal detection system of the present invention.
The key signal detection system provided by the embodiment of the invention comprises a satellite signal acquisition module 10, a satellite signal characteristic analysis module 20 and a key signal detection module 30.
It should be noted that, the satellite signal acquisition module 10 acquires satellite signals, where the satellite signals include a historical satellite signal and a satellite signal to be detected; the satellite signal characteristic analysis module 20 extracts a satellite signal characteristic matrix of the historical satellite signals and the satellite signals to be detected; the key signal detection module 30 performs deep reinforcement learning model training by using the satellite signal feature matrix of the satellite signal to be detected, and performs decision classification tree training by using the satellite signal feature matrix of the historical satellite signal, so as to detect whether the satellite signal has the presence or absence of the key signal.
The satellite signal acquisition module 10 is responsible for acquiring satellite signals; the satellite signal characteristic analysis module 20 is responsible for analyzing the characteristics of the acquired signals and generating prediction data; the accent signal detection module 30 is responsible for finding accent signals, and detecting whether accent signals are present or absent in the current signal segment.
In a preferred embodiment, the satellite signal feature analysis module 20 has: an extraction unit and an encoding unit.
The extracting unit extracts satellite signal characteristics of the historical satellite signals and the satellite signals to be detected; the encoding unit encodes non-floating point features in the extracted satellite signal features.
Wherein, the satellite signal acquisition module 10 acquires the satelliteThe signal data is expressed asThe satellite signal characteristic analysis module 20 will be for +.>And performing characteristic analysis. The feature analysis specifically comprises: the extraction unit extracts->The encoding unit encodes the non-floating point feature in the extracted satellite signal features.
In practical applications, satellite signal characteristics may include: carrier frequencySymbol rate->Bandwidth->Pulse width->Signal power->Polarization mode->Modulation schemeFrame structure->Data Rate->Status of automatic gain control->JumpingFrequency pattern->Doppler Effect->Received signal to noise ratio->Bit error rate->Frequency deviation->Channel type->Signal amplitude modulation index->Signal frequency modulation index->Phase offset->Delay->Time +.>。
In practical applications, when encoding non-floating point data, the LabelEncoder may be used to encode the binary type feature values, i.e., discrete values or text, such as features: channel type。
In a preferred embodiment, the satellite signal feature analysis module 20 further has: the system comprises a satellite signal feature group construction module and a satellite signal feature matrix construction module.
The satellite signal feature set construction module constructs a satellite signal feature set corresponding to each time feature according to the extracted satellite signal features; the satellite signal feature matrix construction module constructs a satellite signal feature matrix comprising each satellite signal feature group according to the constructed satellite signal feature groups, and the satellite signal feature groups in the satellite signal feature matrix are ordered according to the corresponding time features.
In practical application, the satellite signal feature set constructed by the satellite signal feature set construction module may be expressed as:
。
in practical application, the satellite signal feature matrix constructed by the satellite signal feature matrix construction module may be expressed as:
wherein, the liquid crystal display device comprises a liquid crystal display device,by->According to unit time->Sequentially composed of (I)>，/>Is the acquisition step size of the satellite signal acquisition module 10.
In a preferred embodiment, the accent signal detection module 30 has: and a deep reinforcement learning unit.
The deep reinforcement learning unit performs deep reinforcement learning model training according to the acquired satellite signal feature matrixes of the satellite signals to be detected, takes each satellite signal feature group in the satellite signal feature matrixes as a state input, and acquires action execution vector groups of the satellite signal feature matrixes; the motion execution vector group comprises motion execution vectors corresponding to each satellite signal feature group, and the motion execution vectors comprise motion action1 and motion action2.
In this embodiment, the deep reinforcement learning model of the deep reinforcement learning unit adopts the idea of a deep Q neural network (DQN) that represents the Q function using the deep neural network based on the conventional RL method Q learning. The situation that the traditional RL is not well applied when facing continuous motion or large state space is avoided. Q learning is representative of the algorithm based on a value function in RL. For the Markov Decision Process (MDP), the Agent learning strategy is。/>The value function Q indicates that the Agent expects to pass through according to the status +.>Learning strategy->Execution of action->Obtain a jackpot, learn strategy->The Q function of (c) can be expressed as:
（1）
wherein, the liquid crystal display device comprises a liquid crystal display device,and->Representing the current state and the current action, respectively, +.>Representing from->Time to->The cumulative rewards obtained are expressed as follows:
（2）
wherein, the liquid crystal display device comprises a liquid crystal display device,for discounts factor->To observe the bonus function, the bonus function represents a status from +.>Become->The resulting prize. For equation (1), the iterative process of the Q function can be represented by the Bellman optimal equation:
（3）
wherein, the liquid crystal display device comprises a liquid crystal display device,for state set +.>For the next state, ++>For optimal action +.>Is->By->Becomes Q function value, < >>Representing the current state and current action. />The update procedure of (2) can be expressed as:
（4）
wherein, the liquid crystal display device comprises a liquid crystal display device,is the learning rate, and the algorithm obtains the optimal Q function after continuously updating the Q function. Since the signal of each time of the satellite can be used as a state, the problem that the conventional Q learning cannot be stored exists because the conventional Q learning is represented by using a Q table. To solve this problem, the Q function is represented by a deep neural network, which is a convolutional neural network ending in a fully-connected layer, the output of which is a score for each action, the input of which is state->The larger the output value is, the more in the current state +.>Executing the action down causes the state to become +.>Is more preferable. The significance of DQN is that the deep neural network is parameter updated using the idea of Q learning, and the loss function is defined in this embodiment as follows：
（5）
Wherein, the liquid crystal display device comprises a liquid crystal display device,and->Is a network parameter in the training process.
In the present embodiment, training of the deep reinforcement learning modelIs +.>For input, i.e. each +.>All are in a state, action->There are two kinds, one is to judge the state as an important signal action1, the other is an non-important signal action2, the important signal is entered through the non-important signal through the action1 or the non-important signal is entered through the important signal through the action2, the rewarding value is 1, and the other is 0.
In a preferred embodiment, the satellite signal feature matrix of the historical satellite signal has a labeling field that characterizes whether the satellite signal feature matrix of the current historical satellite signal has an action of alternating the occurrence and disappearance of the accent signal.
In practical applications, the historical satellite signals may be expressed as:wherein the label field->For manually marked group-trunk, i.e. in +.>Whether or not there is an alternation of the occurrence and disappearance of the accentuation signal (occurrence +.>1, no occurrence of0).
On this basis, the accent signal detection module 30 further has: and a decision classification tree unit.
It should be noted that, the decision tree unit performs decision tree training by using the satellite signal feature matrix of the historical satellite signal, and determines whether the satellite signal feature matrix of the satellite signal to be detected has an action of alternately disappearance of key signals by using the trained decision tree.
In practical applications, the deep reinforcement learning unit may determine the occurrence or disappearance of the signal, but may not determine the simultaneous occurrence and disappearance of the signal. In this embodiment, the model is enhanced in combination with an HPC that selects a decision classification tree. UsingTraining the decision classification tree, and classifying whether the key signals appear to disappear and alternate (appear +.>1, no->0).
In a preferred embodiment, the accent signal detecting module 30 further has: and an emphasis signal appearance and disappearance judging unit.
It should be noted that, when the satellite signal feature matrix of the satellite signal to be detected has the action of alternating the occurrence and disappearance of the key signals, the key signal occurrence and disappearance judging unit: judging whether the motion execution vector of the satellite signal feature group has motion action1 or not, when the motion execution vector has motion action1, the key signal appears in the first motion action1 and disappears in the last motion action 1; the key signal appearance and disappearance judging unit is used for judging whether the key signal appearance and disappearance are alternated when the satellite signal feature matrix of the satellite signal to be detected does not have the action of the key signal appearance and disappearance alternation: judging whether the previous action of the first action1 is action2 or not, if so, generating an important signal; if not, the action execution vector of the satellite signal feature set has action2, and the key signal disappears.
In practical application, the final output result of the deep reinforcement learning model is thatMotion execution vector of (a). Judging the occurrence and disappearance of the key signal by combining the classification result of the HPC. When->When 1, judge +.>If there is an action1, when there is an action1, it indicates that the key signal appears in the first action1 and disappears in the last action 1. When->When the action is 0, judging the appearance and disappearance by judging the relation before and after the action, and when the action before the first action1 is action2, representing that an important signal appears; when the previous action of the first action1 is action1 and action2 exists, the key signal disappears.
In a preferred embodiment, the method can also be carried out byThe time information contained in the key signal is extracted through the position where the action occurs, so that the key signal appearance law can be conveniently analyzed by staff.
The present embodiment provides a key signal detection system that uses the idea of DQN to identify the occurrence or disappearance of a key signal, and on the basis of this, the alternate process of "occurrence or disappearance" of a key signal is also identified in combination with HPC. On the other hand, a non-image method is used, and only common signal features are used for identification through a deep reinforcement learning model, so that the reasoning speed is improved. And because the adopted signal characteristics contain time, the time information can be extracted through the position where the action occurs, so that the analysis of the occurrence rule of the key signal by the staff is convenient.
An embodiment of the present invention provides a key signal detection system, referring to fig. 2, fig. 2 is a schematic diagram of a second embodiment of the key signal detection system of the present invention.
The key signal detection system according to the embodiment of the present invention further includes: and a manual verification module 40.
It should be noted that, the manual verification module 40 generates correction data according to the detection result of the key signal detection module 30, and performs decision classification tree training in the key signal detection module 30 by using the correction data.
The manual verification module 40 verifies the analysis result and corrects the error data, and further trains the model by using the manually verified data as correction data.
The present embodiment provides a key signal detection system, which can utilize correction data to correct and train the decision classification tree unit in the key signal detection module 30 through the addition of the manual verification module 40, so as to improve the accuracy of model detection.
The embodiment of the invention provides a key signal detection method, and referring to fig. 3, fig. 3 is a schematic flow chart of the key signal detection method embodiment of the invention.
As shown in fig. 3, the key signal detection method according to the embodiment of the present invention is based on the key signal detection system provided in any of the foregoing embodiments, and includes the following steps:
s100: acquiring satellite signals, wherein the satellite signals comprise historical satellite signals and satellite signals to be detected;
s200: extracting a satellite signal characteristic matrix of a historical satellite signal and a satellite signal to be detected;
s300: and training a deep reinforcement learning model by using a satellite signal characteristic matrix of the satellite signal to be detected, and training a decision classification tree by using a satellite signal characteristic matrix of the historical satellite signal to detect whether the satellite signal has the occurrence or the disappearance of an important signal.
The embodiment provides a key signal detection method, which carries out deep reinforcement learning model training and decision classification tree training through the extracted satellite signal feature matrixes of the satellite signals to be detected and the historical satellite signals, identifies the occurrence or disappearance of the key signals, and solves the technical problems that the current key signal detection has higher requirements on hardware and algorithm, larger errors, low reasoning speed and the like.
Other embodiments or specific implementation manners of the key signal detection method of the present invention may refer to the above system embodiments, and are not described herein.
An embodiment of the present invention provides a key signal detection method, referring to fig. 4, and fig. 4 is a schematic flow chart of a second embodiment of the key signal detection method of the present invention.
The key signal detection method provided by the embodiment of the invention is based on the previous embodiment, and further comprises the following steps:
s400: and generating correction data according to the detection result of the key signal detection module, and executing decision classification tree training in the key signal detection module by using the correction data.
The embodiment provides a key signal detection method, which can utilize correction data to correct and train a key signal detection module through the addition of a manual verification module, and improves the accuracy of model prediction and recognition.
Other embodiments or specific implementation manners of the key signal detection method of the present invention may refer to the above system embodiments, and are not described herein.
In describing embodiments of the present invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "center", "top", "bottom", "inner", "outer", "inside", "outside", etc. indicate orientations or positional relationships based on the drawings are merely for convenience in describing the present invention and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a specific orientation, be configured and operated in a specific orientation, and thus should not be construed as limiting the present invention. Wherein "inside" refers to an interior or enclosed area or space. "peripheral" refers to the area surrounding a particular component or region.
In the description of embodiments of the present invention, the terms "first," "second," "third," "fourth" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first", "a second", "a third" and a fourth "may explicitly or implicitly include one or more such feature. In the description of the present invention, unless otherwise indicated, the meaning of "a plurality" is two or more.
In describing embodiments of the present invention, it should be noted that the terms "mounted," "connected," and "assembled" are to be construed broadly, as they may be fixedly connected, detachably connected, or integrally connected, unless otherwise specifically indicated and defined; can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements. The specific meaning of the above terms in the present invention will be understood in specific cases by those of ordinary skill in the art.
In the description of embodiments of the invention, a particular feature, structure, material, or characteristic may be combined in any suitable manner in one or more embodiments or examples.
In describing embodiments of the present invention, it will be understood that the terms "-" and "-" are intended to be inclusive of the two numerical ranges, and that the ranges include the endpoints. For example, "A-B" means a range greater than or equal to A and less than or equal to B. "A-B" means a range of greater than or equal to A and less than or equal to B.
In the description of embodiments of the present invention, the term "and/or" is merely an association relationship describing an association object, meaning that three relationships may exist, e.g., a and/or B, may represent: a exists alone, A and B exist together, and B exists alone. In addition, the character "/" herein generally indicates that the front and rear associated objects are an "or" relationship.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (10)
1. A accent signal detection system, the accent signal detection system comprising:
a satellite signal acquisition module;
a satellite signal characteristic analysis module;
a key signal detection module;
the satellite signal acquisition module acquires satellite signals, wherein the satellite signals comprise historical satellite signals and satellite signals to be detected;
the satellite signal characteristic analysis module extracts a historical satellite signal and a satellite signal characteristic matrix of a satellite signal to be detected;
the key signal detection module performs deep reinforcement learning model training by using a satellite signal feature matrix of a satellite signal to be detected, and performs decision classification tree training by using a satellite signal feature matrix of a historical satellite signal so as to detect whether the satellite signal has the appearance or disappearance of the key signal.
2. The accent signal detection system of claim 1, wherein the satellite signal feature analysis module has:
an extraction unit;
a coding unit;
the extraction unit extracts satellite signal characteristics of historical satellite signals and satellite signals to be detected;
the encoding unit encodes non-floating point characteristics in the extracted satellite signal characteristics.
3. The accent signal detection system of claim 2, wherein the satellite signal feature analysis module further comprises:
the satellite signal feature set construction module;
the satellite signal feature matrix construction module;
the satellite signal feature set construction module constructs a satellite signal feature set corresponding to each time feature according to the extracted satellite signal features;
the satellite signal feature matrix construction module constructs a satellite signal feature matrix comprising each satellite signal feature group according to the constructed satellite signal feature groups, and the satellite signal feature groups in the satellite signal feature matrix are ordered according to the corresponding time features.
4. The accent signal detection system of claim 3, wherein the accent signal detection module has:
a deep reinforcement learning unit;
the deep reinforcement learning unit performs deep reinforcement learning model training according to the acquired satellite signal feature matrixes of the satellite signals to be detected, takes each satellite signal feature group in the satellite signal feature matrixes as a state input, and acquires action execution vector groups of the satellite signal feature matrixes;
the motion execution vector group comprises motion execution vectors corresponding to each satellite signal feature group, and the motion execution vectors comprise motion action1 and non-motion action2.
5. The accent signal detection system of claim 4, wherein the deep reinforcement learning model trains a loss function that is specifically:
wherein->Andfor network parameters in the training process, +.>For discounts factor->For observing the reward function->Representing the current state and the current action, < >>For the next state, ++>For optimal action, Q is the Q function.
6. The accent signal detection system of claim 4, wherein the satellite signal feature matrix of the historical satellite signals has an annotation field that characterizes whether the satellite signal feature matrix of the current historical satellite signals has an action of accent signal occurrence disappearance alternation.
7. The accent signal detection system of claim 6, wherein the accent signal detection module further comprises:
a decision classification tree unit;
the decision classification tree unit performs decision classification tree training by using the satellite signal feature matrix of the historical satellite signals, and judges whether the satellite signal feature matrix of the satellite signals to be detected has actions of alternating key signals appearing and disappearing by using the trained decision classification tree.
8. The accent signal detection system of claim 7, wherein the accent signal detection module further comprises:
a key signal appearance and disappearance judging unit;
the key signal appearance and disappearance judging unit is used for judging whether the key signal appears or disappears when the satellite signal feature matrix of the satellite signal to be detected has the action of alternately appearing and disappearing the key signal: judging whether the motion execution vector of the satellite signal feature group has motion action1 or not, when the motion execution vector has motion action1, the key signal appears in the first motion action1 and disappears in the last motion action 1;
the key signal appearance and disappearance judging unit is used for judging whether the key signal appearance and disappearance are alternated when the satellite signal feature matrix of the satellite signal to be detected does not have the action of the key signal appearance and disappearance alternation: judging whether the previous action of the first action1 is action2 or not, if so, generating an important signal; if not, the action execution vector of the satellite signal feature set has action2, and the key signal disappears.
9. The accent signal detection system of claim 1, wherein the accent signal detection system further comprises:
a manual checking module;
and the manual verification module generates correction data according to the detection result of the key signal detection module, and performs decision classification tree training in the key signal detection module by using the correction data.
10. A method for detecting a key signal, the method comprising the steps of:
acquiring satellite signals, wherein the satellite signals comprise historical satellite signals and satellite signals to be detected;
extracting a satellite signal characteristic matrix of a historical satellite signal and a satellite signal to be detected;
and training a deep reinforcement learning model by using a satellite signal characteristic matrix of the satellite signal to be detected, and training a decision classification tree by using a satellite signal characteristic matrix of the historical satellite signal to detect whether the satellite signal has the occurrence or the disappearance of an important signal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310752613.XA CN116502071B (en) | 2023-06-26 | 2023-06-26 | Key signal detection system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310752613.XA CN116502071B (en) | 2023-06-26 | 2023-06-26 | Key signal detection system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116502071A true CN116502071A (en) | 2023-07-28 |
| CN116502071B CN116502071B (en) | 2023-11-17 |
Family
ID=87323424
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310752613.XA Active CN116502071B (en) | 2023-06-26 | 2023-06-26 | Key signal detection system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116502071B (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150063495A1 (en) * | 2013-08-27 | 2015-03-05 | Blackberry Limited | Soft detection of m-ary dpsk signals |
| US20160203332A1 (en) * | 2015-01-13 | 2016-07-14 | Toyota Motor Engineering & Manufacturing North America, Inc. | Personal information protection for navigation systems |
| CN109348536A (en) * | 2018-11-28 | 2019-02-15 | 成都中星世通电子科技有限公司 | A kind of radio-interference signals automatic station-keeping system waterborne and localization method |
| CN111049595A (en) * | 2019-11-29 | 2020-04-21 | 西安冉科信息技术有限公司 | Signal identification and acquisition system |
| CN113485212A (en) * | 2021-09-07 | 2021-10-08 | 北京扬铭科技发展有限责任公司 | Broadband satellite signal intelligent identification system |
| CN114124661A (en) * | 2021-08-18 | 2022-03-01 | 广东电网有限责任公司广州供电局 | Power grid Ka high-flux satellite network communication fault determination method and device |
| CN114740501A (en) * | 2022-03-14 | 2022-07-12 | 上海第二工业大学 | Satellite navigation deception signal detection method based on frequency domain parameters |
| CN115483964A (en) * | 2022-09-30 | 2022-12-16 | 中国人民解放军陆军工程大学 | Air-space-ground integrated Internet of things communication resource joint allocation method |
| CN115796042A (en) * | 2022-12-07 | 2023-03-14 | 中国电子科技集团公司第十研究所 | Electronic reconnaissance-oriented intelligent game counter deduction system and method |
| WO2023092923A1 (en) * | 2021-11-26 | 2023-06-01 | 北京理工大学 | Composite interference signal recognition method and system |
| CN116223019A (en) * | 2022-12-01 | 2023-06-06 | 中国航空工业集团公司沈阳飞机设计研究所 | Gear box and fault diagnosis method thereof |
-
2023
- 2023-06-26 CN CN202310752613.XA patent/CN116502071B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150063495A1 (en) * | 2013-08-27 | 2015-03-05 | Blackberry Limited | Soft detection of m-ary dpsk signals |
| US20160203332A1 (en) * | 2015-01-13 | 2016-07-14 | Toyota Motor Engineering & Manufacturing North America, Inc. | Personal information protection for navigation systems |
| CN109348536A (en) * | 2018-11-28 | 2019-02-15 | 成都中星世通电子科技有限公司 | A kind of radio-interference signals automatic station-keeping system waterborne and localization method |
| CN111049595A (en) * | 2019-11-29 | 2020-04-21 | 西安冉科信息技术有限公司 | Signal identification and acquisition system |
| CN114124661A (en) * | 2021-08-18 | 2022-03-01 | 广东电网有限责任公司广州供电局 | Power grid Ka high-flux satellite network communication fault determination method and device |
| CN113485212A (en) * | 2021-09-07 | 2021-10-08 | 北京扬铭科技发展有限责任公司 | Broadband satellite signal intelligent identification system |
| WO2023092923A1 (en) * | 2021-11-26 | 2023-06-01 | 北京理工大学 | Composite interference signal recognition method and system |
| CN114740501A (en) * | 2022-03-14 | 2022-07-12 | 上海第二工业大学 | Satellite navigation deception signal detection method based on frequency domain parameters |
| CN115483964A (en) * | 2022-09-30 | 2022-12-16 | 中国人民解放军陆军工程大学 | Air-space-ground integrated Internet of things communication resource joint allocation method |
| CN116223019A (en) * | 2022-12-01 | 2023-06-06 | 中国航空工业集团公司沈阳飞机设计研究所 | Gear box and fault diagnosis method thereof |
| CN115796042A (en) * | 2022-12-07 | 2023-03-14 | 中国电子科技集团公司第十研究所 | Electronic reconnaissance-oriented intelligent game counter deduction system and method |
Non-Patent Citations (3)
| Title |
|---|
| YANYAN HAO 等: "Satellite Fault Detection and Diagnosis based on Data Compression and Improved Decision Tree", 《2020 CHINESE AUTOMATION CONGRESS (CAC)》 * |
| 杨福喜;邵龙才;刘天昀;: "重视细节讲究成效 全面推进抓好落实――扬州市无线电管理处扎实开展台站核查工作纪实", 中国无线电, no. 11 * |
| 苏琳: "基于宽带接收的星载信号处理技术", 《电讯技术》, vol. 51, no. 7 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116502071B (en) | 2023-11-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10769484B2 (en) | Character detection method and apparatus | |
| CN107221326B (en) | Voice awakening method and device based on artificial intelligence and computer equipment | |
| CN111027487B (en) | Behavior recognition system, method, medium and equipment based on multi-convolution kernel residual error network | |
| US8812313B2 (en) | Voice activity detector, voice activity detection program, and parameter adjusting method | |
| US9489598B2 (en) | Systems and methods for object classification, object detection and memory management | |
| US11270218B2 (en) | Mapper component for a neuro-linguistic behavior recognition system | |
| US20170068868A1 (en) | Enhancing handwriting recognition using pre-filter classification | |
| CN108491845B (en) | Character segmentation position determination method, character segmentation method, device and equipment | |
| CN111401259B (en) | Model training method, system, computer readable medium and electronic device | |
| US8938389B2 (en) | Voice activity detector, voice activity detection program, and parameter adjusting method | |
| US11145308B2 (en) | Symbol sequence estimation in speech | |
| CN112650885A (en) | Video classification method, device, equipment and medium | |
| CN112132830A (en) | Air tightness detection water body shaking sensing method and system based on artificial intelligence | |
| CN116502071B (en) | Key signal detection system and method | |
| US20200257956A1 (en) | Anomaly score adjustment across anomaly generators | |
| CN112347826B (en) | Video continuous sign language recognition method and system based on reinforcement learning | |
| CN115239733A (en) | Crack detection method, crack detection device, terminal equipment and storage medium | |
| CN114399491A (en) | Human face image quality evaluation method and device and computer readable storage medium | |
| CN113902044A (en) | Image target extraction method based on lightweight YOLOV3 | |
| CN108520237B (en) | Risk behavior identification method | |
| CN114397964B (en) | Method and device for detecting effective fixation point, electronic equipment and storage medium | |
| US11880654B2 (en) | Acquiring event information from a plurality of texts | |
| US20230118614A1 (en) | Electronic device and method for training neural network model | |
| CN113705513B (en) | Model training and lane line prediction method, electronic device and automatic driving vehicle | |
| CN114140802B (en) | Text recognition method and device, electronic equipment and storage medium |
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 |