CN117527135A - System and method for interfering unmanned aerial vehicle communication based on deep learning - Google Patents
System and method for interfering unmanned aerial vehicle communication based on deep learning Download PDFInfo
- Publication number
- CN117527135A CN117527135A CN202410013749.3A CN202410013749A CN117527135A CN 117527135 A CN117527135 A CN 117527135A CN 202410013749 A CN202410013749 A CN 202410013749A CN 117527135 A CN117527135 A CN 117527135A
- Authority
- CN
- China
- Prior art keywords
- unmanned aerial
- aerial vehicle
- position data
- dimensional space
- space position
- 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
- 238000004891 communication Methods 0.000 title claims abstract description 49
- 230000002452 interceptive effect Effects 0.000 title claims abstract description 44
- 238000013135 deep learning Methods 0.000 title claims abstract description 37
- 238000000034 method Methods 0.000 title claims abstract description 24
- 230000007246 mechanism Effects 0.000 claims abstract description 205
- 230000002093 peripheral effect Effects 0.000 claims abstract description 47
- 230000009471 action Effects 0.000 claims abstract description 19
- 230000000007 visual effect Effects 0.000 claims description 97
- 238000001514 detection method Methods 0.000 claims description 66
- 230000000875 corresponding effect Effects 0.000 claims description 56
- 238000013528 artificial neural network Methods 0.000 claims description 54
- 238000002360 preparation method Methods 0.000 claims description 15
- 230000002596 correlated effect Effects 0.000 claims description 9
- 239000000725 suspension Substances 0.000 claims description 8
- 238000006243 chemical reaction Methods 0.000 claims description 6
- 239000000203 mixture Substances 0.000 claims description 4
- 230000001276 controlling effect Effects 0.000 claims description 3
- 238000004458 analytical method Methods 0.000 description 10
- 238000010586 diagram Methods 0.000 description 10
- 238000000605 extraction Methods 0.000 description 6
- 238000012545 processing Methods 0.000 description 5
- 238000004088 simulation Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000005284 excitation Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000011897 real-time detection Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007123 defense Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000009545 invasion Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000004804 winding Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04K—SECRET COMMUNICATION; JAMMING OF COMMUNICATION
- H04K3/00—Jamming of communication; Counter-measures
- H04K3/40—Jamming having variable characteristics
- H04K3/43—Jamming having variable characteristics characterized by the control of the jamming power, signal-to-noise ratio or geographic coverage area
-
- 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
- G06N3/0499—Feedforward 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Molecular Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a system and a method for interfering unmanned aerial vehicle communication based on deep learning, and belongs to the technical field of wireless communication. The system comprises: the dynamic interference mechanism is used for starting signal interference actions on the peripheral area when the body early warning signal is received; the intelligent identification mechanism is used for intelligently analyzing the position data of the unmanned aerial vehicle at the future moment based on the position data of each past moment; and the signal triggering mechanism is used for sending out a mechanism early warning signal at the current moment when the position data of a certain unmanned aerial vehicle at the future moment is positioned in the peripheral area. According to the method and the device, the spatial position data of the unmanned aerial vehicle at the future moment can be intelligently predicted according to the spatial position data of each past moment of the unmanned aerial vehicle, so that high-strength signal interference is performed in advance when the predicted spatial position data is in the forbidden region, and the probability of forbidden information leakage is reduced.
Description
Technical Field
The invention relates to the technical field of wireless communication, in particular to a system and a method for interfering unmanned aerial vehicle communication based on deep learning.
Background
With the rapid development of unmanned aerial vehicle technology, unmanned aerial vehicles are increasingly widely applied in civil and military fields. However, since some unmanned aerial vehicles may be misused or mishandled, it poses a potential threat to social and public safety. Therefore, unmanned aerial vehicle interference systems, i.e., unmanned aerial vehicle countering systems, have been developed for identifying, tracking, and interfering with unmanned aerial vehicles to maintain safety and order. Generally, unmanned aerial vehicle countering equipment cuts off communication between unmanned aerial vehicle and remote controller through the data link and the navigation link that interfere unmanned aerial vehicle to force unmanned aerial vehicle to drop or drive it away automatically, ensure low altitude airspace safety, provide effectual safety assurance for large-scale activity and important place.
For example, an unmanned aerial vehicle target interference device proposed by chinese patent publication CN212486519U includes an intelligent tripod head, an interference host installed on the intelligent tripod head, and a full-band directional antenna array, where a digital excitation unit, a power amplifier group, a switch module, a power module, a radiator fan, and an LED indicator are disposed in the interference host; the digital excitation signal radio frequency output port of the digital excitation unit is connected with the radio frequency input port of the power amplifier group, the radio frequency output port of the power amplifier is connected with the radio frequency output port of the full-frequency band directional antenna array, and the control port of the power amplifier is in communication connection with the control port of the digital processing unit. The method can be used for carrying out interference aiming at any frequency in the frequency range of 400 MHz-6000 MHz, and is specially designed for unmanned aerial vehicle defense systems; and the interference signals of 3 frequency bands are simultaneously transmitted, so that the multi-target multi-frequency band interference capability is realized.
For example, an omnibearing unmanned aerial vehicle interference device is proposed in chinese patent publication CN209982504U, the device includes a signal transmitting device and an operating host, a power module, a main control board and multiple groups of signal transmitting modules connected to the main control board are arranged in the signal transmitting device, the signal transmitting modules include a signal generator, a radio frequency power amplifier module and directional antennas, the directional antennas are arranged in a circle, the signal transmitting device is connected to the operating host through a gateway port, the main control board and the multiple groups of signal transmitting modules are connected to the power module, and the power module is connected to a power supply port arranged on a shell of the signal transmitting device. The beneficial effects of the utility model are: the operation host computer sends out interference signal's instruction, makes signal generator transmit the interference signal with unmanned aerial vehicle co-frequency through the main control panel, and radio frequency power amplifier module amplifies interference signal, and through a plurality of directional antennas of winding a week transmit simultaneously, to unmanned aerial vehicle transmission interference signal in each direction, the transmission does not have the dead angle, can effectively prevent unmanned aerial vehicle's invasion.
However, the above prior art only relates to a specific structural design of hardware of an unmanned aerial vehicle interference system, and processes real-time detection and real-time interference on an unmanned aerial vehicle, however, from real-time detection of the unmanned aerial vehicle entering a no-fly area, then to the unmanned aerial vehicle transmitting an interference signal in the no-fly area, there is a time difference between the two, and the unmanned aerial vehicle has completed entering the no-fly area and has entered for a certain period of time, so that it is very likely that shooting and returning of forbidden information have been completed, and at this time, the subsequent unmanned aerial vehicle interference appears too late, and the expected countering effect cannot be achieved.
Disclosure of Invention
In order to solve the technical problems in the prior art, the invention provides a system and a method for interfering unmanned aerial vehicle communication based on deep learning, which can intelligently predict the spatial position data of an unmanned aerial vehicle at the future moment by adopting an artificial intelligent model based on the spatial position data of the same unmanned aerial vehicle, which correspond to each past moment respectively, so that when the spatial position data at the future moment is in a no-fly area, high-strength signal interference is carried out on the no-fly area in advance before the future moment arrives, thereby completing advanced countering of the high-risk unmanned aerial vehicle and effectively reducing the possibility that the unmanned aerial vehicle adopts a burst mode to enter and exit the no-fly area to obtain the no-fly information.
According to a first aspect of the present invention, there is provided a deep learning based system for interfering with unmanned aerial vehicle communications, the system comprising:
the dynamic interference mechanism is used for entering an interference preparation state when an organism approaching signal is received, and starting signal interference actions on the peripheral area of the adjacent set range when the organism early warning signal is received;
the visual detection mechanism is arranged at the front end of the dynamic interference mechanism and is connected with the dynamic interference mechanism, and is used for performing large-visual-angle visual data acquisition facing the peripheral area to obtain a wide-area acquisition picture, and sending a body approximation signal to the dynamic interference mechanism when an unmanned aerial vehicle target with a depth of view being smaller than a set depth of field limit exists in the wide-area acquisition picture;
The position conversion mechanism is connected with the visual detection mechanism and is used for acquiring each frame of wide-area acquisition picture which is acquired by the visual detection mechanism at each acquisition time at uniform intervals, determining three-dimensional spatial position data of each unmanned aerial vehicle target in each frame of wide-area acquisition picture based on depth of field, coordinate data of each unmanned aerial vehicle target in the wide-area acquisition picture and three-dimensional spatial position data of the visual detection mechanism, and outputting the three-dimensional spatial position data as three-dimensional spatial position data of each unmanned aerial vehicle target at the corresponding acquisition time of the wide-area acquisition picture;
the intelligent identification mechanism is respectively connected with the dynamic interference mechanism and the position conversion mechanism and is used for intelligently identifying three-dimensional space position data corresponding to the next time of the same unmanned aerial vehicle target by adopting an AI identification model based on the three-dimensional space position data of the visual detection mechanism and the three-dimensional space position data corresponding to the next time of the same unmanned aerial vehicle target before the next time of the current time;
the signal triggering mechanism is respectively connected with the intelligent identification mechanism and the dynamic interference mechanism and is used for sending a mechanism early warning signal at the current moment when three-dimensional space position data corresponding to a certain unmanned aerial vehicle target at the next moment is positioned in the peripheral area, or sending a suspension early warning signal at the current moment;
The AI identification model is a feedforward neural network subjected to multiple learning, and the learning times of the feedforward neural network are positively correlated with the value of a set distance threshold;
when the body early warning signal is received, the signal interference action of the adjacent peripheral area with the set range is started, which comprises the following steps: the surrounding area of the adjacent set range is composed of each position of the distance to the dynamic interference mechanism within the set distance threshold, and the front end of the dynamic interference mechanism is used for transmitting an interference signal for interfering the unmanned aerial vehicle in the surrounding area.
According to a second aspect of the present invention, there is provided a method of interfering with unmanned aerial vehicle communications based on deep learning, the method comprising:
when an organism approaching signal is received, controlling the dynamic interference mechanism to enter an interference preparation state;
a visual detection mechanism arranged at the front end of the dynamic interference mechanism is adopted to perform large-visual-angle visual data acquisition on a peripheral area so as to obtain a wide-area acquisition picture, and when an unmanned aerial vehicle target with a scene depth being smaller than a set depth of field limit exists in the wide-area acquisition picture, an organism approaching signal is sent;
acquiring each frame of wide-area acquisition pictures respectively acquired by the visual detection mechanism at each acquisition time at uniform intervals, determining three-dimensional space position data of each unmanned aerial vehicle target in each frame of wide-area acquisition picture based on depth of field, coordinate data of each unmanned aerial vehicle target in the wide-area acquisition picture and three-dimensional space position data of the visual detection mechanism, and outputting the three-dimensional space position data as three-dimensional space position data of each unmanned aerial vehicle target at the corresponding acquisition time of the wide-area acquisition picture;
After the dynamic interference mechanism enters an interference preparation state, aiming at the same unmanned aerial vehicle target, intelligently identifying the corresponding three-dimensional space position data at the next moment by adopting an AI identification model based on the three-dimensional space position data of the visual detection mechanism, the three-dimensional space position data of the visual detection mechanism and the interval duration between every two adjacent acquisition moments, wherein the three-dimensional space position data respectively correspond to the acquisition moments before the next moment of the current moment;
when a certain unmanned aerial vehicle target exists and the corresponding three-dimensional space position data of the unmanned aerial vehicle target at the next time is located in the peripheral area, a mechanism early warning signal is sent out at the current time, otherwise, a suspension early warning signal is sent out at the current time;
when an organism early warning signal is received, starting signal interference action on a peripheral area of a set range adjacent to the dynamic interference mechanism;
the AI identification model is a feedforward neural network subjected to multiple learning, and the learning times of the feedforward neural network are positively correlated with the value of a set distance threshold;
when the body early warning signal is received, the signal interference action of the peripheral area of the set range adjacent to the dynamic interference mechanism is started, which comprises the following steps: the peripheral area of the set range adjacent to the dynamic interference mechanism is composed of each position of which the distance to the dynamic interference mechanism is within the set distance threshold, and the front end of the dynamic interference mechanism is used for transmitting an interference signal for interfering the unmanned aerial vehicle in the peripheral area.
Compared with the prior art, the invention has at least the following outstanding substantive progress:
first,: the three-dimensional position of the same unmanned aerial vehicle at the next moment is intelligently predicted according to the three-dimensional position of the same unmanned aerial vehicle at each past moment, and when the three-dimensional position of the unmanned aerial vehicle at the next moment is positioned in a set space, high-strength signal interference is started, so that unmanned aerial vehicle interference actions in the set space are extracted and performed based on intelligent prediction results of the three-dimensional positions, and the success probability that the unmanned aerial vehicle suddenly enters a forbidden space and shoots and returns forbidden information is reduced;
secondly: aiming at the same unmanned aerial vehicle target, an AI identification model is adopted to execute corresponding intelligent prediction processing, input data of the model is a plurality of pieces of three-dimensional space position data corresponding to a plurality of acquisition moments of the unmanned aerial vehicle target before the next moment of the current moment, three-dimensional space position data of a visual detection mechanism and interval duration between every two adjacent acquisition moments, the AI identification model is a feedforward neural network subjected to repeated learning, the learning times of the feedforward neural network are positively correlated with the range size of a set space, and therefore the targeted customization of the AI identification model is completed;
Again: in each learning executed on the feedforward neural network, taking three-dimensional space position data corresponding to a certain unmanned aerial vehicle target at a certain past moment as output content of the feedforward neural network, taking a plurality of pieces of three-dimensional space position data corresponding to a plurality of acquisition moments of the certain unmanned aerial vehicle target before the certain past moment, three-dimensional space position data of a visual detection mechanism and interval duration between every two adjacent acquisition moments as a plurality of input content of the feedforward neural network, and executing the learning operation, thereby ensuring the learning effect of each learning operation of the feedforward neural network;
finally: in the input data of the AI identification model, the number of the plurality of acquisition moments is positively correlated with the interval duration between every two adjacent acquisition moments, so that the self-adaptive level of the AI identification model structure is improved, and the effectiveness and stability of the identification result are ensured.
Drawings
Embodiments of the present invention will be described below with reference to the accompanying drawings, in which:
fig. 1 is a technical flow diagram of a deep learning based system and method for interfering with drone communications in accordance with the present invention.
Fig. 2 is an internal structural diagram of a system for interfering with unmanned aerial vehicle communication based on deep learning, which is shown in accordance with a first embodiment of the present invention.
Fig. 3 is an internal structural diagram of a system for interfering with unmanned aerial vehicle communication based on deep learning, according to a second embodiment of the present invention.
Fig. 4 is an internal structural diagram of a system for interfering with unmanned aerial vehicle communication based on deep learning, which is shown in accordance with a third embodiment of the present invention.
Fig. 5 is an internal structural diagram of a system for interfering with unmanned aerial vehicle communication based on deep learning, according to a fourth embodiment of the present invention.
Detailed Description
As shown in fig. 1, a technical flow diagram of a system and method for interfering with unmanned aerial vehicle communications based on deep learning is presented, as shown in the present invention.
As shown in fig. 1, the specific technical process of the present invention is as follows:
the first step: establishing an artificial intelligent model for intelligently predicting the three-dimensional position of the unmanned aerial vehicle at a future moment;
specifically, the artificial intelligent model is an AI identification model, the AI identification model is a feedforward neural network subjected to multiple learning, each item of input data of the AI identification model is each three-dimensional position of the unmanned aerial vehicle corresponding to each time in the past, interval duration of two adjacent times and three-dimensional position of a visual detection mechanism for acquiring visual data for positioning, and output data of the AI identification model is prediction data of the three-dimensional position of the unmanned aerial vehicle at future time;
The learning times of the feedforward neural network are positively correlated with the range size of a set space, so that the targeted customization of the AI identification model is completed;
in each learning executed on the feedforward neural network, taking the three-dimensional position corresponding to a certain unmanned aerial vehicle target at a certain past moment as output content of the feedforward neural network, taking a plurality of three-dimensional positions corresponding to a plurality of acquisition moments of the certain unmanned aerial vehicle target before the certain past moment, the three-dimensional position of a visual detection mechanism and interval time of two adjacent moments as a plurality of input content of the feedforward neural network, and executing the learning operation, thereby ensuring the learning effect of each learning operation of the feedforward neural network;
and a second step of: screening various basic data for intelligent prediction, wherein the basic data comprise various three-dimensional positions of the unmanned aerial vehicle corresponding to various past moments, interval duration of two adjacent moments and three-dimensional positions of a visual detection mechanism for acquiring visual data for positioning;
in each basic data, it is particularly critical that each three-dimensional position of the unmanned aerial vehicle respectively corresponds to each time in the past, as shown in fig. 1, visual detection of each three-dimensional position of the unmanned aerial vehicle respectively corresponds to each time in the past can be completed through a visual positioning mode;
And a third step of: using the artificial intelligent model established in the first step to intelligently predict the three-dimensional position of each unmanned aerial vehicle at the future moment based on the basic data screened in the second step, as shown in figure 1;
by way of example, the more unmanned aerial vehicles exist in the current airspace, the more times of intelligent prediction are executed, and each intelligent prediction execution corresponds to the same unmanned aerial vehicle in the current airspace;
fourth step: when the three-dimensional position of the unmanned aerial vehicle at the future moment is positioned in the set flying forbidden space, high-intensity signal interference on the set flying forbidden space is started before the future moment arrives;
therefore, the following steps are adopted, the unmanned aerial vehicle interference action in the setting space is realized based on the intelligent prediction result of the three-dimensional position, and the probability of success that the unmanned aerial vehicle suddenly enters the forbidden space and shoots and returns forbidden information is reduced.
The key points of the invention are as follows: the method comprises the steps of customizing structures of an artificial intelligent model for intelligently predicting three-dimensional positions of unmanned aerial vehicles at future time, customizing learning modes, performing targeted screening of basic data of intelligent prediction, a visual positioning mechanism of each unmanned aerial vehicle and advanced high-intensity interference processing based on the three-dimensional positions of the unmanned aerial vehicles at the future time.
The system and method for interfering with unmanned aerial vehicle communication based on deep learning of the present invention will be described in detail by way of example.
First embodiment
Fig. 2 is an internal structural diagram of a system for interfering with unmanned aerial vehicle communication based on deep learning, which is shown in accordance with a first embodiment of the present invention.
As shown in fig. 2, the deep learning-based system for interfering with unmanned aerial vehicle communication comprises the following specific components:
the dynamic interference mechanism is used for entering an interference preparation state when an organism approaching signal is received, and starting signal interference actions on the peripheral area of the adjacent set range when the organism early warning signal is received;
the dynamic interference mechanism may be internally provided with a signal receiving and transmitting unit, a state switching unit, an interference executing unit and a microcontroller, wherein the microcontroller is respectively connected with the signal receiving and transmitting unit and the state switching unit and is used for realizing real-time configuration of working parameters of the signal receiving and transmitting unit and the state switching unit;
specifically, the signal receiving and transmitting unit is used for receiving the organism approaching signal or the organism early warning signal;
the state switching unit is connected with the signal receiving and transmitting unit and is used for driving the dynamic interference mechanism to enter an interference preparation state when an organism approaching signal is received and starting the signal interference action of the dynamic interference mechanism on the adjacent peripheral area of the set range when an organism early warning signal is received;
The interference execution unit is connected with the state driving unit and is used for receiving state configuration of the state driving unit;
the visual detection mechanism is arranged at the front end of the dynamic interference mechanism and is connected with the dynamic interference mechanism, and is used for performing large-visual-angle visual data acquisition facing the peripheral area to obtain a wide-area acquisition picture, and sending a body approximation signal to the dynamic interference mechanism when an unmanned aerial vehicle target with a depth of view being smaller than a set depth of field limit exists in the wide-area acquisition picture;
for example, the visual detection mechanism may include an image sensing unit, a target extraction unit, and a depth of field analysis unit, the target extraction unit being connected to the image sensing unit and the depth of field analysis unit, respectively;
the image sensing unit is used for performing visual data acquisition of a large visual angle facing the peripheral area so as to obtain a wide-area acquisition picture;
the target extraction unit is used for extracting each unmanned aerial vehicle target in the wide-area acquisition picture, and the depth of field analysis unit is used for analyzing each depth of field data of each unmanned aerial vehicle target in the wide-area acquisition picture;
The image sensing unit may be a CMOS sensing unit or a CCD sensing unit, for example;
the position conversion mechanism is connected with the visual detection mechanism and is used for acquiring each frame of wide-area acquisition picture which is acquired by the visual detection mechanism at each acquisition time at uniform intervals, determining three-dimensional spatial position data of each unmanned aerial vehicle target in each frame of wide-area acquisition picture based on depth of field, coordinate data of each unmanned aerial vehicle target in the wide-area acquisition picture and three-dimensional spatial position data of the visual detection mechanism, and outputting the three-dimensional spatial position data as three-dimensional spatial position data of each unmanned aerial vehicle target at the corresponding acquisition time of the wide-area acquisition picture;
illustratively, determining three-dimensional space position data of the wide-range acquisition picture based on the depth of field, coordinate data of the wide-range acquisition picture and three-dimensional space position data of the visual detection mechanism, and outputting the three-dimensional space position data as three-dimensional space position data of the wide-range acquisition picture at corresponding acquisition time comprises: a numerical analysis formula can be adopted to determine three-dimensional space position data of the visual detection mechanism based on the depth of field, coordinate data of the visual detection mechanism in the wide-area acquisition picture and the three-dimensional space position data;
Specifically, a numerical analysis formula may be adopted to determine three-dimensional spatial position data of the vision detection mechanism based on the depth of field thereof, coordinate data thereof in the wide-area acquisition picture, and three-dimensional spatial position data of the vision detection mechanism, including: the smaller the value of the depth of field is, the shorter the distance between the three-dimensional space position data of the depth of field and the three-dimensional space position data of the visual detection mechanism is;
the intelligent identification mechanism is respectively connected with the dynamic interference mechanism and the position conversion mechanism and is used for intelligently identifying three-dimensional space position data corresponding to the next time of the same unmanned aerial vehicle target by adopting an AI identification model based on the three-dimensional space position data of the visual detection mechanism and the three-dimensional space position data corresponding to the next time of the same unmanned aerial vehicle target before the next time of the current time;
for example, a numerical simulation mechanism can be selected to complete simulation of a data processing process of intelligently identifying three-dimensional space position data corresponding to the next time by adopting an AI identification model based on a plurality of pieces of three-dimensional space position data corresponding to a plurality of acquisition times before the next time of the current time, three-dimensional space position data of a visual detection mechanism and interval duration between every two adjacent acquisition times;
The signal triggering mechanism is respectively connected with the intelligent identification mechanism and the dynamic interference mechanism and is used for sending a mechanism early warning signal at the current moment when three-dimensional space position data corresponding to a certain unmanned aerial vehicle target at the next moment is positioned in the peripheral area, or sending a suspension early warning signal at the current moment;
the AI identification model is a feedforward neural network subjected to multiple learning, and the learning times of the feedforward neural network are positively correlated with the value of a set distance threshold;
for example, the forward correlation of the learning times of the feedforward neural network with the value of the set distance threshold includes: setting the value of the distance threshold to be 100 meters, setting the learning frequency of the feedforward neural network to be 50, setting the value of the distance threshold to be 120 meters, setting the learning frequency of the feedforward neural network to be 60, setting the value of the distance threshold to be 150 meters, and setting the learning frequency of the feedforward neural network to be 70;
when the body early warning signal is received, the signal interference action of the adjacent peripheral area with the set range is started, which comprises the following steps: the adjacent peripheral area of the set range is composed of all positions of which the distance to the dynamic interference mechanism is within a set distance threshold, and the front end of the dynamic interference mechanism is used for transmitting interference signals for interfering unmanned aerial vehicles in the peripheral area;
Wherein, adopt AI authentication model intelligence to appraise its three-dimensional space position data that corresponds at the moment of next moment based on its a plurality of three-dimensional space position data that correspond respectively of a plurality of collection moments before the moment of current moment, the three-dimensional space position data of visual detection mechanism and every interval duration between two adjacent collection moments includes: the plurality of acquisition moments before the next moment of the current moment comprise the current moment; the longer the interval duration between every two adjacent acquisition moments, the more the number of the plurality of acquisition moments;
and wherein the front end of the dynamic interference mechanism is configured to send an interference signal that interferes with the unmanned aerial vehicle in the peripheral area, including: and the transmitting power of the interference signal is higher than or equal to a set power threshold.
Second embodiment
Fig. 3 is an internal structural diagram of a system for interfering with unmanned aerial vehicle communication based on deep learning, according to a second embodiment of the present invention.
As shown in fig. 3, unlike the embodiment in fig. 2, the deep learning-based system for interfering with unmanned aerial vehicle communications further includes:
the wireless notification mechanism is connected with the signal triggering mechanism and is used for transmitting three-dimensional space position data corresponding to the unmanned aerial vehicle target at the next moment to a remote cloud computing server through a wireless communication link when the signal triggering mechanism transmits a mechanism early warning signal;
For example, the wireless notification mechanism may be a frequency division duplex notification mechanism or a time division duplex notification mechanism;
the method for transmitting the three-dimensional space position data corresponding to the next moment of the certain unmanned aerial vehicle target to the remote cloud computing server through the wireless communication link while the signal triggering mechanism transmits the mechanism early warning signal comprises the following steps: the wireless communication link is a frequency division duplex communication link or a time division duplex communication link.
Third embodiment
Fig. 4 is an internal structural diagram of a system for interfering with unmanned aerial vehicle communication based on deep learning, which is shown in accordance with a third embodiment of the present invention.
As shown in fig. 4, unlike the embodiment in fig. 3, the deep learning-based system for interfering with unmanned aerial vehicle communications further includes:
the cloud computing server is respectively connected with each wireless notification mechanism corresponding to each dynamic interference mechanism and is used for wirelessly receiving the three-dimensional space position data sent by each wireless notification mechanism through a wireless communication link;
alternatively, substitution of the cloud computing server may be implemented with a big data server or a blockchain server.
Fourth embodiment
Fig. 5 is an internal structural diagram of a system for interfering with unmanned aerial vehicle communication based on deep learning, according to a fourth embodiment of the present invention.
As shown in fig. 5, unlike the embodiment in fig. 2, the deep learning-based system for interfering with unmanned aerial vehicle communications further includes:
the model composition mechanism is connected with the intelligent identification mechanism and used for carrying out repeated learning on the feedforward neural network to obtain the feedforward neural network after repeated learning, and sending the feedforward neural network to the intelligent identification mechanism as an AI identification model for use;
the parameter storage mechanism is connected with the model composition mechanism and used for storing various model parameters of the AI identification model;
for example, the parameter storage mechanism may be selected as an MMC storage mechanism, a FLASH memory or a TF storage mechanism, and is configured to store each model parameter of the AI identification model;
the method for learning the feedforward neural network for multiple times to obtain the feedforward neural network after multiple times of learning, and sending the feedforward neural network to the intelligent authentication mechanism as an AI authentication model for use comprises the following steps: in each learning executed on the feedforward neural network, taking three-dimensional space position data corresponding to a certain unmanned aerial vehicle target at a certain past moment as output content of the feedforward neural network, taking a plurality of pieces of three-dimensional space position data corresponding to a plurality of acquisition moments of the certain unmanned aerial vehicle target before the certain past moment, three-dimensional space position data of a visual detection mechanism and interval duration between every two adjacent acquisition moments as a plurality of pieces of input content of the feedforward neural network, and executing the learning operation.
Fifth embodiment
Unlike the embodiment in fig. 2, the system for interfering with unmanned aerial vehicle communication based on deep learning according to the fifth embodiment of the present invention further comprises:
the power supply mechanism is connected with the dynamic interference mechanism and is used for supplying power to the dynamic interference mechanism;
for example, an uninterruptible power supply may alternatively be used to implement the power supply mechanism.
Next, a detailed description of the respective method embodiments of the present invention will be continued.
In a deep learning based system for interfering with drone communications according to any of the above embodiments of the present invention:
performing visual data acquisition of a large visual angle facing the peripheral area to obtain a wide-area acquisition picture, and transmitting an organism approximation signal to the dynamic interference mechanism when an unmanned aerial vehicle target with a scene depth smaller than a set depth limit exists in the wide-area acquisition picture comprises: the large visual angle is a visual acquisition visual angle with a numerical value exceeding or equal to a preset visual angle threshold;
illustratively, the large viewing angle is a visual acquisition viewing angle with a value exceeding or equal to a preset viewing angle threshold value, including: the preset visual angle reading value is 270 degrees, and the large visual angle is a visual acquisition visual angle with a numerical value exceeding or equal to 270 degrees;
Wherein, execute the visual data acquisition of big visual angle in order to obtain wide region and gather the picture in the face of peripheral region, when there is the scene depth in wide region to be less than the unmanned aerial vehicle target of setting for depth of field limit in gathering the picture, send organism approaching signal to dynamic interference mechanism still includes: detecting more than one unmanned aerial vehicle target in the wide-area acquisition picture based on each standard outline pattern corresponding to each type of unmanned aerial vehicle;
wherein, detect more than one unmanned aerial vehicle target in the wide-area acquisition picture based on each standard appearance pattern that each type unmanned aerial vehicle corresponds respectively includes: the resolution of each standard outline pattern corresponding to each type of unmanned aerial vehicle is the same, and the definition of each standard outline pattern corresponding to each type of unmanned aerial vehicle is the same.
In a deep learning based system for interfering with drone communications according to any of the above embodiments of the present invention:
the visual detection mechanism is also used for sending a body far-away signal to the dynamic interference mechanism when no unmanned aerial vehicle target with the depth of view being less than the set depth of field limit exists in the wide-area acquisition picture;
the dynamic interference mechanism is also used for exiting the interference preparation state when the machine body far-away signal is received, and the dynamic interference mechanism in the interference preparation state is a dynamic interference mechanism for preparing to transmit an interference signal;
And when the dynamic interference mechanism receives the suspension early warning signal, ending/interrupting the signal interference action on the peripheral area of the adjacent set range.
And in a deep learning based system for interfering with drone communications according to any of the above embodiments of the present invention:
based on a plurality of three-dimensional space position data which respectively correspond to a plurality of acquisition moments before the next moment of the current moment, three-dimensional space position data of a visual detection mechanism, and interval duration between every two adjacent acquisition moments, intelligent identification of the three-dimensional space position data which corresponds to the next moment by adopting an AI identification model comprises the following steps: the AI identification model is input in parallel by a plurality of pieces of three-dimensional space position data corresponding to a plurality of acquisition moments before the next moment of the current moment, three-dimensional space position data of a visual detection mechanism and interval duration between every two adjacent acquisition moments;
wherein, adopt AI authentication model intelligence to appraise its three-dimensional space position data that corresponds at the moment of next moment based on its a plurality of three-dimensional space position data that correspond respectively of a plurality of collection moments before the moment of current moment, the three-dimensional space position data of visual detection mechanism and every interval duration between two adjacent collection moments includes: and executing the AI identification model to obtain three-dimensional space position data which is output by the AI identification model and corresponds to the AI identification model at the next moment.
Sixth embodiment
A sixth embodiment of the present invention shows a method of interfering with unmanned aerial vehicle communication based on deep learning, comprising the steps of:
s61: when an organism approaching signal is received, controlling the dynamic interference mechanism to enter an interference preparation state;
the dynamic interference mechanism may be internally provided with a signal receiving and transmitting unit, a state switching unit, an interference executing unit and a microcontroller, wherein the microcontroller is respectively connected with the signal receiving and transmitting unit and the state switching unit and is used for realizing real-time configuration of working parameters of the signal receiving and transmitting unit and the state switching unit;
s62: a visual detection mechanism arranged at the front end of the dynamic interference mechanism is adopted to perform large-visual-angle visual data acquisition on a peripheral area so as to obtain a wide-area acquisition picture, and when an unmanned aerial vehicle target with a scene depth being smaller than a set depth of field limit exists in the wide-area acquisition picture, an organism approaching signal is sent;
for example, the visual detection mechanism may include an image sensing unit, a target extraction unit, and a depth of field analysis unit, the target extraction unit being connected to the image sensing unit and the depth of field analysis unit, respectively;
the image sensing unit is used for performing visual data acquisition of a large visual angle facing the peripheral area so as to obtain a wide-area acquisition picture;
The target extraction unit is used for extracting each unmanned aerial vehicle target in the wide-area acquisition picture, and the depth of field analysis unit is used for analyzing each depth of field data of each unmanned aerial vehicle target in the wide-area acquisition picture;
the image sensing unit may be a CMOS sensing unit or a CCD sensing unit, for example;
s63: acquiring each frame of wide-area acquisition pictures respectively acquired by the visual detection mechanism at each acquisition time at uniform intervals, determining three-dimensional space position data of each unmanned aerial vehicle target in each frame of wide-area acquisition picture based on depth of field, coordinate data of each unmanned aerial vehicle target in the wide-area acquisition picture and three-dimensional space position data of the visual detection mechanism, and outputting the three-dimensional space position data as three-dimensional space position data of each unmanned aerial vehicle target at the corresponding acquisition time of the wide-area acquisition picture;
illustratively, determining three-dimensional space position data of the wide-range acquisition picture based on the depth of field, coordinate data of the wide-range acquisition picture and three-dimensional space position data of the visual detection mechanism, and outputting the three-dimensional space position data as three-dimensional space position data of the wide-range acquisition picture at corresponding acquisition time comprises: a numerical analysis formula can be adopted to determine three-dimensional space position data of the visual detection mechanism based on the depth of field, coordinate data of the visual detection mechanism in the wide-area acquisition picture and the three-dimensional space position data;
Specifically, a numerical analysis formula may be adopted to determine three-dimensional spatial position data of the vision detection mechanism based on the depth of field thereof, coordinate data thereof in the wide-area acquisition picture, and three-dimensional spatial position data of the vision detection mechanism, including: the smaller the value of the depth of field is, the shorter the distance between the three-dimensional space position data of the depth of field and the three-dimensional space position data of the visual detection mechanism is;
s64: after the dynamic interference mechanism enters an interference preparation state, aiming at the same unmanned aerial vehicle target, intelligently identifying the corresponding three-dimensional space position data at the next moment by adopting an AI identification model based on the three-dimensional space position data of the visual detection mechanism, the three-dimensional space position data of the visual detection mechanism and the interval duration between every two adjacent acquisition moments, wherein the three-dimensional space position data respectively correspond to the acquisition moments before the next moment of the current moment;
for example, a numerical simulation mechanism can be selected to complete simulation of a data processing process of intelligently identifying three-dimensional space position data corresponding to the next time by adopting an AI identification model based on a plurality of pieces of three-dimensional space position data corresponding to a plurality of acquisition times before the next time of the current time, three-dimensional space position data of a visual detection mechanism and interval duration between every two adjacent acquisition times;
S65: when a certain unmanned aerial vehicle target exists and the corresponding three-dimensional space position data of the unmanned aerial vehicle target at the next time is located in the peripheral area, a mechanism early warning signal is sent out at the current time, otherwise, a suspension early warning signal is sent out at the current time;
s66: when an organism early warning signal is received, starting signal interference action on a peripheral area of a set range adjacent to the dynamic interference mechanism;
the signal transceiver unit is configured to receive the body approaching signal or the body early warning signal;
the state switching unit is connected with the signal receiving and transmitting unit and is used for driving the dynamic interference mechanism to enter an interference preparation state when an organism approaching signal is received and starting the signal interference action of the dynamic interference mechanism on the adjacent peripheral area of the set range when an organism early warning signal is received;
the interference execution unit is connected with the state driving unit and is used for receiving state configuration of the state driving unit;
the AI identification model is a feedforward neural network subjected to multiple learning, and the learning times of the feedforward neural network are positively correlated with the value of a set distance threshold;
For example, the forward correlation of the learning times of the feedforward neural network with the value of the set distance threshold includes: setting the value of the distance threshold to be 100 meters, setting the learning frequency of the feedforward neural network to be 50, setting the value of the distance threshold to be 120 meters, setting the learning frequency of the feedforward neural network to be 60, setting the value of the distance threshold to be 150 meters, and setting the learning frequency of the feedforward neural network to be 70;
when the body early warning signal is received, the signal interference action of the peripheral area of the set range adjacent to the dynamic interference mechanism is started, which comprises the following steps: the peripheral area of the set range adjacent to the dynamic interference mechanism is formed by each position of the distance reaching the dynamic interference mechanism within a set distance threshold, and the front end of the dynamic interference mechanism is used for transmitting an interference signal for interfering the unmanned aerial vehicle in the peripheral area;
wherein, adopt AI authentication model intelligence to appraise its three-dimensional space position data that corresponds at the moment of next moment based on its a plurality of three-dimensional space position data that correspond respectively of a plurality of collection moments before the moment of current moment, the three-dimensional space position data of visual detection mechanism and every interval duration between two adjacent collection moments includes: the plurality of acquisition moments before the next moment of the current moment comprise the current moment; the longer the interval duration between every two adjacent acquisition moments, the more the number of the plurality of acquisition moments;
And wherein the front end of the dynamic interference mechanism is configured to send an interference signal that interferes with the unmanned aerial vehicle in the peripheral area, including: and the transmitting power of the interference signal is higher than or equal to a set power threshold.
In addition, the present invention may also cite the following technical matters to highlight the significant technical progress of the present invention:
detecting more than one unmanned aerial vehicle target in the wide-area acquisition picture based on each standard outline pattern respectively corresponding to each type of unmanned aerial vehicle comprises: when the edge contour of a certain image block in the wide-area acquisition picture is matched with the edge contour of a standard outline pattern corresponding to a certain type of unmanned aerial vehicle, judging that the certain image block is an image area with a single unmanned aerial vehicle target;
in each learning performed on the feedforward neural network, taking three-dimensional space position data corresponding to a certain unmanned aerial vehicle target at a certain past moment as output content of the feedforward neural network, taking a plurality of pieces of three-dimensional space position data corresponding to a plurality of acquisition moments of the certain unmanned aerial vehicle target before the certain past moment, three-dimensional space position data of a visual detection mechanism and interval duration between every two adjacent acquisition moments as a plurality of pieces of input content of the feedforward neural network, and performing the learning operation includes: and selecting MATLAB tool boxes to realize simulation and test of each learning executed on the feedforward neural network.
The foregoing description of the exemplary embodiments of the invention has been presented for the purposes of illustration and description. It is not intended to be exhaustive or to limit the invention to the precise form disclosed. Obviously, many modifications and variations will be apparent to practitioners skilled in the art. The exemplary embodiments were chosen and described in order to best explain the principles of the invention and its practical application, thereby enabling others skilled in the art to understand the invention for various embodiments and with various modifications as are suited to the particular use contemplated. It is intended that the scope of the invention be defined by the following claims and their equivalents.
Claims (10)
1. A system for interfering with unmanned aerial vehicle communications based on deep learning, the system comprising:
the dynamic interference mechanism is used for entering an interference preparation state when an organism approaching signal is received, and starting signal interference actions on the peripheral area of the adjacent set range when the organism early warning signal is received;
the visual detection mechanism is arranged at the front end of the dynamic interference mechanism and is connected with the dynamic interference mechanism, and is used for performing large-visual-angle visual data acquisition facing the peripheral area to obtain a wide-area acquisition picture, and sending a body approximation signal to the dynamic interference mechanism when an unmanned aerial vehicle target with a depth of view being smaller than a set depth of field limit exists in the wide-area acquisition picture;
The position conversion mechanism is connected with the visual detection mechanism and is used for acquiring each frame of wide-area acquisition picture which is acquired by the visual detection mechanism at each acquisition time at uniform intervals, determining three-dimensional spatial position data of each unmanned aerial vehicle target in each frame of wide-area acquisition picture based on depth of field, coordinate data of each unmanned aerial vehicle target in the wide-area acquisition picture and three-dimensional spatial position data of the visual detection mechanism, and outputting the three-dimensional spatial position data as three-dimensional spatial position data of each unmanned aerial vehicle target at the corresponding acquisition time of the wide-area acquisition picture;
the intelligent identification mechanism is respectively connected with the dynamic interference mechanism and the position conversion mechanism and is used for intelligently identifying three-dimensional space position data corresponding to the next time of the same unmanned aerial vehicle target by adopting an AI identification model based on the three-dimensional space position data of the visual detection mechanism and the three-dimensional space position data corresponding to the next time of the same unmanned aerial vehicle target before the next time of the current time;
the signal triggering mechanism is respectively connected with the intelligent identification mechanism and the dynamic interference mechanism and is used for sending a mechanism early warning signal at the current moment when three-dimensional space position data corresponding to a certain unmanned aerial vehicle target at the next moment is positioned in the peripheral area, or sending a suspension early warning signal at the current moment;
The AI identification model is a feedforward neural network subjected to multiple learning, and the learning times of the feedforward neural network are positively correlated with the value of a set distance threshold;
when the body early warning signal is received, the signal interference action of the adjacent peripheral area with the set range is started, which comprises the following steps: the surrounding area of the adjacent set range is composed of each position of the distance to the dynamic interference mechanism within the set distance threshold, and the front end of the dynamic interference mechanism is used for transmitting an interference signal for interfering the unmanned aerial vehicle in the surrounding area.
2. The deep learning-based system for interfering with unmanned aerial vehicle communications of claim 1, wherein:
based on a plurality of three-dimensional space position data which respectively correspond to a plurality of acquisition moments before the next moment of the current moment, three-dimensional space position data of a visual detection mechanism, and interval duration between every two adjacent acquisition moments, intelligent identification of the three-dimensional space position data which corresponds to the next moment by adopting an AI identification model comprises the following steps: the plurality of acquisition moments before the next moment of the current moment comprise the current moment; the longer the interval duration between every two adjacent acquisition moments, the more the number of the plurality of acquisition moments;
The front end of the dynamic interference mechanism is configured to send an interference signal that interferes with the unmanned aerial vehicle in the peripheral area, where the interference signal includes: and the transmitting power of the interference signal is higher than or equal to a set power threshold.
3. The deep learning based system for interfering with drone communications of claim 2, wherein the system further comprises:
and the wireless notification mechanism is connected with the signal triggering mechanism and is used for transmitting the three-dimensional space position data corresponding to the unmanned aerial vehicle target at the next moment to a remote cloud computing server through a wireless communication link when the signal triggering mechanism transmits the mechanism early warning signal.
4. The deep learning based system for interfering with drone communications of claim 3, further comprising:
the cloud computing server is connected with each wireless notification mechanism corresponding to each dynamic interference mechanism respectively and is used for receiving the three-dimensional space position data sent by each wireless notification mechanism through a wireless communication link.
5. The deep learning based system for interfering with drone communications of claim 2, wherein the system further comprises:
The model composition mechanism is connected with the intelligent identification mechanism and used for carrying out repeated learning on the feedforward neural network to obtain the feedforward neural network after repeated learning, and sending the feedforward neural network to the intelligent identification mechanism as an AI identification model for use;
the parameter storage mechanism is connected with the model composition mechanism and used for storing various model parameters of the AI identification model;
the method for learning the feedforward neural network for multiple times to obtain the feedforward neural network after multiple times of learning, and sending the feedforward neural network to the intelligent authentication mechanism as an AI authentication model for use comprises the following steps: in each learning executed on the feedforward neural network, taking three-dimensional space position data corresponding to a certain unmanned aerial vehicle target at a certain past moment as output content of the feedforward neural network, taking a plurality of pieces of three-dimensional space position data corresponding to a plurality of acquisition moments of the certain unmanned aerial vehicle target before the certain past moment, three-dimensional space position data of a visual detection mechanism and interval duration between every two adjacent acquisition moments as a plurality of pieces of input content of the feedforward neural network, and executing the learning operation.
6. The deep learning based system for interfering with drone communications of claim 2, wherein the system further comprises:
And the power supply mechanism is connected with the dynamic interference mechanism and is used for providing power supply for the dynamic interference mechanism.
7. The deep learning based system for interfering with unmanned aerial vehicle communications of any of claims 2-6, wherein:
performing visual data acquisition of a large visual angle facing the peripheral area to obtain a wide-area acquisition picture, and transmitting an organism approximation signal to the dynamic interference mechanism when an unmanned aerial vehicle target with a scene depth smaller than a set depth limit exists in the wide-area acquisition picture comprises: the large visual angle is a visual acquisition visual angle with a numerical value exceeding or equal to a preset visual angle threshold;
wherein, execute the visual data acquisition of big visual angle in order to obtain wide region and gather the picture in the face of peripheral region, when there is the scene depth in wide region to be less than the unmanned aerial vehicle target of setting for depth of field limit in gathering the picture, send organism approaching signal to dynamic interference mechanism still includes: detecting more than one unmanned aerial vehicle target in the wide-area acquisition picture based on each standard outline pattern corresponding to each type of unmanned aerial vehicle;
wherein, detect more than one unmanned aerial vehicle target in the wide-area acquisition picture based on each standard appearance pattern that each type unmanned aerial vehicle corresponds respectively includes: the resolution of each standard outline pattern corresponding to each type of unmanned aerial vehicle is the same, and the definition of each standard outline pattern corresponding to each type of unmanned aerial vehicle is the same.
8. The deep learning based system for interfering with unmanned aerial vehicle communications of any of claims 2-6, wherein:
the visual detection mechanism is also used for sending a body far-away signal to the dynamic interference mechanism when no unmanned aerial vehicle target with the depth of view being less than the set depth of field limit exists in the wide-area acquisition picture;
the dynamic interference mechanism is also used for exiting the interference preparation state when the machine body far-away signal is received, and the dynamic interference mechanism in the interference preparation state is a dynamic interference mechanism for preparing to transmit an interference signal;
and when the dynamic interference mechanism receives the suspension early warning signal, ending/interrupting the signal interference action on the peripheral area of the adjacent set range.
9. The deep learning based system for interfering with unmanned aerial vehicle communications of any of claims 2-6, wherein:
based on a plurality of three-dimensional space position data which respectively correspond to a plurality of acquisition moments before the next moment of the current moment, three-dimensional space position data of a visual detection mechanism, and interval duration between every two adjacent acquisition moments, intelligent identification of the three-dimensional space position data which corresponds to the next moment by adopting an AI identification model comprises the following steps: the AI identification model is input in parallel by a plurality of pieces of three-dimensional space position data corresponding to a plurality of acquisition moments before the next moment of the current moment, three-dimensional space position data of a visual detection mechanism and interval duration between every two adjacent acquisition moments;
Wherein, adopt AI authentication model intelligence to appraise its three-dimensional space position data that corresponds at the moment of next moment based on its a plurality of three-dimensional space position data that correspond respectively of a plurality of collection moments before the moment of current moment, the three-dimensional space position data of visual detection mechanism and every interval duration between two adjacent collection moments includes: and executing the AI identification model to obtain three-dimensional space position data which is output by the AI identification model and corresponds to the AI identification model at the next moment.
10. A method of interfering with unmanned aerial vehicle communications based on deep learning, the method comprising:
when an organism approaching signal is received, controlling the dynamic interference mechanism to enter an interference preparation state;
a visual detection mechanism arranged at the front end of the dynamic interference mechanism is adopted to perform large-visual-angle visual data acquisition on a peripheral area so as to obtain a wide-area acquisition picture, and when an unmanned aerial vehicle target with a scene depth being smaller than a set depth of field limit exists in the wide-area acquisition picture, an organism approaching signal is sent;
acquiring each frame of wide-area acquisition pictures respectively acquired by the visual detection mechanism at each acquisition time at uniform intervals, determining three-dimensional space position data of each unmanned aerial vehicle target in each frame of wide-area acquisition picture based on depth of field, coordinate data of each unmanned aerial vehicle target in the wide-area acquisition picture and three-dimensional space position data of the visual detection mechanism, and outputting the three-dimensional space position data as three-dimensional space position data of each unmanned aerial vehicle target at the corresponding acquisition time of the wide-area acquisition picture;
After the dynamic interference mechanism enters an interference preparation state, aiming at the same unmanned aerial vehicle target, intelligently identifying the corresponding three-dimensional space position data at the next moment by adopting an AI identification model based on the three-dimensional space position data of the visual detection mechanism, the three-dimensional space position data of the visual detection mechanism and the interval duration between every two adjacent acquisition moments, wherein the three-dimensional space position data respectively correspond to the acquisition moments before the next moment of the current moment;
when a certain unmanned aerial vehicle target exists and the corresponding three-dimensional space position data of the unmanned aerial vehicle target at the next time is located in the peripheral area, a mechanism early warning signal is sent out at the current time, otherwise, a suspension early warning signal is sent out at the current time;
when an organism early warning signal is received, starting signal interference action on a peripheral area of a set range adjacent to the dynamic interference mechanism;
the AI identification model is a feedforward neural network subjected to multiple learning, and the learning times of the feedforward neural network are positively correlated with the value of a set distance threshold;
when the body early warning signal is received, the signal interference action of the peripheral area of the set range adjacent to the dynamic interference mechanism is started, which comprises the following steps: the peripheral area of the set range adjacent to the dynamic interference mechanism is composed of each position of which the distance to the dynamic interference mechanism is within the set distance threshold, and the front end of the dynamic interference mechanism is used for transmitting an interference signal for interfering the unmanned aerial vehicle in the peripheral area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410013749.3A CN117527135B (en) | 2024-01-04 | 2024-01-04 | System and method for interfering unmanned aerial vehicle communication based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410013749.3A CN117527135B (en) | 2024-01-04 | 2024-01-04 | System and method for interfering unmanned aerial vehicle communication based on deep learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117527135A true CN117527135A (en) | 2024-02-06 |
| CN117527135B CN117527135B (en) | 2024-03-22 |
Family
ID=89753495
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410013749.3A Active CN117527135B (en) | 2024-01-04 | 2024-01-04 | System and method for interfering unmanned aerial vehicle communication based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117527135B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106205217A (en) * | 2016-06-24 | 2016-12-07 | 华中科技大学 | Unmanned plane automatic testing method based on machine vision and unmanned plane method of control |
| CN109360240A (en) * | 2018-09-18 | 2019-02-19 | 华南理工大学 | A kind of small drone localization method based on binocular vision |
| US20210097664A1 (en) * | 2019-09-30 | 2021-04-01 | AO Kaspersky Lab | System and method for counteracting unmanned aerial vehicles |
| CN113268081A (en) * | 2021-05-31 | 2021-08-17 | 中国人民解放军32802部队 | Small unmanned aerial vehicle prevention and control command decision method and system based on reinforcement learning |
| US20210350162A1 (en) * | 2020-05-07 | 2021-11-11 | Skydio, Inc. | Visual observer for unmanned aerial vehicles |
| CN115297266A (en) * | 2022-10-10 | 2022-11-04 | 成都本原聚能科技有限公司 | Panoramic wide-scene-depth image acquisition system and image acquisition method |
| CN115390582A (en) * | 2022-07-15 | 2022-11-25 | 江西理工大学 | Point cloud-based multi-rotor unmanned aerial vehicle tracking and intercepting method and system |
| CN115902786A (en) * | 2022-11-07 | 2023-04-04 | 江西理工大学 | Acoustic tracking and anti-braking interception device, system and method for invading unmanned aerial vehicle |
-
2024
- 2024-01-04 CN CN202410013749.3A patent/CN117527135B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106205217A (en) * | 2016-06-24 | 2016-12-07 | 华中科技大学 | Unmanned plane automatic testing method based on machine vision and unmanned plane method of control |
| CN109360240A (en) * | 2018-09-18 | 2019-02-19 | 华南理工大学 | A kind of small drone localization method based on binocular vision |
| US20210097664A1 (en) * | 2019-09-30 | 2021-04-01 | AO Kaspersky Lab | System and method for counteracting unmanned aerial vehicles |
| US20210350162A1 (en) * | 2020-05-07 | 2021-11-11 | Skydio, Inc. | Visual observer for unmanned aerial vehicles |
| CN113268081A (en) * | 2021-05-31 | 2021-08-17 | 中国人民解放军32802部队 | Small unmanned aerial vehicle prevention and control command decision method and system based on reinforcement learning |
| CN115390582A (en) * | 2022-07-15 | 2022-11-25 | 江西理工大学 | Point cloud-based multi-rotor unmanned aerial vehicle tracking and intercepting method and system |
| CN115297266A (en) * | 2022-10-10 | 2022-11-04 | 成都本原聚能科技有限公司 | Panoramic wide-scene-depth image acquisition system and image acquisition method |
| CN115902786A (en) * | 2022-11-07 | 2023-04-04 | 江西理工大学 | Acoustic tracking and anti-braking interception device, system and method for invading unmanned aerial vehicle |
Non-Patent Citations (1)
| Title |
|---|
| 康凯;杨磊;李红艳;: "基于视觉显著性的小型无人机目标检测方法", 光学与光电技术, no. 03, 10 June 2020 (2020-06-10) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117527135B (en) | 2024-03-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112797846B (en) | Unmanned aerial vehicle prevention and control method and system | |
| CN110673626B (en) | GPS deception trapping method for unmanned aerial vehicle | |
| CN110719136B (en) | Unmanned aerial vehicle defense system | |
| US11800062B2 (en) | Systems, methods, apparatuses, and devices for radar-based identifying, tracking, and managing of unmanned aerial vehicles | |
| CN108037545B (en) | Directional optical interference system for unmanned aerial vehicle | |
| CN112068111A (en) | Unmanned aerial vehicle target detection method based on multi-sensor information fusion | |
| KR102292117B1 (en) | Drone control system and method for detecting and identifying of drone using the same | |
| CN112000133B (en) | Low-altitude aircraft/flyer identification system, counter-control system and identification method | |
| CN109282700A (en) | A kind of unmanned plane of cognitive radio is counter to chase orientation method and counter device | |
| CN110716583A (en) | Unmanned aerial vehicle counter-braking method and system, electronic equipment and storage medium | |
| CN109981212B (en) | Low-slow small prevention and control system and method based on data chain detection and countermeasures | |
| US20220075392A1 (en) | Reconnaissance unmanned aerial vehicle and surveillance flight method thereof | |
| CN114911267A (en) | Intelligent autonomous unmanned aerial vehicle counter-braking system | |
| CN113156417B (en) | Anti-unmanned aerial vehicle detection system, method and radar equipment | |
| CN108198402A (en) | Region Intelligent unattended machine detecting system | |
| CN111175737A (en) | Vehicle-mounted anti-unmanned aerial vehicle monitoring system | |
| CN113220010A (en) | Intelligent prevention and control system for unmanned aerial vehicle | |
| CN107943084A (en) | A kind of trailing type electromagnetic interference system and method for civilian multi-rotor unmanned aerial vehicle | |
| CN117527135B (en) | System and method for interfering unmanned aerial vehicle communication based on deep learning | |
| CN202958578U (en) | Bird situation monitoring and bird repelling system for airport | |
| CN112945015A (en) | Unmanned aerial vehicle monitoring system, method, device and storage medium | |
| CN115902786A (en) | Acoustic tracking and anti-braking interception device, system and method for invading unmanned aerial vehicle | |
| CN104267395A (en) | Method for achieving C-mode low-voice calling of traffic collision avoidance system | |
| KR20240021673A (en) | Object monitoring device providing information about unidentified objects | |
| CN115733575A (en) | Unmanned aerial vehicle detection and countercheck system and method |
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 |