Wireless communication system and method for nuclear radiation scene monitoring based on unmanned aerial vehicle

Abstract

The invention discloses a wireless communication system and method for nuclear radiation scene monitoring based on an unmanned aerial vehicle, and belongs to the technical field of unmanned aerial vehicle application, and the system comprises a cooperative monitoring network composed of an unmanned aerial vehicle end, a wireless communication relay layer and a ground operation end. The unmanned aerial vehicle end carries a high-sensitivity radiation measurement device, a Lora wireless unvarnished transmission device and a Beidou positioning module, and the radiation measurement device is connected to the Lora wireless unvarnished transmission device through a military aviation plug interface and an industrial RS485 cable. Monomer radiation measurement equipment is integrated with a Lora wireless transparent transmission device through a military-grade aviation interface to 485 cable, the use scene of the radiation measurement equipment is increased, the problem that the use scene of the radiation measurement equipment is limited is effectively solved, a matched software system is developed, and the application range of the radiation measurement equipment is expanded. According to the system, the radiation data and the coordinate information of the Beidou satellite navigation system can be visually displayed in real time, so that visual and real-time monitoring information is provided for a user.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicle (UAV) applications, and in particular relates to a wireless communication system and method for monitoring nuclear radiation scenes based on UAVs. Background Art

[0002] The currently used drone radiation monitoring systems face several problems that limit their performance and application scope. The use scenarios of single devices are single, and ordinary single detection equipment does not have an interface for real-time transmission of detection information to the outside world, resulting in a single and restricted use scenario of the equipment. The transmission distance is limited. Existing wireless communication technologies are mostly based on Wi-Fi or Bluetooth, and their coverage is usually limited to a short distance, with a radius of less than 100 meters. Such limitations make it impossible to effectively carry out monitoring work when large-scale monitoring is required in the event of a nuclear accident, and the deployment is complicated. In order to expand the monitoring range, additional relay equipment is usually required, which not only increases costs, but also makes the deployment and maintenance of the entire system more complicated.

[0003] Existing single-unit detection equipment has a single usage scenario and can usually only be used for handheld use and detection. It has poor scalability, and the read data cannot be uploaded automatically and can only be recorded and reported manually. Traditional wireless communication technologies Wi-Fi / Bluetooth and 4G / 5G, Wi-Fi / Bluetooth has a short transmission distance, a radius of less than 100 meters, and weak penetration ability, and cannot pass through concrete buildings. 4G / 5G networks rely on base station coverage, and nuclear accident scenarios are prone to signal interruption. Data integration is insufficient. Existing systems mostly use independent modules to process radiation data and positioning information, resulting in time synchronization errors.

[0004] Based on this, the present invention designs a wireless communication system and method for nuclear radiation scene monitoring based on drones to solve the above problems. Summary of the Invention

[0005] The purpose of the present invention is to solve the problems that the existing single-unit detection equipment has a single usage scenario and can usually only be used for handheld use and detection, and has poor scalability. The read data cannot be automatically uploaded and can only be recorded and reported manually. Traditional wireless communication technologies Wi-Fi / Bluetooth and 4G / 5G, Wi-Fi / Bluetooth have a short transmission distance, a radius of less than 100 meters, weak penetration ability, and cannot pass through concrete buildings. 4G / 5G networks rely on base station coverage, and nuclear accident scenarios are prone to signal interruption and insufficient data integration. Existing systems mostly use independent modules to process radiation data and positioning information, resulting in time synchronization errors. A wireless communication system and method for nuclear radiation scene monitoring based on drones is proposed.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A wireless communication system for monitoring nuclear radiation scenes based on drones, comprising a collaborative monitoring network consisting of a drone terminal, a wireless communication relay layer, and a ground operation terminal;

[0008] The drone is equipped with a high-sensitivity radiation measurement device, a Lora wireless transparent transmission device, and a Beidou positioning module. The radiation measurement device is connected to the Lora wireless transparent transmission device via a military-grade aviation plug-in interface and an industrial RS485 cable, and is integrated with a Beidou positioning module for spatial positioning.

[0009] The ground operation terminal is based on a Windows tablet computer and establishes a physical link with the receiving end of the Lora wireless transparent transmission device through a USB to 485 adapter.

[0010] As a further description of the above technical solution:

[0011] The physical link supports baud rate adjustment, has optoelectronic isolation protection, and is surge-proof with voltage ≥ 1500V. After the USB port is inserted into a Windows tablet, the corresponding driver is installed. The Windows system assigns a virtual COM port number. The differential signal voltage range is -7V to +12V. It uses twisted-pair shielded cable for connection, with a terminal resistance of 120Ω. It supports hot-plug detection, meets electromagnetic compatibility standards, and a dedicated software system developed in conjunction with it supports multi-source data fusion and real-time visualization.

[0012] As a further description of the above technical solution:

[0013] In the physical link, the Lora wireless transparent transmission device serves as the core hub and adopts a star networking mode to realize the collaborative operation of multiple drones. On the drone side, the radiation measurement equipment encapsulates the radiation dose rate, nuclide type and Beidou coordinates into customized data frames through the 410-525MHz frequency band and sends them to the ground side at a rate of once every 2 seconds.

[0014] As a further description of the above technical solution:

[0015] The ground operation software uses multi-threaded parsing technology to synchronously process radiation data streams and map information, and adopts an adaptive packet loss compensation algorithm to perform packet loss detection based on dual verification of serial numbers and timestamps. It dynamically selects and forward-predicts radiation data and spatially interpolated map data based on the packet loss location to ensure data integrity.

[0016] As a further description of the above technical solution:

[0017] The UAV end is powered by a large-capacity lithium battery, and the ground operation end is also equipped with an ultra-large-capacity mobile lithium battery power supply to ensure continuous operation capabilities in extreme environments.

[0018] As a further description of the above technical solution:

[0019] The military-grade aviation plug interface is integrated with the industrialized RS485 cable. The military-grade aviation plug interface has a built-in shockproof silicone sleeve and an impact resistance rating of IP67.

[0020] The radiation measurement device is connected to a military-grade aviation plug-in interface to export the measurement data of the radiation measurement device and forward it to the Windows tablet in real time through a physical data link, thereby improving the versatility of the radiation measurement device.

[0021] Based on the Lora communication protocol optimization of the Lora wireless transparent transmission device, it uses the 410-525MHz frequency band, the maximum transmission power is 20dBm, the transmission rate is 19200bps, and the bit error rate is <0.01%;

[0022] The software data fusion algorithm synchronizes the radiometric data with the BeiDou coordinates through time stamps. Data outside this window is considered asynchronous and discarded or triggers a compensation mechanism, using a linear regression model to predict and compensate for clock drift.

[0023] As a further description of the above technical solution:

[0024] The adaptive packet loss compensation algorithm includes network status perception, dynamic compensation strategy and resource optimization allocation;

[0025] In the dynamic network perception described above, the algorithm dynamically evaluates the current network quality by monitoring multiple parameters such as network packet loss rate, delay and jitter in real time. It triggers a compensation mechanism when the packet loss rate exceeds a set threshold, and reduces redundant operations to save resources when the network is stable.

[0026] As a further description of the above technical solution:

[0027] In the dynamic compensation strategy, the algorithm adaptively selects a compensation method based on the type and severity of packet loss. For short-term packet loss, interpolation is used to generate approximate data packets. For long-term packet loss, redundancy or forward error correction technology is used to recover the packet.

[0028] In the resource optimization allocation, in resource-constrained scenarios, the algorithm prioritizes the transmission of key data packets in radiation data and map data, and balances bandwidth and quality by dynamically adjusting the encoding rate, frame rate or compression ratio.

[0029] As a further description of the above technical solution:

[0030] The adaptive packet loss compensation algorithm predicts the content of lost data packets through a recommendation model. The recommendation model is formed by combining a long short-term memory network and an adversarial network. The long short-term memory network takes the behavior pattern of the drone end in response to the radiation data scene as input and outputs a long-term and short-term dynamic sequence of the drone end's map data scene;

[0031] In the long short-term memory network, L=(l1,l2,…,l n ) represents a massive radiation data scene, x t represents the input gate, y t represents the forget gate, z t Represents the output gate, Q t Represents a memory unit, which is used to schedule and optimize the memory and past radiation data information in the long-term memory unit;

[0032] In the long short-term memory network, the memory unit Q at each time step t is t The value is passed to the input gate x t Adjusted transition data and the Forgotten Gate t The adjusted memory data Q of the previous time step t-1 t-1 , and the combination process expression is:

[0033]

[0034] The input of the new radiation data scene is obtained by the input gate x t Control, the forget gate of the old radiation data scene is controlled by the forget gate, when the memory unit Q is corrected t After that, the output gate m t The expression is:

[0035] m t =α(W o Q t +V o h t-1 )

[0036] Where α(·) represents the hyperbolic tangent function, W o represents the output weight, V o represents the cycle weight, h t-1 Represents the hidden state at the previous time point;

[0037] The input sample of the adversarial network is the long-term and short-term dynamic sequence m of the radiation data scene monitored by the drone output by the long-term and short-term memory network model. t , train the input samples through the adversarial network to generate the typical form of the adversarial network structure;

[0038] In the recommendation process of massive radiation data scenes, the massive radiation data scenes are used as the training set of the recommendation model, and L=(l1,l2,…,l n ) description, where n is the total number of samples in L, assuming that the radiation data scene on the drone side has a certain distribution relationship P r , assuming that the long-term and short-term dynamic sequence m of the radiation data scene monitored by the drone t According to Gaussian distribution P0, m t As the input of the generator, the output result is a fake radiation data scene sample that meets P r Distribution, using deep neural network to build P0 to P r The mapping relationship;

[0039] The role of the generator K is to simulate and generate false data samples that are highly similar to the radiation data scene of the drone end, making it difficult for the discriminator to distinguish whether the input data is true or false, thereby increasing the difficulty of the model's discrimination. Therefore, the ultimate goal of the generator K is to achieve the loss function Minimize, the formula is as follows

[0040]

[0041] The role of the discriminator D is to distinguish whether its input is a true data scene sample or a false data scene sample, and l i Data scenario S with the same dimension D As the input of the discriminator D, then S D It may be a true data scene sample, or it may be a false data scene sample output by the generator K. The output result of the generator K is a scalar value between 0 and 1, which is used to represent the probability that the input of D is a true data scene sample or a false data scene sample. Since the discriminator needs to distinguish between true data scene samples and false data scene samples as much as possible, its training purpose is to ensure that At the same time, Max, the objective function of the discriminator is as follows:

[0042]

[0043] The training of a generative adversarial network is equivalent to a binary zero-sum game process. Its generator strives to reduce the distribution difference between the created fake data scenarios and the real data scenarios to confuse the discriminator, while the discriminator needs to work hard to learn and identify the distribution difference between the two and do its best to distinguish them. The objective function of this game process is expressed as:

[0044]

[0045] From the objective function of the above formula, we can see that the generator K attempts to generate data scenarios that meet the actual user's interests, thereby making it impossible for the discriminator D to distinguish whether the input data scenario is the actual radiation data scenario monitored by the drone end or a fake data scenario.

[0046] A wireless communication method for monitoring nuclear radiation scenes based on an unmanned aerial vehicle, the wireless communication method comprising:

[0047] System startup:

[0048] The ground operation software initializes the communication parameters;

[0049] The Lora wireless transparent transmission device on the drone performs a self-test. The normal operating voltage of the Lora wireless transparent transmission device is greater than 24V and the signal strength is greater than -110dBm.

[0050] Data collection and transmission:

[0051] The radiometer samples once every 2 seconds and sends a data frame via RS485. The format of the data frame is 0xAA+dose value+CRC16;

[0052] Beidou positioning module output statement;

[0053] Data fusion and display:

[0054] The software parsing module extracts the latitude and longitude and dose values;

[0055] Abnormal alarm:

[0056] When the dose rate is detected to be >10μGy / h, the software will alarm and record the GPS coordinates.

[0057] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0058] 1. The present invention can achieve stable communication beyond line of sight, thereby overcoming the limitations of traditional wireless communication distance. A single radiation measurement device is integrated with a Lora wireless transparent transmission device through a military-grade aviation plug-in interface to a 485 cable, which increases the use scenarios of the radiation measurement device and effectively solves the problem of limited use scenarios of the radiation measurement device. A supporting software system is developed, which can visualize the radiation data and the coordinate information of the Beidou satellite navigation system in real time, thereby providing users with intuitive and real-time monitoring information.

[0059] 2. In the present invention, when predicting the long-term and short-term dynamic changes of the radiation data from the drone end, the behavior pattern of the drone end at each time step t is used as the input radiation data scene. The behavior pattern can be regarded as the monitoring record summary of the radiation data scene by the drone end in a specific time period, and the output gate of the long short-term memory network model outputs m t, that is, the long-term and short-term dynamic sequences of the radiation data scenes on the drone side, so as to more accurately capture the changing trends of the radiation data scenes monitored by the drone side. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 This is a flow chart of a wireless communication system and method for monitoring nuclear radiation scenes based on drones proposed by the present invention;

[0061] Figure 2 This is a schematic diagram of a collaborative monitoring network in a wireless communication system and method for monitoring nuclear radiation scenes based on drones proposed in the present invention;

[0062] Figure 3 This is a schematic diagram of the connection between a Windows tablet computer and a Lora wireless transparent transmission device in a wireless communication system and method for monitoring nuclear radiation scenes based on drones proposed by the present invention;

[0063] Figure 4 This is a schematic diagram of an adaptive packet loss compensation algorithm in a wireless communication system and method for monitoring nuclear radiation scenes based on drones proposed by the present invention. DETAILED DESCRIPTION

[0064] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0065] Please see the attached Figure 1 -Attached Figure 4 , the present invention provides a technical solution: a wireless communication system for nuclear radiation scene monitoring based on a UAV, comprising a collaborative monitoring network consisting of a UAV terminal, a wireless communication relay layer, and a ground operation terminal;

[0066] The drone is equipped with a high-sensitivity radiation measurement device, a Lora wireless transparent transmission device, and a Beidou positioning module. The radiation measurement device is connected to the Lora wireless transparent transmission device via a military-grade aviation plug-in interface and an industrial RS485 cable, and is integrated with a Beidou positioning module for spatial positioning.

[0067] The ground operation terminal is based on a Windows tablet computer and establishes a physical link with the receiving end of the Lora wireless transparent transmission device through a USB to 485 adapter.

[0068] As a further description of the above technical solution:

[0069] The physical link supports baud rate adjustment, has optoelectronic isolation protection, and is surge-proof with voltage ≥ 1500V. After the USB port is inserted into a Windows tablet, the corresponding driver is installed. The Windows system assigns a virtual COM port number, which complies with the TIA / EIA-485-A standard. The differential signal voltage range is -7V to +12V, and it uses twisted-pair shielded cable for connection. The terminal resistance is 120Ω, hot-plug detection is supported, and electromagnetic compatibility meets standards. The specially developed dedicated software system supports multi-source data fusion and real-time visualization.

[0070] As a further description of the above technical solution:

[0071] In the physical link, the Lora wireless transparent transmission device serves as the core hub and adopts a star networking mode to realize the collaborative operation of multiple drones. On the drone side, the radiation measurement equipment encapsulates the radiation dose rate, nuclide type and Beidou coordinates into customized data frames through the 410-525MHz frequency band and sends them to the ground side at a rate of once every 2 seconds.

[0072] As a further description of the above technical solution:

[0073] The ground operation software uses multi-threaded parsing technology to synchronously process radiation data streams and map information, and adopts an adaptive packet loss compensation algorithm to perform packet loss detection based on dual verification of serial numbers and timestamps. It dynamically selects and forward-predicts radiation data and spatially interpolated map data based on the packet loss location to ensure data integrity.

[0074] As a further description of the above technical solution:

[0075] The UAV end is powered by a large-capacity lithium battery, and the ground operation end is also equipped with an ultra-large-capacity mobile lithium battery power supply to ensure continuous operation capabilities in extreme environments.

[0076] As a further description of the above technical solution:

[0077] The military-grade aviation plug interface is integrated with the industrialized RS485 cable. The military-grade aviation plug interface has a built-in shockproof silicone sleeve and an impact resistance rating of IP67.

[0078] The radiation measurement device is connected to a military-grade aviation plug-in interface to export the measurement data of the radiation measurement device and forward it to the Windows tablet in real time through a physical data link, thereby improving the versatility of the radiation measurement device.

[0079] Based on the Lora communication protocol optimization of the Lora wireless transparent transmission device, the 410-525MHz frequency band is used;

[0080] The software data fusion algorithm synchronizes the radiometric data with the BeiDou coordinates through time stamps. Data outside this window is considered asynchronous and discarded or triggers a compensation mechanism, using a linear regression model to predict and compensate for clock drift.

[0081] As a further description of the above technical solution:

[0082] The adaptive packet loss compensation algorithm includes network status perception, dynamic compensation strategy and resource optimization allocation;

[0083] In the dynamic network perception described above, the algorithm dynamically evaluates the current network quality by monitoring multiple parameters such as network packet loss rate, delay and jitter in real time. It triggers a compensation mechanism when the packet loss rate exceeds a set threshold, and reduces redundant operations to save resources when the network is stable.

[0084] As a further description of the above technical solution:

[0085] In the dynamic compensation strategy, the algorithm adaptively selects a compensation method based on the type and severity of packet loss. For short-term packet loss, interpolation is used to generate approximate data packets. For long-term packet loss, redundancy or forward error correction technology is used to recover the packet.

[0086] In the resource optimization allocation, in resource-constrained scenarios, the algorithm prioritizes the transmission of key data packets in radiation data and map data, and balances bandwidth and quality by dynamically adjusting the encoding rate, frame rate or compression ratio.

[0087] As a further description of the above technical solution:

[0088] The adaptive packet loss compensation algorithm predicts the content of lost data packets through a recommendation model. The recommendation model is formed by combining a long short-term memory network and an adversarial network. The long short-term memory network takes the behavior pattern of the drone end in response to the radiation data scene as input and outputs a long-term and short-term dynamic sequence of the drone end's map data scene;

[0089] In the long short-term memory network, L=(l1,l2,…,l n ) represents a massive radiation data scene, x t represents the input gate, y t represents the forget gate, z t Represents the output gate, Q t Represents a memory unit, which is used to schedule and optimize the memory and past radiation data information in the long-term memory unit;

[0090] In the long short-term memory network, the memory unit Q at each time step t is t The value is passed to the input gate x t Adjusted transition data and the Forgotten Gate tThe adjusted memory data Q of the previous time step t-1 t-1 , and the combination process expression is:

[0091]

[0092] The input of the new radiation data scene is obtained by the input gate x t Control, the forget gate of the old radiation data scene is controlled by the forget gate, when the memory unit Q is corrected t After that, the output gate m t The expression is:

[0093] m t =α(W o Q t +V o h t-1 )

[0094] Where α(·) represents the hyperbolic tangent function, W o represents the output weight, V o represents the cycle weight, h t-1 Represents the hidden state at the previous time point;

[0095] The input sample of the adversarial network is the long-term and short-term dynamic sequence m of the radiation data scene monitored by the drone output by the long-term and short-term memory network model. t , train the input samples through the adversarial network to generate the typical form of the adversarial network structure;

[0096] In the recommendation process of massive radiation data scenes, the massive radiation data scenes are used as the training set of the recommendation model, and L=(l1,l2,…,l n ) description, where n is the total number of samples in L, assuming that the radiation data scene on the drone side has a certain distribution relationship P r , assuming that the long-term and short-term dynamic sequence m of the radiation data scene monitored by the drone t According to Gaussian distribution P0, m t As the input of the generator, the output result is a fake radiation data scene sample that meets P r Distribution, using deep neural network to build P0 to P r The mapping relationship;

[0097] The role of the generator K is to simulate and generate false data samples that are highly similar to the radiation data scene of the drone end, making it difficult for the discriminator to distinguish whether the input data is true or false, thereby increasing the difficulty of the model's discrimination. Therefore, the ultimate goal of the generator K is to achieve the loss function Minimize, the formula is as follows

[0098]

[0099] The role of the discriminator D is to distinguish whether its input is a true data scene sample or a false data scene sample, and l i Data scenario S with the same dimension D As the input of the discriminator D, then S D It may be a true data scene sample, or it may be a false data scene sample output by the generator K. The output result of the generator K is a scalar value between 0 and 1, which is used to represent the probability that the input of D is a true data scene sample or a false data scene sample. Since the discriminator needs to distinguish between true data scene samples and false data scene samples as much as possible, its training purpose is to ensure that At the same time, Max, the objective function of the discriminator is as follows:

[0100]

[0101] The training of a generative adversarial network is equivalent to a binary zero-sum game process. Its generator strives to reduce the distribution difference between the created fake data scenarios and the real data scenarios to confuse the discriminator, while the discriminator needs to work hard to learn and identify the distribution difference between the two and do its best to distinguish them. The objective function of this game process is expressed as:

[0102]

[0103] From the objective function of the above formula, we can see that the generator K attempts to generate data scenarios that meet the actual user's interests, thereby making it impossible for the discriminator D to distinguish whether the input data scenario is the actual radiation data scenario monitored by the drone end or a fake data scenario.

[0104] In this embodiment:

[0105] Communication performance improvement:

[0106] The transmission distance reaches 5 kilometers in urban environments and 10 kilometers in open areas, which is 50 times higher than Wi-Fi.

[0107] The power consumption is lower in standby mode, down to 0.1W, supporting 24 hours of continuous operation;

[0108] Reliability enhancements:

[0109] The plug-in and pull-out life of the aviation plug interface is greater than 10,000 times, and the cable tensile strength is greater than 50N;

[0110] The data packet loss compensation mechanism achieves a data integrity rate of over 95%.

[0111] A wireless communication method for monitoring nuclear radiation scenes based on an unmanned aerial vehicle, the wireless communication method comprising:

[0112] System startup:

[0113] The ground operation software initializes the communication parameters;

[0114] The Lora wireless transparent transmission device on the drone performs a self-test. The normal operating voltage of the Lora wireless transparent transmission device is greater than 24V and the signal strength is greater than -110dBm.

[0115] Data collection and transmission:

[0116] The radiometer samples once every 2 seconds and sends a data frame via RS485. The format of the data frame is 0xAA+dose value+CRC16;

[0117] Beidou positioning module output statement;

[0118] Data fusion and display:

[0119] The software parsing module extracts the latitude and longitude and dose values;

[0120] Abnormal alarm:

[0121] When the dose rate is detected to be >10μGy / h, the software will alarm and record the GPS coordinates;

[0122] This embodiment includes:

[0123] Hardware connection: the Lora wireless transparent transmission device on the drone is physically connected to the radiometer through a military-grade aviation plug interface, and the operation end is connected to the Windows tablet through a USB to 485 adapter;

[0124] Software protocol, custom data frame structure, including radiation value, longitude and latitude and CRC check code, frame interval 50ms.

[0125] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A wireless communication system for monitoring nuclear radiation scenes based on drones, characterized in that: It includes a collaborative monitoring network consisting of a UAV terminal, a wireless communication relay layer, and a ground operation terminal; The drone is equipped with a high-sensitivity radiation measurement device, a Lora wireless transparent transmission device, and a Beidou positioning module. The radiation measurement device is connected to the Lora wireless transparent transmission device via a military-grade aviation plug-in interface and an industrial RS485 cable, and is integrated with a Beidou positioning module for spatial positioning. The ground operation terminal is based on a Windows tablet computer and establishes a physical link with the receiving end of the Lora wireless transparent transmission device through a USB to 485 adapter.

2. The wireless communication system for nuclear radiation scene monitoring based on an unmanned aerial vehicle according to claim 1, characterized in that: The physical link supports baud rate adjustment, has optoelectronic isolation protection, and is surge-proof with voltage ≥ 1500V. After the USB port is inserted into a Windows tablet, the corresponding driver is installed. The Windows system assigns a virtual COM port number. The differential signal voltage range is -7V to +12V. It uses twisted-pair shielded cable for connection, with a terminal resistance of 120Ω. It supports hot-plug detection, meets electromagnetic compatibility standards, and a dedicated software system developed in conjunction with it supports multi-source data fusion and real-time visualization.

3. The wireless communication system for nuclear radiation scene monitoring based on an unmanned aerial vehicle according to claim 2, characterized in that: In the physical link, the Lora wireless transparent transmission device serves as the core hub and adopts a star networking mode to realize the collaborative operation of multiple drones. On the drone side, the radiation measurement equipment encapsulates the radiation dose rate, nuclide type and Beidou coordinates into customized data frames through the 410-525MHz frequency band and sends them to the ground side at a rate of once every 2 seconds.

4. The wireless communication system for nuclear radiation scene monitoring based on an unmanned aerial vehicle according to claim 3, characterized in that: The ground operation software uses multi-threaded parsing technology to synchronously process radiation data streams and map information, and adopts an adaptive packet loss compensation algorithm to perform packet loss detection based on dual verification of serial numbers and timestamps. It dynamically selects and forward-predicts radiation data and spatially interpolated map data based on the packet loss location to ensure data integrity.

5. The wireless communication system for nuclear radiation scene monitoring based on an unmanned aerial vehicle according to claim 4, characterized in that: The UAV end is powered by a large-capacity lithium battery, and the ground operation end is also equipped with an ultra-large-capacity mobile lithium battery power supply to ensure continuous operation capabilities in extreme environments.

6. The wireless communication system for nuclear radiation scene monitoring based on an unmanned aerial vehicle according to claim 5, characterized in that: The military-grade aviation plug interface is integrated with the industrialized RS485 cable. The military-grade aviation plug interface has a built-in shockproof silicone sleeve and an impact resistance level of IP67. The radiation measurement device is connected to a military-grade aviation plug-in interface to export the measurement data of the radiation measurement device and forward it to the Windows tablet in real time through a physical data link, thereby improving the versatility of the radiation measurement device. Based on the Lora communication protocol optimization of the Lora wireless transparent transmission device, the 410-525MHz frequency band is used; The software data fusion algorithm synchronizes the radiometric data with the BeiDou coordinates through time stamps. Data outside this window is considered asynchronous and discarded or triggers a compensation mechanism, using a linear regression model to predict and compensate for clock drift.

7. The wireless communication system for nuclear radiation scene monitoring based on an unmanned aerial vehicle according to claim 6, characterized in that: The adaptive packet loss compensation algorithm includes network status perception, dynamic compensation strategy and resource optimization allocation; In the dynamic network perception described above, the algorithm dynamically evaluates the current network quality by monitoring multiple parameters such as network packet loss rate, delay and jitter in real time. It triggers a compensation mechanism when the packet loss rate exceeds a set threshold, and reduces redundant operations to save resources when the network is stable.

8. The wireless communication system for nuclear radiation scene monitoring based on an unmanned aerial vehicle according to claim 7, characterized in that: In the dynamic compensation strategy, the algorithm adaptively selects a compensation method based on the type and severity of packet loss. For short-term packet loss, interpolation is used to generate approximate data packets. For long-term packet loss, redundancy or forward error correction technology is used to recover the packet. In the resource optimization allocation, in resource-constrained scenarios, the algorithm prioritizes the transmission of key data packets in radiation data and map data, and balances bandwidth and quality by dynamically adjusting the encoding rate, frame rate or compression ratio.

9. The wireless communication system for nuclear radiation scene monitoring based on an unmanned aerial vehicle according to claim 8, characterized in that: The adaptive packet loss compensation algorithm predicts the content of lost data packets through a recommendation model. The recommendation model is formed by combining a long short-term memory network and an adversarial network. The long short-term memory network takes the behavior pattern of the drone end in response to the radiation data scene as input and outputs a long-term and short-term dynamic sequence of the drone end's map data scene; In the long short-term memory network, L=(l1, l2, ..., l n ) represents a massive radiation data scene, x t represents the input gate, y t represents the forget gate, z t Represents the output gate, Q t Represents a memory unit, which is used to schedule and optimize the memory and past radiation data information in the long-term memory unit; In the long short-term memory network, the memory unit Q at each time step t is t The value is passed to the input gate x t Adjusted transition data and the Forgotten Gate t The adjusted memory data Q of the previous time step t-1 t-1 , and the combination process expression is: The input of the new radiation data scene is obtained by the input gate x t Control, the forget gate of the old radiation data scene is controlled by the forget gate, when the memory unit Q is corrected t After that, the output gate m t The expression is: m t =α(W o Q t +V o h t-1 ) Where α(·) represents the hyperbolic tangent function, W o represents the output weight, V o represents the cycle weight, h t-1 Represents the hidden state at the previous time point; The input sample of the adversarial network is the long-term and short-term dynamic sequence m of the radiation data scene monitored by the drone output by the long-term and short-term memory network model. t , train the input samples through the adversarial network to generate the typical form of the adversarial network structure; In the recommendation process of massive radiation data scenes, the massive radiation data scenes are used as the training set of the recommendation model, and L = (l1, l2, ..., l n ) description, where n is the total number of samples in L, assuming that the radiation data scene on the drone side has a certain distribution relationship P r , assuming that the long-term and short-term dynamic sequence m of the radiation data scene monitored by the drone t According to Gaussian distribution P0, m t As the input of the generator, the output result is a fake radiation data scene sample that meets P r Distribution, using deep neural network to build P0 to P r The mapping relationship; The role of the generator K is to simulate and generate false data samples that are highly similar to the radiation data scene of the drone end, making it difficult for the discriminator to distinguish whether the input data is true or false, thereby increasing the difficulty of the model's discrimination. Therefore, the ultimate goal of the generator K is to achieve the loss function Minimize, the formula is as follows The role of the discriminator D is to distinguish whether its input is a true data scene sample or a false data scene sample, and l i Data scenario S with the same dimension D As the input of the discriminator D, then S D It may be a true data scene sample, or it may be a false data scene sample output by the generator K. The output result of the generator K is a scalar value between 0 and 1, which is used to represent the probability that the input of D is a true data scene sample or a false data scene sample. Since the discriminator needs to distinguish between true data scene samples and false data scene samples as much as possible, its training purpose is to ensure that At the same time, Max, the objective function of the discriminator is as follows: The training of a generative adversarial network is equivalent to a binary zero-sum game process. Its generator strives to reduce the distribution difference between the created fake data scenarios and the real data scenarios to confuse the discriminator, while the discriminator needs to work hard to learn and identify the distribution difference between the two and do its best to distinguish them. The objective function of this game process is expressed as: From the objective function of the above formula, we can see that the generator K attempts to generate data scenarios that meet the actual user's interests, thereby making it impossible for the discriminator D to distinguish whether the input data scenario is the actual radiation data scenario monitored by the drone end or a fake data scenario.

10. A wireless communication method for monitoring nuclear radiation scenes based on drones, a wireless communication system for monitoring nuclear radiation scenes based on drones according to claim 9, characterized in that: The wireless communication method includes: System startup: The ground operation software initializes the communication parameters; The Lora wireless transparent transmission device on the drone performs a self-test. The normal operating voltage of the Lora wireless transparent transmission device is greater than 24V and the signal strength is greater than -110dBm. Data collection and transmission: The radiometer samples once every 2 seconds and sends a data frame via RS485. The format of the data frame is OxAA+dose value+CRC16; Beidou positioning module output statement; Data fusion and display: The software parsing module extracts the latitude and longitude and dose values; Abnormal alarm: When the dose rate is detected to be greater than 10 μGy / h, the software will alarm and record the GPS coordinates.

Images