A closed-loop management method and system for intelligent low-altitude air defense systems at airports

Abstract

This application relates to the field of airport airspace security and control technology, specifically to a closed-loop management method and system for an intelligent low-altitude airspace control system. The method includes a perception process, a discrimination process, an execution process, and an evaluation process. This application constructs a complete process of "perception -> discrimination -> execution -> evaluation," and maps the evaluation results back to the perception and discrimination modules. This enables the system to automatically adjust its parameters and strategies based on historical handling effectiveness data, thereby achieving continuous improvement and adaptation of its defense capabilities without human intervention when facing new or changing low-altitude threats. The perception process processes multi-source heterogeneous data through high-precision spatiotemporal fusion; the discrimination process integrates geographical and flight dynamic information for quantitative evaluation; and the execution process achieves multi-means coordinated response and collects feedback. This enables the system to form a high-precision closed-loop handling chain for low-altitude targets, significantly improving its overall response capability to complex threats.

Description

Technical Field

[0001] This application relates to the field of airport airspace safety and control technology, specifically to a closed-loop management method and system for an intelligent low-altitude control system for airports. Background Technology

[0002] In the existing field of low-altitude security, with the popularization of drone technology, the low-altitude security risks faced by airport areas are increasing. Existing technologies disclose a low-altitude collaborative monitoring system, which integrates multiple subsystems such as fire fighting, bird surveillance, and drone monitoring, utilizing a cloud platform to achieve collaborative management of multiple services, focusing on information sharing and integration among multiple departments under administrative management regulations. A low-altitude defense command and control system is also disclosed. This technical solution, by establishing weapon effectiveness models and tactical deployment models, realizes the detection, identification, analysis, and handling processes of low-altitude targets, with its core focusing on the deployment logic and strike effectiveness of tactical weapons.

[0003] However, the aforementioned existing technologies still have limitations in practical applications. Most existing solutions follow a linear logic from detection to countermeasure, lacking a quantitative feedback and effectiveness evaluation mechanism for the results of the response actions. This results in the system being unable to automatically adjust its prevention and control strategies based on the actual effects of the response. Furthermore, existing prevention and control methods often rely on manually preset fixed parameters. When faced with the complex and dynamically changing electromagnetic environment and meteorological conditions at airports, it is difficult to achieve real-time evolution and self-optimization of the prevention and control logic, resulting in insufficient quantitative management of the entire process and difficulty in dealing with highly intelligent low-altitude intrusion threats. Summary of the Invention

[0004] The purpose of this application is to provide a closed-loop management method for an airport low-altitude intelligent control system, including: a perception process, which involves real-time acquisition of multi-source heterogeneous detection data based on radar detection payloads, radio detection payloads, and photoelectric tracking payloads deployed in the airport perimeter and core area, and using a high-precision clock synchronization mechanism to perform spatial coordinate alignment and timestamp association processing on the multi-source heterogeneous detection data to generate a fused perception dataset; and a discrimination process, which, in response to the generation of the fused perception dataset, uses a deep neural network model to extract multi-dimensional features of low-altitude targets and performs target attribute recognition and behavioral intent prediction, and combines real-time access to airport geofence data and civil aviation flight operation dynamic information to perform quantitative assessment of the target threat level to generate prevention and control measures. The decision-making process involves: 1) Decision-making instructions; 2) Execution process: Based on the decision-making instructions, automatically or assistedly triggering a coordinated response process for electromagnetic signal suppression, protocol navigation deception, or physical interception devices targeting low-altitude targets, and real-time collecting closed-loop feedback information from the countermeasure execution mechanism and the controlled state data of the target to form a dynamic trajectory dataset; 3) Evaluation process: Based on the attribution comparison analysis of the dynamic trajectory dataset and the fused sensing dataset, quantitatively evaluating the response delay, accuracy, and impact range on the airport's electromagnetic environment of the response action to generate an evaluation feedback vector, and mapping the evaluation feedback vector inversely to the detection sensitivity parameters of the sensing process and the classification weight matrix of the discrimination process to achieve continuous convergence of the execution model and self-evolution of the prevention and control strategy.

[0005] By adopting the above technical solutions, a complete process of "perception -> discrimination -> execution -> evaluation" was constructed, and the evaluation results were back-mapped to the perception and discrimination modules. This enabled the system to automatically adjust its parameters and strategies based on historical response effectiveness data, thereby achieving continuous improvement and adaptation of defense capabilities without human intervention when facing new or changing low-altitude threats. The perception process processed multi-source heterogeneous data through high-precision spatiotemporal fusion, the discrimination process integrated geographical and flight dynamic information for quantitative evaluation, and the execution process realized multi-means coordinated response and collected feedback. This enabled the system to form a high-precision closed-loop response chain for low-altitude targets, significantly improving the overall response capability to complex threats. An independent evaluation process was established to quantitatively evaluate the time delay, accuracy, and environmental impact of response actions and generate feedback vectors. This ensured that the effect of each prevention and control action could be measured, providing a clear data-driven basis for system optimization.

[0006] Optionally, the clock synchronization mechanism in the sensing process includes: performing a millisecond-level time reference alignment process for multi-source heterogeneous detection data based on a two-level synchronization architecture constructed using a precise time protocol and a network time protocol; wherein, the sensing process specifically includes obtaining Coordinated Universal Time (UTC) information using a base clock source deployed at the sensing node and injecting it into the header of the original data stream of the detection payload, and realizing data fusion of radar reflectance data, radio spectrum feature data, and photoelectric pixel point cloud data on the same time axis by calculating the transmission delay compensation value from each detection payload to the data fusion center.

[0007] By adopting the above technical solution, a two-level synchronization architecture based on a precise time protocol and a network time protocol is used, and the transmission delay is compensated, so that detection data from radar, radio, photoelectric and other sources with different principles or different sampling frequencies can be aligned with the same time reference with millisecond-level accuracy.

[0008] Optionally, the discrimination process includes a target false alarm suppression process, specifically including: constructing an environmental background noise template based on the airport aircraft take-off and landing frequency and bird migration seasonal characteristics; using a moving target indication algorithm to filter out fixed ground object echoes and known aircraft features from the fused perception dataset; and using a multi-class support vector machine algorithm to identify and remove non-threatening interference targets caused by balloons, kites, or birds to reduce the system false alarm rate.

[0009] By adopting the above technical solution, a background noise template based on the specific airport environment (take-off and landing frequency, bird migration) is introduced in the discrimination process. The moving target indication algorithm and multi-class support vector machine algorithm are used for layer-by-layer filtering, which enables the system to intelligently distinguish between real threat targets (such as drones) and non-threat interference (such as birds, kites, and fixed objects), effectively reducing unnecessary alarms and resource waste, and ensuring that the normal operation of the airport is not interrupted by frequent false alarms.

[0010] Optionally, the quantitative assessment in the discrimination process includes a dynamic adjustment process for the threat weights, specifically including: establishing a multi-level protection zone model based on the geographical coordinates of the airport runway configuration, taxiway layout, and key airspace in the terminal area; responding to the physical distance, entry speed, and track deviation of the target relative to the multi-level protection zone model; calling the corresponding risk weight coefficients from the preset weight library in real time for weighted calculation; and dynamically adjusting the threat assessment weight factors for different time periods and different areas in combination with the current airport operation level.

[0011] By adopting the above technical solutions, threat assessment is strongly correlated with airport physical structure (runways, taxiways, critical airspace) and multi-dimensional dynamic factors (target distance, speed, track deviation, airport operation level), and the weight coefficients are dynamically called and adjusted. This enables the system to simulate the thinking of human security experts and to perform differentiated risk quantification for targets entering different areas with different behavioral patterns, thereby achieving precise and efficient allocation of prevention and control resources (such as countermeasures).

[0012] Optionally, the execution process includes a deterministic low-latency distribution process for executing countermeasure instructions, specifically including: preprocessing the prevention and control decision instructions using computing nodes deployed at the airport edge; mapping the prevention and control decision instructions into control primitives conforming to the underlying communication protocol of the execution agency by establishing a dedicated high-speed control link based on a virtual local area network; and using a task priority scheduling algorithm to ensure that the response time of the countermeasure execution agency after receiving the instructions remains within the millisecond range.

[0013] By adopting the above technical solution, decision commands are preprocessed at the airport edge and a dedicated high-speed control link is established to map the commands into underlying primitives that can be recognized by the execution agency. At the same time, combined with task priority scheduling, the end-to-end response time from decision generation to execution agency activation is strictly controlled at the millisecond level, solving the problem of processing delay caused by network latency or slow protocol conversion.

[0014] Optionally, the execution process also includes an interference protection process for the communication links of surrounding aircraft, specifically including: real-time acquisition of airport civil aviation radio frequency occupancy monitoring data, using beamforming technology to control the energy distribution during the electromagnetic signal suppression process, and setting null gain points in the direction of civil aviation dedicated frequency bands to avoid electromagnetic interference to legitimate aircraft communication and navigation links.

[0015] By adopting the above technical solution, the civil aviation frequency occupancy is monitored in real time during electromagnetic suppression, and beamforming technology is used for precise energy spatial pointing control to form a signal "zero trap" (extremely low gain point) in the civil aviation communication / navigation direction that needs to be protected. This allows the system to effectively suppress illegal drones while minimizing co-channel interference to the critical radio services of legitimate aircraft, thus ensuring the safety of the main airport operations.

[0016] Optionally, the effectiveness quantitative assessment in the evaluation process includes establishing a multi-dimensional indicator evaluation system, specifically including: a statistical unit for the success rate of target interception; an assessment unit for the impact of the execution and disposal process on the surrounding airspace traffic flow; a comparison unit for the energy consumption of the countermeasures and the coverage of the prevention and control; the evaluation process generates a comprehensive evaluation value that reflects the overall effectiveness of the prevention and control action by nonlinearly fusing the evaluation results of the above units.

[0017] By adopting the above technical solutions, the evaluation system incorporates multi-dimensional indicators such as the degree of impact on airspace traffic and the energy consumption and coverage of countermeasures. Through nonlinear fusion, a comprehensive evaluation value is generated, enabling the evaluation of single or phased prevention and control actions to comprehensively reflect the overall effectiveness of the prevention and control actions from multiple perspectives such as safety, operational efficiency, and economy, providing a more scientific basis for management decisions.

[0018] Optionally, the self-evolution in the evaluation process includes an automatic conversion process for the optimization parameters of the execution strategy. Specifically, this includes: establishing a correlation mapping function between the comprehensive evaluation value and the system configuration parameters; in response to the situation where the comprehensive evaluation value is lower than a preset threshold, extracting the corresponding failure feature dimension in the evaluation feedback vector; and using a genetic algorithm to search in the parameter space for the optimal combination of detection parameters and identification model weights that make the performance function tend to be optimal, thereby realizing closed-loop automatic optimization of the prevention and control strategy.

[0019] By adopting the above technical solution, a mapping relationship between evaluation results and system configuration parameters was established. When the performance is not up to standard, the genetic algorithm is automatically used to perform a global search for optimization in the parameter space. This enables the system to get rid of its dependence on manual experience in parameter tuning and continuously and automatically find and converge to the optimal or better working state in the current environment, truly realizing a self-evolutionary closed loop.

[0020] Optionally, the method further includes an evidence chain preservation process for the entire closed-loop management process, specifically including: collecting the original waveform data of the perception process, the reasoning logic log of the discrimination process, the control instruction snapshot of the execution process, and the performance analysis report of the evaluation process; using a hash algorithm to generate unique fingerprint information for the data at each stage; and writing it into a distributed blockchain ledger to ensure the tamper-proof nature and judicial traceability of evidence data in the prevention and control business process.

[0021] By adopting the above technical solution, the original data and process logs of the four key stages of perception, judgment, execution and evaluation are collected and solidified, and the tamper-proof fingerprints generated by the hash algorithm are stored in the blockchain. This ensures that all electronic evidence of the entire prevention and control operation is complete, authentic and non-repudiable, providing a technical evidence chain for post-event liability determination, accident investigation, compliance audit and possible legal proceedings.

[0022] The second objective of this application is to provide a closed-loop management system for an airport low-altitude intelligent control system, including a memory and a processor. The memory stores a computer program that can be loaded by the processor and executed as described above for the closed-loop management method of the airport low-altitude intelligent control system. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the closed-loop management method of the airport low-altitude intelligent control system in this application. Detailed Implementation

[0024] The following will describe in detail a closed-loop management method and system for an airport low-altitude intelligent control system, in conjunction with the accompanying drawings and specific embodiments. It should be noted that the following description is intended to provide those skilled in the art with sufficient information to implement this application, and the embodiments and parameters involved are merely illustrative examples and are not intended to limit the scope of protection of this application.

[0025] The embodiments of this application construct a closed-loop process from data perception, intelligent judgment, precise execution to performance evaluation, enabling the system to perform self-evaluation and dynamic optimization based on the actual effect of each action, thereby adapting to the complex and ever-changing low-altitude safety threats in airport airspace, and is particularly suitable for dealing with typical scenarios such as the increasing illegal intrusion of drones, disturbances by unidentified flying objects, and the risk of large bird strikes.

[0026] like Figure 1 As shown in the figure, this application provides a closed-loop management method for an airport low-altitude intelligent control system, including a perception process, a judgment process, an execution process, and an evaluation process.

[0027] S01: The perception process is a real-time acquisition process of multi-source heterogeneous detection data based on radar detection payloads, radio detection payloads, and photoelectric tracking payloads deployed in the airport perimeter and core area. A high-precision clock synchronization mechanism is used to perform spatial coordinate alignment and timestamp association processing on the multi-source heterogeneous detection data to generate a fused perception dataset.

[0028] S02: The discrimination process, in response to the generation of the fused sensing dataset, uses a deep neural network model to extract multi-dimensional features of low-altitude targets and performs target attribute recognition and behavioral intent prediction. It combines real-time access to airport geofence data and civil aviation flight operation dynamic information to conduct quantitative assessment of target threat level in order to generate prevention and control decision instructions.

[0029] S03: Execution process, based on the prevention and control decision instructions, automatically or assistedly triggers the coordinated response process of electromagnetic signal suppression, protocol navigation deception or physical interception devices against low-altitude targets, and collects the closed-loop feedback information of the countermeasure execution mechanism and the controlled status data of the target in real time to form a dynamic trajectory dataset for the disposal.

[0030] S04: Evaluation process, based on the attribution comparison analysis of the dynamic trajectory dataset and the fused sensing dataset, quantitatively evaluates the response delay, accuracy and impact range of the response action on the airport electromagnetic environment to generate an evaluation feedback vector, and then maps the evaluation feedback vector back to the detection sensitivity parameters of the sensing process and the classification weight matrix of the discrimination process to achieve continuous convergence of the model and self-evolution of the prevention and control strategy.

[0031] Specifically, the system achieves all-weather and all-airspace coverage through a network of sensors physically deployed at both ends of the airport runway, around the perimeter of the perimeter, and on the top of the terminal building. Among them, the radar detection payload can typically be a phased array radar with a minimum detection range of no more than 50 meters and a maximum detection range of no less than 15 kilometers, and a detection probability of no less than 90% for targets with a radar cross-section of 0.01 square meters. The radio detection payload operates in the frequency band commonly used by civilian UAVs from 400 MHz to 6 GHz, and can scan and analyze radio signals that conform to specific protocols in real time. Its spectrum scanning period can be set to the millisecond level. The electro-optical tracking payload includes a visible light and infrared dual-spectrum camera equipped with a gimbal device. Its optical zoom range is continuously adjustable from 1x to 30x, and its infrared thermal imaging resolution is no less than 320 x 240 pixels, which can provide clear thermal outline images of targets at night or in low visibility conditions. Understandably, a high-precision clock synchronization mechanism can ensure that the aforementioned radar point cloud data, radio spectrum fingerprint data, and optoelectronic video stream data are precisely aligned in the spatiotemporal dimension. This mechanism can utilize a master clock server deployed in the airport's core computer room, employing a two-tier synchronization network built with both precise time protocols and network time protocols, to distribute Coordinated Universal Time (UTC) signals to every sensing node, ensuring that the local clock deviation of all detection payloads is controlled within ±1 millisecond. During the data fusion phase, the system first integrates the target polar coordinates provided by the radar, the signal source azimuth angle provided by radio detection, and the image pixel coordinates provided by the optoelectronic system. The data is uniformly converted to an East-North-Sky Cartesian coordinate system with the center point of the airport's main runway as the origin. The East-North-Sky Cartesian coordinate system has axes pointing due east, due north, and vertically upward. The conversion process requires strict coordinate transformation calculations based on the installation position and attitude angle of each sensor. At the same time, each frame of data is timestamped to milliseconds based on clock synchronization information. Finally, a structured fusion sensing dataset is generated. The fusion sensing dataset includes at least 20 dimensions of field information, such as timestamp, globally unique target identifier, three-dimensional spatial coordinates, motion velocity vector, radar cross-section, radio signal signature, and infrared thermal radiation intensity.

[0032] It is understandable that the clock synchronization mechanism in the sensing process may include: a two-level synchronization architecture based on a precise time protocol and a network time protocol to perform a millisecond-level time reference alignment process for multi-source heterogeneous detection data; wherein, the sensing process specifically includes using the basic clock source deployed at the sensing node to obtain Coordinated Universal Time information and injecting it into the header of the raw data stream of the detection payload, and by calculating the transmission delay compensation value from each detection payload to the data fusion center, realizing the data fusion of radar reflectance data, radio spectrum feature data, and photoelectric pixel point cloud data on the same time axis.

[0033] Specifically, the first level of the two-tier synchronization architecture includes a highly stable master clock server located in the airport's core data center. This server obtains Coordinated Universal Time (UTC) by receiving signals from the Global Positioning System (GPS). Its internal clock crystal has a long-term stability better than ±0.1 seconds per day and serves as the root time source for the entire sensing network. The second level of synchronization is achieved by boundary clock devices deployed on aggregation switches in various physical areas. These boundary clocks maintain synchronization with the master clock server via fiber optic or Category 5e twisted-pair cables using a precise time protocol. The synchronization message sending interval can be configured between one-sixteenth and one-half a second, thus building a low-jitter and highly reliable clock distribution network within the airport. Each... The detection payload, radio detection payload, and photoelectric tracking payload all access the nearest boundary clock through their built-in ordinary clock client and obtain precise time via Network Time Protocol (NTP) or a dedicated hardware timestamp interface. During actual data injection, at the beginning of each detection cycle, the payload controller writes the currently acquired Coordinated Universal Time (UTC), accurate to microseconds, into a specific field in the protocol header of the data packet to be generated. This timestamp marks the absolute moment when data sampling begins. Due to differences in network path length and switch forwarding delay between different payloads to the data fusion center, the system calculates the time from leaving its network interface to arriving at the fusion center by sending measurement frames during the initialization phase. The one-way transmission delay of the fusion center server is typically between 0.5 and 5 milliseconds and is stored as a fixed compensation value in the system's configuration database. During data fusion, after receiving the original data packet with a header timestamp, the fusion center server first reads the timestamp and then adds a pre-defined transmission delay compensation value corresponding to the payload. This allows it to calculate the accurate absolute time of the physical event represented by the original data packet. For example, a detection data packet from a remote radar has a header timestamp of 13:30:15:123:456 microseconds and a transmission delay of 2.1 milliseconds. The fusion center then determines that the physical event corresponding to the detection data packet occurred at 13:00. At 30 minutes, 15 seconds, 125 milliseconds, and 556 microseconds; in this way, although radar data may have an inherent delay of several milliseconds due to signal processing, optoelectronic data may have a delay of tens of milliseconds due to image encoding, and radio data is almost real-time, after transmission delay compensation, the corresponding event times of all data are unified onto the same high-precision time axis. This allows subsequent algorithms to accurately correlate the observation results of different sensors on the same target at the same time. For example, it can correlate a high-speed moving point detected by radar within a certain 1 millisecond, a frequency-hopping signal detected by radio at the same time, and a small heat source in the same airspace captured by the optoelectronic system, thereby forming a more complete and reliable description of the target.

[0034] Understandably, the discrimination process may include the execution of a target false alarm suppression process, specifically including: constructing an environmental background noise template based on the airport's aircraft take-off and landing frequency and the seasonal characteristics of bird migration; using a moving target indication algorithm to filter out fixed ground object echoes and known aircraft features from the fused perception dataset; and using a multi-class support vector machine algorithm to identify and remove non-threatening interference targets caused by balloons, kites, or birds in order to reduce the system's false alarm rate.

[0035] Specifically, after receiving the fused sensing dataset, the discrimination process first initiates a false alarm suppression process to improve the accuracy of subsequent analysis. The construction of the environmental background noise template is a dynamic learning process. During normal operation without threat alarms, the system continuously records and statistically analyzes the sensing data. The template content includes, but is not limited to: the radar reflection characteristics of fixed buildings around the airport, such as control towers, hangars, and jet bridges, and their small fluctuation range with temperature changes; and a spatiotemporal distribution model of known civil aviation flight takeoff and landing routes established based on airport flight schedules and real-time broadcast automatic correlation surveillance data. This model can predict takeoff and landing routes at specific times... The probability and characteristics of legal aircraft appearing in specific airspaces; a seasonal bird activity hotspot map and typical trajectory database established based on historical bird observation data and phenological information, for example, during the spring migration season, flocks of birds may appear at specific altitudes within 5 kilometers east of the airport, with their flight speeds concentrated between 10 and 20 meters per second; a moving target indication algorithm identifies points whose positions change by comparing fused sensing datasets from multiple consecutive time frames. For points whose positions remain unchanged or whose change patterns are consistent with the swaying characteristics of buildings over multiple consecutive periods, the system marks them as fixed ground feature echoes and removes them from the list of targets to be processed; for those with already... Targets with known civil aviation flight identification and whose movement trajectories closely match the planned routes are also identified as legitimate targets and allowed to proceed without entering the threat assessment process. For suspicious targets remaining after the above filtering, the system initiates a multi-class support vector machine algorithm for fine classification. The input feature vectors of this multi-class support vector machine algorithm are extracted from the fused perception dataset, and the dimensions include: the mean and variance of the target's movement speed, the rate of change of radar cross-section with viewing angle, the contrast between infrared thermal radiation intensity and background, the curvature and coherence of the movement trajectory, and whether the radio signal has a specific modulation pattern. The support vector machine uses a large number of labeled samples in the offline phase. During training, the sample categories may include at least seven categories: "consumer drones," "industrial drones," "single birds," "bird flocks," "balloons," "kites," and "unknown debris." During online classification, the multi-class support vector machine algorithm calculates the distance from the input feature vector to the optimal classification hyperplane for each category and outputs the probability of it belonging to each category. When a target is classified as "single bird," "bird flock," "balloon," or "kite" with a confidence level exceeding 85%, the system determines it as a non-threatening interference target, generates a low-priority log record, but does not trigger a high-level alarm, thereby effectively reducing false alarms caused by natural environmental factors.

[0036] Understandably, the quantitative assessment in the discrimination process includes the dynamic adjustment of threat weights. Specifically, this includes: establishing a multi-level protection zone model based on the geographical coordinates of the airport runway configuration, taxiway layout, and key airspace in the terminal area; responding to the target's physical distance, entry speed, and track deviation relative to the multi-level protection zone model; calling the corresponding risk weight coefficients from the preset weight library in real time for weighted calculation; and dynamically adjusting the threat assessment weight factors for different time periods and different areas in combination with the current airport operation level.

[0037] Specifically, the multi-level protection zone model divides airport airspace into three core layers: the core zone is a rectangular airspace extending 150 meters to each side of the runway centerline and 1000 meters to each end, plus a three-dimensional area at an altitude of 0 to 200 meters. Intrusion into this zone directly endangers aircraft taking off or landing, and it has the highest risk weight. The buffer zone extends 500 meters beyond the core zone and includes airspace at an altitude of 200 to 500 meters. This area is crucial for aircraft approach and climb, and has the next highest risk weight. The surveillance zone encompasses the remaining airspace within a 5-kilometer radius of the airport perimeter and higher airspace. Entry into this zone... Early warning and tracking are conducted with the lowest risk weight. The system accurately marks the boundary coordinates of these areas on a digital map, forming electronic fences. When a target is confirmed as a potential threat during the identification process, its spatial relationship with these electronic fences is immediately calculated. The dynamic adjustment of threat weight is based on the calculation of a composite risk index, which is obtained by weighted summation of a base threat value, a distance attenuation factor, a speed enhancement factor, and a heading threat factor. The base threat value is determined according to the target classification results; for example, the base threat value for "suspicious drones" is set to 80, and for "verified drones" it is set to 60. The distance attenuation factor... The factor is calculated using a negative exponential function based on the distance from the target's centroid to the nearest core area boundary, decreasing by 15% for every 100 meters of distance increase. The velocity bonus factor is proportional to the target's ground velocity; when the velocity exceeds 30 meters per second, the velocity bonus factor begins to increase linearly to reflect the characteristic that high-speed targets are more difficult to intercept. The heading threat factor is determined by calculating the angle between the target's velocity vector and the vector pointing towards the nearest runway center point; the smaller the angle, the more directly the target is flying towards the runway, and the larger the heading threat factor value. The system's preset weight library assigns initial weight coefficients to the above factors, such as the distance factor weight. The weights are 0.4 for the speed factor and 0.3 for the heading factor. In addition, the system accesses airport operation level information in real time. When the airport is in peak operation period or the weather conditions are low visibility, the system will automatically increase the risk weight of the core area by 20% globally and lower the response threshold accordingly to improve the sensitivity of prevention and control. Conversely, during nighttime shutdown periods, the system can appropriately reduce the weights of the buffer zone and the monitoring zone to save prevention and control resources and reduce unnecessary alarms. This enables more refined and contextualized risk assessment, ensuring that limited prevention and control resources can be prioritized for the time and space with the highest risk.

[0038] It is understandable that the execution process includes a deterministic low-latency distribution process for countermeasure instructions, specifically including: preprocessing prevention and control decision instructions using computing nodes deployed at the airport edge; mapping prevention and control decision instructions into control primitives that conform to the underlying communication protocol of the execution agency by establishing a dedicated high-speed control link based on a virtual local area network; and using a task priority scheduling algorithm to ensure that the response time of the countermeasure execution agency after receiving the instructions remains within the millisecond range.

[0039] Specifically, the execution process begins with a prevention and control decision instruction generated by the discrimination module, which includes the target identifier, recommended countermeasures, and expected coordinates of the point of action. This instruction is first sent to edge computing nodes deployed near the target's physical location to reduce network transmission latency. The edge computing nodes run an instruction preprocessing engine, which stores a detailed protocol dictionary for all controllable countermeasures. This dictionary defines the control primitive formats, parameter ranges, and verification methods that each device can understand. For example, for a directional radio frequency suppression device, its control primitives might include a series of specific binary or text commands such as "set center frequency," "set transmit bandwidth," "set output power," "set beam pointing angle," and "start transmission." The preprocessing engine translates the abstract decision, such as "implement spectrum suppression on target X," into a sequence of control primitives arranged in a time sequence. To ensure the accuracy of the instructions... To ensure deterministic and low-latency command distribution, the system delineates a dedicated Virtual Local Area Network (VLAN) on the airport backbone network specifically for control signaling transmission. This VLAN has the highest quality of service priority, and its network switches are configured with strict priority queuing to minimize queuing delays for control packets. The task scheduler within the edge computing nodes assigns a priority label to each control primitive to be sent based on the urgency of the countermeasure action. The priority is dynamically calculated based on the target's threat level and the estimated arrival time in the critical area; the higher the threat level or the more urgent the time, the higher the command priority. The scheduler sends the control primitives to the corresponding execution agencies via a dedicated VLAN link in descending order of priority. Upon receiving the command, the execution agency, such as an RF suppressor, navigation decoy transmitter, or drone capture base station, immediately parses and executes it, sending back an acknowledgment frame including a reception timestamp and execution status to the edge node.

[0040] Understandably, the execution process also includes interference protection for communication links of surrounding aircraft. Specifically, this includes: acquiring real-time monitoring data on the occupancy of civil aviation radio frequencies at the airport; using beamforming technology to control the energy distribution during the electromagnetic signal suppression process; and setting null gain points in the direction of civil aviation dedicated frequency bands to avoid electromagnetic interference to legitimate aircraft communication and navigation links. When executing electromagnetic signal suppression countermeasures, the system strictly follows spectrum security guidelines, and this protection process is initiated synchronously with the suppression process.

[0041] Specifically, the system uses an independent spectrum monitoring network to scan and record in real time the signal strength and occupancy status of civil aviation-specific frequency bands such as high-frequency communication bands, instrument landing system bands, and microwave landing system bands around the airport. The monitoring data is updated to the edge computing nodes once per second. When directional radio frequency suppression needs to be initiated, the protection coordination module within the edge computing node first checks whether there are any currently in-use civil aviation communication or navigation channels near the target direction pointed to by the suppression device. If so, the coordination module calculates one or more silent angle intervals that need to be protected. Subsequently, the coordination module sends the suppression command along with the azimuth and elevation angles that need to be nulled to the radio frequency suppression device. The radio frequency suppression device typically uses a phased array antenna or a digital beamforming antenna array, and its beam pointing and shape can be flexibly controlled by software. The device generates suppression interference signals... Simultaneously, based on the received null angle information, the weighting coefficients of the antenna array are adjusted in real time. This ensures that while maximizing the main lobe gain of the interference beam pointing towards the target, the antenna forms an extremely low gain point, or null, in the spatial direction pointing towards the civil aviation communication link that needs protection. The depth of this null is typically required to be below -20 dB, sufficient to attenuate the interference energy leaking into that direction to below the harmless level stipulated by the International Telecommunication Union. For example, when it is necessary to suppress a drone intruding on the south side of an airport runway, while a flight is communicating with the control tower on the north side of the runway using VHF, the system will control the suppression antenna to form a null at the angle pointing north while pointing towards the target to the south. This ensures that the effective interference with the drone will not affect normal flight communication. The entire process is automated and requires no manual intervention, achieving both the countermeasure objective and strictly protecting the security of critical civil aviation radio services.

[0042] Understandably, the effectiveness quantification assessment in the evaluation process includes establishing a multi-dimensional indicator evaluation system, specifically including: a statistical unit for the success rate of intercepting the target; an assessment unit for the impact of the execution and disposal process on the surrounding airspace traffic flow; a comparison unit for the energy consumption of the countermeasures and the coverage of the prevention and control; the evaluation process generates a comprehensive evaluation value that reflects the overall effectiveness of the prevention and control action by nonlinearly fusing the evaluation results of the above units.

[0043] Specifically, after each response operation, the evaluation module is activated to conduct a comprehensive quantitative review of the operation. The target interception success rate statistics unit compares the dynamic trajectory dataset and the fused perception dataset. The specific calculation logic is as follows: within a certain time window after the response command is issued, if the target's motion state changes from "uncontrolled" to "controlled," including but not limited to signal loss, trajectory deviation, hovering, or forced landing, and this change is highly correlated with the effectiveness of the countermeasures in time and space, it is judged as a successful interception. The success rate is calculated as the ratio of the number of successful interceptions to the total number of responses. This statistics unit also further subdivides the success rates of different countermeasures against different types of targets. The assessment unit for the impact on surrounding airspace traffic flow analyzes abnormal changes in the trajectories of civil aviation flights in the pre-set monitoring airspace during and after the response operation by accessing automatic dependent surveillance (ADS) data. For example, whether any flights have performed unplanned detours, climbs, or descents due to the control operation, and calculates the number of affected flights, total delay time, and other factors. The average flight path deviation distance is quantified into an impact coefficient ranging from 0 to 1, with a higher coefficient indicating a greater negative impact. A unit comparing the energy consumption of countermeasures with the coverage of the control measures records the working parameters and duration of each countermeasure operation, such as the transmission power and duration of radio frequency suppression, the emission energy of laser equipment, and the number of sorties and flight distances of net-trapping drones, converting these into standardized energy consumption points. Simultaneously, based on the dynamic trajectory data, the actual airspace volume or area protected by the operation is calculated. This comparison unit outputs an energy efficiency ratio, i.e., the effective protected space obtained per unit of energy consumption. Finally, the evaluation module uses a pre-trained nonlinear fusion model, typically a three-layer feedforward neural network trained based on historical expert rating data, to input the aforementioned success rate, impact coefficient, energy efficiency ratio, and other primary indicators, outputting a comprehensive evaluation value between 0 and 100. This comprehensive evaluation value directly reflects the overall effectiveness of the control operation in terms of effectiveness, safety, and economy.

[0044] It is understandable that the self-evolution in the evaluation process includes the automatic conversion process of the execution strategy optimization parameters. Specifically, this includes: establishing a correlation mapping function between the comprehensive evaluation value and the system configuration parameters; in response to the situation where the comprehensive evaluation value is lower than the preset threshold, extracting the corresponding failure feature dimension in the evaluation feedback vector; and using a genetic algorithm to search in the parameter space for the optimal combination of detection parameters and identification model weights that make the efficiency function tend to be optimal, thereby realizing the closed-loop automatic optimization of the prevention and control strategy.

[0045] Specifically, the evaluation feedback vector is a structured data object that, in addition to the comprehensive evaluation value, records detailed scores for each primary indicator and key features analyzed from the handling data, such as "excessive delay in this response," "low target classification confidence leading to misjudgment," and "excessive leakage of suppression energy in non-target directions." The system maintains a parameter knowledge base, which defines optimizable system configuration parameters and their allowable adjustment ranges. For example, the radar detection threshold during perception can be adjusted to ±10 dB, the motion detection sensitivity of the photoelectric system can be adjusted to 0 to 100, the penalty factor of the support vector machine classifier during discrimination can be adjusted to 0.1 to 10, and the weight matrix of a specific layer in the neural network. The system presets a comprehensive evaluation value threshold, such as 70 points. When the evaluation of an action falls below this threshold, the self-evolutionary process is triggered. The system first analyzes the evaluation feedback vector to locate the main failure dimension leading to the low score, assuming that the analysis indicates "low target classification confidence" is the main cause. Then, the system constructs a classification-based... Accuracy is the core performance function; then, a genetic algorithm is launched to optimize parameters; the genetic algorithm encodes a set of system parameters into an individual, and the initial population is generated by the current parameter configuration and its random perturbation; each individual is decoded into a specific set of parameter values. In the simulation environment or historical data playback, the system applies this set of parameters to run the discrimination process and calculates its classification accuracy in the current failure scenario as a fitness value. By selecting individuals with high fitness for crossover and mutation operations, a new population is generated iteratively. After dozens to hundreds of generations of evolution, the algorithm will converge to a new set of parameter combinations that can improve the classification accuracy in similar scenarios; after optimization, the system automatically generates parameter update instructions. After security review, the new radar detection threshold, classifier parameters, etc. are pushed to the corresponding perception and discrimination modules in a gradual or timed batch processing manner to complete a strategy iteration; this process runs periodically, enabling the system to continuously learn from successful and failed experiences, gradually adapt to new threat patterns and complex environments, and achieve autonomous evolution of prevention and control capabilities.

[0046] It is understood that the method in this application embodiment also includes an evidence chain storage process for the entire closed-loop management process, specifically including: collecting the original waveform data of the perception process, the reasoning logic log of the discrimination process, the control instruction snapshot of the execution process, and the performance analysis report of the evaluation process; using a hash algorithm to generate unique fingerprint information of the data at each stage; and writing it into a distributed blockchain ledger to ensure the tamper-proof nature and judicial traceability of evidence data in the prevention and control business process.

[0047] Specifically, the system performs data preservation at the source of data generation. For the perception process, the preservation subsystem extracts a data segment after the radar intermediate frequency signal completes analog-to-digital conversion, after the radio detection equipment completes spectral energy calculation, and after the optoelectronic equipment completes image frame capture. This data, along with its precise timestamp and device serial number, is then used to generate a fixed-length digital fingerprint using a secure hash algorithm. For the discrimination process, the system records the model version number loaded by the inference engine, the input feature vector, the key output probabilities of the intermediate layer, and the final classification and threat assessment conclusions. These logical logs are serialized and their hash values ​​are calculated. For the execution process, the preservation captures every control primitive issued from the edge computing node and every status confirmation frame returned by the actuator. For the evaluation process, the preservation objects are a complete performance analysis report and self-evolving decision records. All these hash values ​​generated at different stages and at different times... The data is packaged into a chronological evidence block, which also includes the hash value of the previous block to form a chain structure. This evidence block is then submitted to a private blockchain network jointly maintained by multiple trusted nodes, including airport management agencies, air traffic control departments, and equipment suppliers. After verifying the validity of the block through a consensus mechanism, the nodes in the network append it to their respective copies of the blockchain ledger. Once the data is recorded on the chain, due to its hashing characteristics and distributed storage, any tampering with the original data will cause its hash value to change, thus making it inconsistent with the on-chain record and easily detectable. When traceability is required, the authorized party can query the blockchain ledger, quickly locate all relevant evidence hashes for a specific event according to the time index, and compare and verify them with the locally securely stored original data copy, thereby completely and reliably reproducing the entire process details of the prevention and control event, meeting high standards of security auditing and regulatory compliance requirements.

[0048] This application also discloses an airport low-altitude intelligent prevention and control closed-loop management system, including a memory and a processor. The memory stores a computer program that can be loaded by the processor and executed by the closed-loop management method of the airport low-altitude intelligent prevention and control system.

[0049] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device, and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0050] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0051] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0052] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0053] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.

[0054] The above are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Any feature disclosed in this specification (including the abstract and drawings) may be replaced by other equivalent or similar features unless specifically stated otherwise. That is, unless specifically stated otherwise, each feature is only one example of a series of equivalent or similar features.

Claims

1. A closed-loop management method for an airport low-altitude intelligent control system, characterized in that, include: The perception process involves the real-time acquisition of multi-source heterogeneous detection data based on radar detection payloads, radio detection payloads, and photoelectric tracking payloads deployed in the airport perimeter and core area. A high-precision clock synchronization mechanism is used to perform spatial coordinate alignment and timestamp association processing on the multi-source heterogeneous detection data to generate a fused perception dataset. The discrimination process, in response to the generation of the fused perception dataset, uses a deep neural network model to extract multi-dimensional features of low-altitude targets and performs target attribute recognition and behavioral intent prediction. It combines real-time access to airport geofence data and civil aviation flight operation dynamic information to conduct quantitative assessment of target threat level in order to generate prevention and control decision instructions. During the execution process, the coordinated response process of electromagnetic signal suppression, protocol navigation deception, or physical interception devices against low-altitude targets is automatically or assisted in triggering based on the prevention and control decision instructions, and the closed-loop feedback information of the countermeasure execution mechanism and the controlled status data of the target are collected in real time to form a dynamic trajectory dataset for the disposal. The evaluation process, based on the attribution comparison analysis of the dynamic trajectory dataset and the fused sensing dataset, quantifies the effectiveness of the response delay, accuracy, and impact range on the airport electromagnetic environment of the response action to generate an evaluation feedback vector. The evaluation feedback vector is then back-mapped to the detection sensitivity parameters of the sensing process and the classification weight matrix of the discrimination process to achieve continuous convergence of the execution model and self-evolution of the control strategy.

2. The closed-loop management method of the airport low-altitude intelligent control system according to claim 1, characterized in that, The clock synchronization mechanism in the sensing process includes: a two-level synchronization architecture based on a precise time protocol and a network time protocol to perform a millisecond-level time reference alignment process for multi-source heterogeneous detection data; wherein, the sensing process specifically includes using the basic clock source deployed at the sensing node to obtain Coordinated Universal Time information and injecting it into the header of the original data stream of the detection payload, and calculating the transmission delay compensation value from each detection payload to the data fusion center to achieve data fusion of radar reflectance data, radio spectrum feature data, and photoelectric pixel point cloud data on the same time axis.

3. The closed-loop management method of the airport low-altitude intelligent control system according to claim 1, characterized in that, The discrimination process includes a target false alarm suppression process, which specifically includes: constructing an environmental background noise template based on the airport's aircraft take-off and landing frequency and the seasonal characteristics of bird migration; using a moving target indication algorithm to filter out fixed ground object echoes and known aircraft features from the fused perception dataset; and using a multi-class support vector machine algorithm to identify and remove non-threatening interference targets caused by balloons, kites, or birds to reduce the system's false alarm rate.

4. The closed-loop management method of the airport low-altitude intelligent control system according to claim 1, characterized in that, The quantitative assessment in the discrimination process includes a dynamic adjustment process for the execution threat weights. Specifically, it includes: establishing a multi-level protection zone model based on the geographical coordinates of the airport runway configuration, taxiway layout, and key airspace in the terminal area; responding to the target's physical distance, entry speed, and track deviation relative to the multi-level protection zone model; calling the corresponding risk weight coefficients from the preset weight library in real time for weighted calculation; and dynamically adjusting the threat assessment weight factors for different time periods and different areas in combination with the current airport operation level.

5. The closed-loop management method of the airport low-altitude intelligent control system according to claim 1, characterized in that, The execution process includes a deterministic low-latency distribution process for executing countermeasure instructions, specifically including: preprocessing the prevention and control decision instructions using computing nodes deployed at the airport edge; mapping the prevention and control decision instructions into control primitives that conform to the underlying communication protocol of the execution agency by establishing a dedicated high-speed control link based on a virtual local area network; and using a task priority scheduling algorithm to ensure that the response time of the countermeasure execution agency after receiving the instructions remains within the millisecond range.

6. The closed-loop management method of the airport low-altitude intelligent control system according to claim 5, characterized in that, The execution process also includes an interference protection process for the communication links of surrounding aircraft, specifically including: real-time acquisition of airport civil aviation radio frequency occupancy monitoring data, using beamforming technology to control the energy distribution during the electromagnetic signal suppression process, and setting null gain points in the direction of civil aviation dedicated frequency bands to avoid electromagnetic interference to legitimate aircraft communication and navigation links.

7. The closed-loop management method of the airport low-altitude intelligent control system according to claim 1, characterized in that, The effectiveness quantification assessment in the evaluation process includes establishing a multi-dimensional indicator evaluation system, specifically including: a statistical unit for the success rate of target interception; an assessment unit for the impact of the execution and disposal process on the surrounding airspace traffic flow; and a comparison unit for the energy consumption of countermeasures and the coverage of prevention and control. The evaluation process generates a comprehensive evaluation value that reflects the overall effectiveness of the prevention and control action by nonlinearly fusing the evaluation results of the above units.

8. The closed-loop management method of the airport low-altitude intelligent control system according to claim 7, characterized in that, The self-evolution in the evaluation process includes an automatic conversion process for optimizing execution strategy parameters. Specifically, it includes: establishing a correlation mapping function between the comprehensive evaluation value and the system configuration parameters; in response to the situation where the comprehensive evaluation value is lower than a preset threshold, extracting the corresponding failure feature dimension in the evaluation feedback vector; and using a genetic algorithm to search in the parameter space for the optimal combination of detection parameters and identification model weights that make the performance function tend to be optimal, thereby realizing closed-loop automatic optimization of the prevention and control strategy.

9. The closed-loop management method of the airport low-altitude intelligent control system according to claim 1, characterized in that, The method also includes an evidence chain storage process for the entire closed-loop management process, specifically including: collecting the original waveform data of the perception process, the reasoning logic log of the discrimination process, the control instruction snapshot of the execution process, and the performance analysis report of the evaluation process; using a hash algorithm to generate unique fingerprint information for the data at each stage; and writing it into a distributed blockchain ledger to ensure the tamper-proof nature and judicial traceability of evidence data in the prevention and control business process.

10. A closed-loop management system for an airport low-altitude intelligent control system, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executed as described in any one of claims 1 to 9, the closed-loop management method of the airport low-altitude intelligent control system.

Images