Unmanned aircraft integrated comprehensive supervision system and method thereof

Abstract

The invention discloses an unmanned aircraft integrated comprehensive supervision system and a method thereof, and relates to the technical field of unmanned aircraft management. The system synchronously broadcasts standard 5G synchronization signal blocks, receives airspace target echoes, and generates a local detection spectrum; carrying out symbol-level fusion processing on the local detection spectrums of the plurality of base stations, carrying out joint spectral deviation minimization processing by adopting an improved spatial smoothing multiple signal classification algorithm, and extracting the position and speed of a target unmanned aerial vehicle; identifying the identity legality and the flight type of the target unmanned aerial vehicle; the sensing capacity is dynamically estimated, and flow scheduling is executed during overload; generating a flight instruction or triggering a security linkage measure according to an identification result; and processing task load data returned by the unmanned aerial vehicle and carrying out anomaly detection. According to the method, the coverage and precision of synchronous sensing of the low-altitude airspace unmanned aerial vehicle can be improved, accurate identification and quick response to illegal flight behaviors are realized, and the airspace safety supervision capability in a complex environment is enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of data acquisition and processing, and in particular to a high-reliability lossless data acquisition and processing system and method. Background Art

[0002] In recent years, unmanned aerial vehicles (UAVs) have been widely used in logistics and distribution, urban management, environmental monitoring, and security inspections, creating new demands for low-altitude airspace management and monitoring technologies. Existing UAV monitoring methods primarily include ground-based radar detection, wireless positioning from base stations, and active broadcast identification by aircraft. These methods are supplemented by single-point data detection or rule-based abnormal behavior detection to achieve UAV positioning, identification, and management. These traditional methods have, to a certain extent, supported the regulatory needs of early, small-scale UAV applications and formed a foundational flight monitoring system.

[0003] However, with the rapid growth of drones in low-altitude airspace and the increasing complexity of their application scenarios, existing monitoring technologies are gradually exposing multiple limitations. First, ground-based sensing methods have limited airspace coverage, blind spots, and insufficient synchronization accuracy, making it difficult to achieve high-precision detection of multiple targets simultaneously. Second, multi-base station collaborative sensing relies solely on simple post-detection fusion, lacking underlying symbol-level data fusion optimization. This leads to significant deviations in joint detection results, impacting the accuracy of target parameter extraction. Furthermore, traditional drone identification methods rely on static broadcasts or single-point data analysis, lacking intelligent classification and legitimacy determination mechanisms based on historical track features. Regarding system capacity management, existing technologies generally lack dynamic capacity estimation and traffic scheduling strategies based on real-time sensing results, making them unable to effectively address overload issues in high-density flight scenarios. Finally, in terms of anomaly detection, existing methods primarily rely on single-sensor thresholds or fixed models, lacking in-depth structured analysis of drone payload data and dynamic adaptive detection mechanisms. Consequently, anomaly identification accuracy and real-time performance are insufficient.

[0004] Therefore, it is urgent to propose an integrated comprehensive supervision system that can realize low-altitude and wide-coverage synchronous perception, symbol-level data fusion optimization, flight identification based on track characteristics, capacity dynamic management and scheduling control, and intelligent anomaly detection based on mission payload data, so as to improve the overall efficiency, safety and intelligence level of UAV airspace management. Summary of the Invention

[0005] The purpose of the present invention is to provide an integrated unmanned aerial vehicle (UAV) supervision system and method thereof, in response to the problems existing in existing UAV supervision systems in terms of airspace perception coverage, data fusion accuracy, flight identification intelligence, capacity dynamic management, and mission anomaly detection, so as to achieve low-altitude and wide-coverage synchronous perception, symbol-level data fusion optimization, flight identification based on track characteristics, capacity dynamic management and scheduling control, and intelligent anomaly detection of mission payload data.

[0006] The present invention achieves the above-mentioned purpose through the following technical solutions:

[0007] In one aspect, the present invention provides an integrated unmanned aerial vehicle (UAV) monitoring system, comprising:

[0008] The base station sensing module is used to synchronously broadcast the standard 5G synchronization signal block (SSB) and receive the target echo in the airspace to generate a local detection spectrum.

[0009] A data fusion module is used to perform symbol-level fusion processing on the local detection spectra of multiple base stations and extract the position and velocity of the UAV target based on the fusion results. The data fusion module adopts a symbol-level fusion method and performs joint spectrum deviation minimization processing based on an improved spatial smoothing multiple signal classification (MUSIC) algorithm.

[0010] Capacity management module, which is used to dynamically estimate the base station sensing capacity and perform drone traffic scheduling when overloaded;

[0011] The flight identification module is used to combine the fusion detection results and the regulatory database to identify the identity legitimacy and flight category of the target drone;

[0012] The dispatch control module is used to issue flight instructions or trigger security linkage measures based on target classification results;

[0013] The intelligent analysis module is used to process the mission payload data sent back by the drone and perform anomaly detection.

[0014] A further improvement of the present invention is that the base station sensing module includes:

[0015] A multi-beam transmitting unit, configured to synchronously broadcast a standard 5G synchronization signal block (SSB) in a predetermined beam sequence;

[0016] An echo receiving unit is used to switch to a receiving mode after each transmission cycle to receive an echo signal of an airspace target;

[0017] The time synchronization unit is used to perform timing synchronization control based on the 5G NR system synchronization signal or the global navigation satellite system signal, so that the time synchronization error between base stations is controlled within the set threshold;

[0018] A signal processing unit, used for pre-processing and frequency domain transformation of the received wavelet;

[0019] The spatial spectrum estimation unit is used to perform spatial spectrum estimation on the processed signal based on the improved spatial smoothing multiple signal classification MUSIC algorithm to generate local detection spectrum data.

[0020] A further improvement of the present invention is that the improved spatial smoothing multiple signal classification algorithm includes:

[0021] Two-dimensional parameter spectrum estimation is introduced, and a dual-parameter detection spectrum is constructed based on the joint search of distance parameters and speed parameters;

[0022] Applying composite spatial smoothing, the array space is divided into multiple overlapping sub-arrays, and the frequency space is divided into sub-bands, and the sub-array covariance matrices are generated and averaged respectively;

[0023] The covariance matrix is reconstructed based on multi-order statistics, and the covariance matrix is corrected by fusing the statistical information of each sub-array and sub-band under finite sample conditions.

[0024] Adaptive optimization of noise subspace, dynamically setting the division threshold between noise feature subspace and signal feature subspace according to the local detection spectrum energy;

[0025] Adaptive scanning with local energy sensing limits the search area based on the initial detection of spectral energy distribution, and only performs a refined search in areas where the spectral energy exceeds the set threshold;

[0026] Target relocalization mechanism, when there are multiple detection peaks, re-estimates the initially detected target parameters based on the subspace matched filter and outputs the revised target parameters;

[0027] Parameter estimation error compensation, based on the joint spectral gradient calculation of the initial estimated point correction, update the final estimated values of the distance parameter and velocity parameter.

[0028] A further improvement of the present invention is that the data fusion module includes:

[0029] An input interface unit, configured to receive local detection spectrum data from a plurality of base stations;

[0030] The joint spectrum calculation unit is used to calculate the joint detection spectrum according to the following formula:

[0031]

[0032] Where: P m (r, v) represents the local detection spectrum calculated by the m-th base station based on the radial distance parameter r and the radial velocity parameter v; P sum(r, v) represents the joint detection spectrum obtained by fusing the local detection spectra of multiple base stations; M represents the number of base stations involved in the fusion process;

[0033] The deviation suppression unit is used to introduce spectrum deviation minimization processing based on the joint detection spectrum, and fusion estimate the target parameters. The fusion estimation formula is:

[0034]

[0035] Where: λ represents the spectrum deviation suppression weight factor, which is used to suppress the impact of single base station measurement errors on the fusion results; represents the radial distance parameter estimation value and radial velocity parameter estimation value of the target UAV obtained by fusion estimation;

[0036] The output interface unit is used to output the radial distance parameters and radial velocity parameters of the target UAV obtained by fusion estimation.

[0037] A further improvement of the present invention is that the capacity management module includes:

[0038] The drone number calculation unit is used to calculate the maximum number of drones N that can be supported based on the dynamic estimation of the base station's carrying capacity. max , where N max Calculated by the following formula:

[0039]

[0040] Where: γ avail To effectively perceive the signal-to-noise ratio, is the ambient noise power, γ min is the minimum detection threshold;

[0041] Control unit, used to calculate the maximum number of drones N max , perform real-time traffic control, when the number of drones approaches N max When the detection is complete, delay or guidance control is performed to avoid system overload and ensure detection feasibility.

[0042] A further improvement of the present invention is that the flight identification module includes:

[0043] A track feature extraction unit is used to extract the speed change rate, track curvature and flight altitude change rate as track feature parameters based on the historical track data of the UAV target;

[0044] a classifier unit, configured to perform category determination on the track characteristic parameters based on a trained machine learning model, wherein the machine learning model comprises a support vector machine (SVM) model or a convolutional neural network (CNN) model;

[0045] A data comparison unit is used to compare the identification result of the target with the flight permission information and no-fly zone information in the supervision database to verify the target's ID number, flight permission status and the legitimacy of the flight zone;

[0046] The output unit is used to output the legality classification results and flight category labels of the target drone.

[0047] A further improvement of the present invention is that the scheduling control module includes:

[0048] The illegal flight emergency response unit is used to automatically generate flight trajectory optimization instructions when illegal or abnormal drones approach, and control the drone to adjust its speed and heading to avoid entering restricted areas;

[0049] Illegal flight warning unit, used to compare flight targets with the geographical restriction database in real time, trigger illegal flight warnings and initiate relevant handling mechanisms;

[0050] The path adjustment unit activates alternative flight path planning or adjusts the scheduling order of flight missions to avoid congested areas when the flight density is high.

[0051] A further improvement of the present invention is that the intelligent analysis module includes:

[0052] The data processing unit is used to receive multimodal sensor data and fuse the different sensor data using a weighted decision algorithm. The calculation result formula is as follows:

[0053]

[0054] Where: w i is the feature weight of each sensing modality, f i (I) is the modal characteristic function, and the anomaly score D is calculated by weighted fusion; N is the total number of sensing modalities involved in weighted fusion;

[0055] The dynamic evaluation unit is used to set a dynamic threshold according to the fusion evaluation value and perform anomaly detection on the sensor data to determine whether the system status is abnormal.

[0056] The exception handling unit is used to initiate predetermined emergency measures and feedback alarm information when a system abnormality is detected.

[0057] A further improvement of the present invention is that the intelligent analysis module further includes:

[0058] Mission data parsing unit, used to perform structured analysis on mission payload data sent back by the target UAV and extract multi-dimensional feature sets;

[0059] A dynamic feature selection unit, configured to dynamically determine a feature subset for anomaly detection based on real-time statistical characteristics of the extracted multidimensional feature set;

[0060] An adaptive threshold setting unit, used to dynamically adjust the decision threshold of anomaly detection based on the distribution characteristics of feature subsets;

[0061] The anomaly recognition unit is used to perform anomaly detection on the mission payload data and generate anomaly recognition results based on the results of dynamic feature selection and adaptive threshold setting.

[0062] In another aspect, the present invention provides an integrated and comprehensive monitoring method for unmanned aerial vehicles, comprising the following steps:

[0063] Step 1: Control the base station perception module to synchronously broadcast the standard 5G synchronization signal block (SSB), receive the target echo signal in the airspace, and generate a local detection spectrum;

[0064] Step 2: Receive the local detection spectra output by multiple base station sensing modules, perform symbol-level fusion processing on the local detection spectra, and use the improved spatial smoothing multiple signal classification (MUSIC) algorithm to minimize the joint spectrum deviation to extract the position and velocity parameters of the target UAV;

[0065] Step 3: Based on the position and speed parameters, combined with the flight permission information and no-fly zone information stored in the regulatory database, the target drone is identified for legitimacy and its flight category is classified;

[0066] Step 4: Based on the local detection spectrum, dynamically estimate the base station perception capacity and generate traffic scheduling instructions when the number of detected drones approaches or exceeds the capacity threshold;

[0067] Step 5: Generate flight instructions or security linkage measures based on the target drone classification information and traffic scheduling instructions, and send the instructions to the ground control equipment or security system;

[0068] Step 6: Receive the mission payload data sent back by the target UAV, perform data structure analysis and abnormal feature extraction, and generate anomaly detection results for subsequent regulatory processing.

[0069] This system enables wide-coverage, high-precision, real-time perception of multiple targets within low-altitude airspace, enhancing airspace monitoring continuity and perception integrity. Local detection spectra generated by multiple base stations are fused at the symbol level and jointly minimized using a modified spatially smoothed multiple signal classification (MUSIC) algorithm. This significantly improves the consistency and detection accuracy of multi-base station perception data fusion, reduces target parameter estimation bias caused by observation errors, and accurately extracts the position and velocity of target drones. Combining the fused detection results with regulatory database data, it dynamically identifies the legitimacy and flight category of target drones, enhancing illegal flight identification and airspace management capabilities. Regarding capacity management, the system dynamically estimates base station sensing capacity based on real-time sensing data and promptly schedules drone traffic when overload trends are detected, effectively avoiding the risk of monitoring failures caused by resource constraints and improving system stability in high-density flight scenarios. The scheduling control module generates flight instructions or triggers security measures based on classification results, enabling rapid intervention and response to illegal or anomalous flight targets, improving the timeliness and safety of airspace anomaly handling. At the same time, the intelligent analysis module performs structured analysis on the mission payload data transmitted back by the drone and uses an anomaly detection method that combines dynamic feature selection and adaptive threshold adjustment to achieve real-time identification and early warning of abnormal activity in mission data, further improving the system's sensitivity and response speed to potential risky behaviors. The various modules of this invention form a complete closed loop of perception, fusion, identification, management, control, and intelligent analysis, effectively improving the overall efficiency, intelligence level, and airspace safety control capabilities of the integrated unmanned aerial vehicle monitoring system. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0071] in:

[0072] Figure 1 It is a system modular diagram of the present invention;

[0073] Figure 2 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0074] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the described embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field are within the scope of protection of the present invention.

[0075] like Figure 1 FIG. 1 is an embodiment of the present invention, which provides an integrated unmanned aerial vehicle (UAV) monitoring system, including:

[0076] (1) Base station perception module

[0077] Used to synchronize broadcast standard 5G synchronization signal block SSB and receive spatial target echoes to generate local detection spectrum;

[0078] A multi-beam transmitting unit, configured to synchronously broadcast a standard 5G synchronization signal block (SSB) in a predetermined beam sequence;

[0079] An echo receiving unit is used to switch to a receiving mode after each transmission cycle to receive an echo signal of an airspace target;

[0080] The time synchronization unit is used to perform timing synchronization control based on the 5G NR system synchronization signal or the global navigation satellite system signal, so that the time synchronization error between base stations is controlled within the set threshold;

[0081] A signal processing unit, used for pre-processing and frequency domain transformation of the received wavelet;

[0082] The spatial spectrum estimation unit is used to perform spatial spectrum estimation on the processed signal based on the improved spatial smoothing multiple signal classification MUSIC algorithm to generate local detection spectrum data.

[0083] In one embodiment, the improved spatial smoothing multiple signal classification algorithm includes:

[0084] Two-dimensional parameter spectrum estimation is introduced, and a dual-parameter detection spectrum is constructed based on the joint search of distance parameters and speed parameters;

[0085] Applying composite spatial smoothing, the array space is divided into multiple overlapping sub-arrays, and the frequency space is divided into sub-bands, and the sub-array covariance matrices are generated and averaged respectively;

[0086] The covariance matrix is reconstructed based on multi-order statistics, and the covariance matrix is corrected by fusing the statistical information of each sub-array and sub-band under finite sample conditions.

[0087] Adaptive optimization of noise subspace, dynamically setting the division threshold between noise feature subspace and signal feature subspace according to the local detection spectrum energy;

[0088] Adaptive scanning with local energy sensing limits the search area based on the initial detection of spectral energy distribution, and only performs a refined search in areas where the spectral energy exceeds the set threshold;

[0089] Target relocalization mechanism, when there are multiple detection peaks, re-estimates the initially detected target parameters based on the subspace matched filter and outputs the revised target parameters;

[0090] Parameter estimation error compensation, based on the joint spectral gradient calculation of the initial estimated point correction, update the final estimated values of the distance parameter and velocity parameter.

[0091] Through the above configuration, the base station sensing module in this embodiment uses multi-beam transmission and rapid mode switching during standard 5G synchronization signal block (SSB) broadcast and echo reception to achieve continuous coverage and real-time detection of airspace targets. The introduced improved spatial smoothing multiple signal classification (MUSIC) algorithm performs a dual-parameter search using both distance and velocity parameters to achieve simultaneous estimation of the target's spatial position and dynamic characteristics. Composite spatial smoothing is employed to divide and average substructures in both array and frequency space, improving the robustness and resolution of covariance matrix estimation. Furthermore, by fusing multi-order statistical information from subarrays and subbands, estimation bias under finite sample conditions is effectively mitigated, enhancing the stability of spectral estimation. The noise subspace adaptive optimization mechanism dynamically sets the feature subspace boundaries based on the local detection spectrum energy, enhancing the detectability of weak signal targets. The local energy sensing scanning strategy reduces interference and improves detection efficiency by limiting the search area. The target relocation and parameter estimation error compensation mechanism further optimizes the target parameter estimation accuracy through subspace matched filtering and joint spectral gradient correction. Compared with traditional base station perception and single spatial spectrum estimation algorithms, this embodiment significantly improves the accuracy, stability and reliability of low-altitude drone target detection through multi-dimensional processing optimization, and has excellent application value and expansion potential.

[0092] (2) Data fusion module

[0093] It is used to perform symbol-level fusion processing on the local detection spectra of multiple base stations and extract the position and speed of the UAV target based on the fusion results; wherein, the data fusion module adopts the symbol-level fusion method and performs joint spectrum deviation minimization processing based on the improved spatial smoothing multiple signal classification (MUSIC) algorithm;

[0094] An input interface unit, configured to receive local detection spectrum data from a plurality of base stations;

[0095] The joint spectrum calculation unit is used to calculate the joint detection spectrum according to the following formula:

[0096]

[0097] Where: P m (r, v) represents the local detection spectrum calculated by the m-th base station based on the radial distance parameter r and the radial velocity parameter v; P sum (r, v) represents the joint detection spectrum obtained by fusing the local detection spectra of multiple base stations; M represents the number of base stations involved in the fusion process;

[0098] The deviation suppression unit is used to introduce spectrum deviation minimization processing based on the joint detection spectrum, and fusion estimate the target parameters. The fusion estimation formula is:

[0099]

[0100] Where: λ represents the spectrum deviation suppression weight factor, which is used to suppress the impact of single base station measurement errors on the fusion results; represents the radial distance parameter estimation value and radial velocity parameter estimation value of the target UAV obtained by fusion estimation;

[0101] The output interface unit is used to output the radial distance parameters and radial velocity parameters of the target UAV obtained by fusion estimation.

[0102] Through the above configuration, the data fusion module in this embodiment receives local detection spectrum data generated by multiple base station sensing modules and performs joint processing based on a symbol-level fusion method. Compared to traditional solutions that focus on decision-level fusion, this method can achieve information fusion at the underlying signal level, fully preserving the phase and amplitude characteristics of each base station's detection signal. During the fusion process, an improved spatially smoothed multiple signal classification (MUSIC) algorithm is used to perform joint spectral deviation minimization estimation on the local detection spectrum data. By introducing spectral deviation suppression weights during the fusion process, the impact of single-base station measurement errors on the joint spectral estimation results is effectively reduced, improving the consistency and accuracy of the fused detection. Specifically, during the fusion estimation process, the joint spectrum calculation unit first calculates the total joint detection spectrum based on each local detection spectrum. Then, the deviation suppression unit adjusts the contribution of each base station signal based on a set spectral deviation weight factor, thereby achieving optimal extraction of target parameters. Through the above processing, this embodiment can achieve high-precision target drone position and velocity parameter estimation even when multi-base station observation information contains random errors, synchronization deviations, or partial interference. Compared with traditional methods, it achieves significant improvements in fusion accuracy, stability, and weak signal detection sensitivity. It is particularly suitable for joint detection and tracking of multi-source drone targets in complex airspace environments.

[0103] (3) Capacity Management Module

[0104] Used to dynamically estimate base station sensing capacity and perform drone traffic scheduling when overloaded;

[0105] The drone number calculation unit is used to calculate the maximum number of drones N that can be supported based on the dynamic estimation of the base station's carrying capacity. max , where N max Calculated by the following formula:

[0106]

[0107] Where: γ avail To effectively perceive the signal-to-noise ratio, is the ambient noise power, γ min is the minimum detection threshold;

[0108] Control unit, used to calculate the maximum number of drones N max , perform real-time traffic control, when the number of drones approaches N max When the detection is complete, delay or guidance control is performed to avoid system overload and ensure detection feasibility.

[0109] The capacity management module in this embodiment dynamically estimates the maximum number of drones each base station can currently support based on real-time sensor data. Specifically, the drone number calculation unit uses a capacity estimation formula to calculate the maximum drone capacity in real time based on three metrics: the effective sensor signal-to-noise ratio, the ambient noise power, and the minimum detection threshold. When the system detects that the number of connected drones approaches or exceeds the capacity threshold, the control unit triggers a flow control mechanism based on the capacity estimation results. This mechanism implements measures such as delayed mission start, flight path guidance adjustments, or mandatory flow reduction, effectively mitigating the risk of monitoring failure caused by system overload. Compared to existing technologies that set capacity limits based solely on static parameters, this embodiment introduces a real-time dynamic estimation mechanism that adaptively adjusts the sensor capacity limit based on airspace environmental changes, significantly improving system stability and reliability in complex flight scenarios. Furthermore, by linking capacity control with flow scheduling, intelligent flow distribution is achieved in high-density drone flight environments, enhancing system resource utilization efficiency and flight safety assurance.

[0110] (4) Flight identification module

[0111] Used to combine the fusion detection results and the regulatory database to identify the identity legitimacy and flight category of the target drone;

[0112] A track feature extraction unit is used to extract the speed change rate, track curvature and flight altitude change rate as track feature parameters based on the historical track data of the UAV target;

[0113] a classifier unit, configured to perform category determination on the track characteristic parameters based on a trained machine learning model, wherein the machine learning model comprises a support vector machine (SVM) model or a convolutional neural network (CNN) model;

[0114] A data comparison unit is used to compare the identification result of the target with the flight permission information and no-fly zone information in the supervision database to verify the target's ID number, flight permission status and the legitimacy of the flight zone;

[0115] The output unit is used to output the legality classification results and flight category labels of the target drone.

[0116] The flight identification module in this embodiment combines fused detection results with data from a regulatory database to comprehensively determine the target drone's identity and flight category. Specifically, the track feature extraction unit extracts dynamic characteristic parameters such as velocity change rate, track curvature, and altitude change rate from the target drone's historical track data to form a track feature vector for identification. The classifier unit classifies and determines the extracted track feature vector using a trained machine learning model. Supported models include support vector machines (SVMs) or convolutional neural networks (CNNs) to accurately classify different flight categories. The data comparison unit automatically matches the identification results with flight permit information and no-fly zone demarcations stored in the regulatory database to verify the target drone's ID code and flight permit status, and determine its legality in the current airspace. Based on the combined comparison and classification results, the output unit outputs the target drone's legality classification result and flight category label. Compared with traditional flight identification methods that rely on single track features or static rule comparison, this embodiment significantly improves the accuracy of flight identification and the sensitivity of illegal target detection by introducing multi-dimensional dynamic feature extraction and machine learning classification, combined with real-time database comparison, and can adapt to the needs of drone supervision in highly dynamic and complex airspace environments.

[0117] (5) Scheduling control module

[0118] Used to issue flight instructions or trigger security linkage measures based on target classification results;

[0119] The illegal flight emergency response unit is used to automatically generate flight trajectory optimization instructions when illegal or abnormal drones approach, and control the drone to adjust its speed and heading to avoid entering restricted areas;

[0120] Illegal flight warning unit, used to compare flight targets with the geographical restriction database in real time, trigger illegal flight warnings and initiate relevant handling mechanisms;

[0121] The path adjustment unit activates alternative flight path planning or adjusts the scheduling order of flight missions to avoid congested areas when the flight density is high.

[0122] The dispatch control module in this embodiment issues corresponding flight instructions or triggers security linkage measures based on the classification results of the target drone, implementing graded responses for different abnormal situations. Specifically, the illegal flight emergency response unit automatically generates flight trajectory optimization instructions when detecting an illegal or abnormal drone approaching a restricted area, and controls the target drone to adjust its speed and course, guiding it away from the controlled area and reducing the risk of intrusion. The illegal flight warning unit compares flight targets with information in a geographic restriction database in real time to promptly identify violations of flight regulations. When illegal flights are identified, a warning mechanism is activated, and the ground control system or security equipment is coordinated to take further action. When the flight density exceeds a preset threshold, the path adjustment unit dynamically plans alternative flight paths or adjusts the flight mission scheduling sequence to guide drones away from congested areas and reduce the risk of flight conflicts. Compared to traditional dispatch methods that rely primarily on static no-fly zone warnings, this embodiment combines real-time target classification, anomaly detection, and dynamic flight adjustments to form an integrated flight dispatch control system that integrates early warning, avoidance, and linkage measures. This significantly improves the system's response speed and safety control capabilities in complex and dynamic airspace environments.

[0123] (6) Intelligent Analysis Module

[0124] Used to process the mission payload data sent back by the drone and perform anomaly detection.

[0125] The data processing unit is used to receive multimodal sensor data and fuse the different sensor data using a weighted decision algorithm. The calculation result formula is as follows:

[0126]

[0127] Where: w i is the feature weight of each sensing modality, f i (I) is the modal characteristic function, and the anomaly score D is calculated by weighted fusion; N is the total number of sensing modalities involved in weighted fusion;

[0128] The dynamic evaluation unit is used to set a dynamic threshold according to the fusion evaluation value and perform anomaly detection on the sensor data to determine whether the system status is abnormal.

[0129] In a specific embodiment, the mission data parsing unit is used to perform structured parsing on the mission payload data transmitted back by the target UAV and extract a multi-dimensional feature set;

[0130] A dynamic feature selection unit, configured to dynamically determine a feature subset for anomaly detection based on real-time statistical characteristics of the extracted multidimensional feature set;

[0131] An adaptive threshold setting unit, used to dynamically adjust the decision threshold of anomaly detection based on the distribution characteristics of feature subsets;

[0132] The anomaly recognition unit is used to perform anomaly detection on the mission payload data and generate anomaly recognition results based on the results of dynamic feature selection and adaptive threshold setting.

[0133] The intelligent analysis module in this embodiment performs anomaly detection on the mission payload data transmitted back by the drone, achieving early detection and identification of abnormal conditions through multi-module collaboration. The data processing unit receives status data from multimodal sensors and fuses the different sensory data using a weighted decision algorithm to calculate a comprehensive anomaly score. The dynamic assessment unit sets dynamic thresholds based on the comprehensive score, monitors changes in the characteristics of the fused data, and determines in real time whether the system status is abnormal. In specific implementation, the mission data analysis unit first performs structured analysis on the transmitted mission payload data, extracting multidimensional feature information such as speed, position, and mission progress. The dynamic feature selection unit then selects the feature subset most relevant to anomaly detection from this multidimensional feature set, improving detection efficiency and accuracy. Subsequently, the adaptive threshold setting unit dynamically adjusts the anomaly detection threshold based on the statistical characteristics of the feature subset, achieving adaptive sensitivity adjustment for different mission conditions. Finally, the anomaly identification unit, based on the results of the dynamic feature selection and adaptive threshold setting, comprehensively analyzes the payload data behavior and generates an anomaly detection conclusion. Compared with traditional anomaly detection methods based on fixed rules or single feature monitoring, this embodiment significantly improves the accuracy and response speed of anomaly detection by introducing dynamic feature optimization, adaptive threshold adjustment and multimodal fusion analysis. It is particularly suitable for drone risk warning and abnormal behavior identification in complex dynamic flight mission environments.

[0134] like Figure 2 FIG. 1 is another embodiment of the present invention, which provides an integrated comprehensive monitoring method for unmanned aerial vehicles. The method uses the above-mentioned integrated comprehensive monitoring system for unmanned aerial vehicles, including the following steps:

[0135] Step 1: Control the base station perception module to synchronously broadcast the standard 5G synchronization signal block (SSB), receive the target echo signal in the airspace, and generate a local detection spectrum;

[0136] Step 2: Receive the local detection spectra output by multiple base station sensing modules, perform symbol-level fusion processing on the local detection spectra, and use the improved spatial smoothing multiple signal classification (MUSIC) algorithm to minimize the joint spectrum deviation to extract the position and velocity parameters of the target UAV;

[0137] Step 3: Based on the position and speed parameters, combined with the flight permission information and no-fly zone information stored in the regulatory database, the target drone is identified for legitimacy and its flight category is classified;

[0138] Step 4: Based on the local detection spectrum, dynamically estimate the base station perception capacity and generate traffic scheduling instructions when the number of detected drones approaches or exceeds the capacity threshold;

[0139] Step 5: Generate flight instructions or security linkage measures based on the target drone classification information and traffic scheduling instructions, and send the instructions to the ground control equipment or security system;

[0140] Step 6: Receive the mission payload data sent back by the target UAV, perform data structure analysis and abnormal feature extraction, and generate anomaly detection results for subsequent regulatory processing.

[0141] In summary, the present invention addresses the deficiencies of existing technologies in terms of airspace synchronous perception, data fusion accuracy, flight identification intelligence, capacity dynamic management, and mission payload data anomaly detection. By constructing a comprehensive supervision architecture with multi-module collaboration, it achieves wide coverage, high precision, and intelligent supervision of low-altitude airspace. The present invention can effectively improve the consistency and accuracy of drone target detection, enhance the ability to identify illegal flight behaviors, ensure the stability of system operation in high-density flight environments, and improve the level of real-time identification of abnormal activities in mission data. With its complete closed-loop chain of perception, fusion, identification, management, scheduling, and intelligent analysis, the present invention has good scalability and adaptability, and can be widely used in urban airspace traffic management, low-altitude protection of important areas, emergency response command, and intelligent supervision of various drone application scenarios. It has high practical application value and promotion prospects.

[0142] In the description of this specification, the reference terms "one embodiment," "some embodiments," "example," "specific example," or "some examples" mean that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials, or characteristics described may be combined in any appropriate manner in any one or more embodiments or examples. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification, as well as features of different embodiments or examples, unless they are mutually inconsistent.

[0143] Any process or method description in the flowchart or otherwise described herein can be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process. The scope of the preferred embodiments of the present application includes additional implementations in which the functions may be performed in a different order than shown or discussed, including in a substantially simultaneous manner or in a reverse order depending on the functions involved.

[0144] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various modifications or substitutions within the technical scope disclosed in this application, and such modifications or substitutions should be included within the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. An integrated unmanned aerial vehicle monitoring system, characterized in that: include: The base station sensing module is used to synchronously broadcast the standard 5G synchronization signal block (SSB) and receive the target echo in the airspace to generate a local detection spectrum. A data fusion module is used to perform symbol-level fusion processing on the local detection spectra of multiple base stations and extract the position and velocity of the UAV target based on the fusion results. The data fusion module adopts a symbol-level fusion method and performs joint spectrum deviation minimization processing based on an improved spatial smoothing multiple signal classification (MUSIC) algorithm. Capacity management module, which is used to dynamically estimate the base station sensing capacity and perform drone traffic scheduling when overloaded; The flight identification module is used to combine the fusion detection results and the regulatory database to identify the identity legitimacy and flight category of the target drone; The dispatch control module is used to issue flight instructions or trigger security linkage measures based on target classification results; The intelligent analysis module is used to process the mission payload data sent back by the drone and perform anomaly detection.

2. The integrated unmanned aerial vehicle monitoring system according to claim 1, characterized in that: The base station sensing module includes: A multi-beam transmitting unit, configured to synchronously broadcast a standard 5G synchronization signal block (SSB) in a predetermined beam sequence; An echo receiving unit is used to switch to a receiving mode after each transmission cycle to receive an echo signal of an airspace target; The time synchronization unit is used to perform timing synchronization control based on the 5G NR system synchronization signal or the global navigation satellite system signal, so that the time synchronization error between base stations is controlled within the set threshold; A signal processing unit, used for pre-processing and frequency domain transformation of the received wavelet; The spatial spectrum estimation unit is used to perform spatial spectrum estimation on the processed signal based on the improved spatial smoothing multiple signal classification MUSIC algorithm to generate local detection spectrum data.

3. The integrated unmanned aerial vehicle monitoring system according to claim 2, characterized in that: The improved spatial smoothing multiple signal classification algorithm includes: Two-dimensional parameter spectrum estimation is introduced, and a dual-parameter detection spectrum is constructed based on the joint search of distance parameters and speed parameters; Applying composite spatial smoothing, the array space is divided into multiple overlapping sub-arrays, and the frequency space is divided into sub-bands, and the sub-array covariance matrices are generated and averaged respectively; The covariance matrix is reconstructed based on multi-order statistics, and the covariance matrix is corrected by fusing the statistical information of each sub-array and sub-band under finite sample conditions. Adaptive optimization of noise subspace, dynamically setting the division threshold between noise feature subspace and signal feature subspace according to the local detection spectrum energy; Adaptive scanning with local energy sensing limits the search area based on the initial detection of spectral energy distribution, and only performs a refined search in areas where the spectral energy exceeds the set threshold; Target relocalization mechanism, when there are multiple detection peaks, re-estimates the initially detected target parameters based on the subspace matched filter and outputs the revised target parameters; Parameter estimation error compensation, based on the joint spectral gradient calculation of the initial estimated point correction, update the final estimated values of the distance parameter and velocity parameter.

4. The integrated unmanned aerial vehicle monitoring system according to claim 1, characterized in that: The data fusion module includes: An input interface unit, configured to receive local detection spectrum data from a plurality of base stations; The joint spectrum calculation unit is used to calculate the joint detection spectrum according to the following formula: Where: P m (r, v) represents the local detection spectrum calculated by the m-th base station based on the radial distance parameter r and the radial velocity parameter v; P sum (r, v) represents the joint detection spectrum obtained by fusing the local detection spectra of multiple base stations; M represents the number of base stations involved in the fusion process; The deviation suppression unit is used to introduce spectrum deviation minimization processing based on the joint detection spectrum, and fusion estimate the target parameters. The fusion estimation formula is: Where: λ represents the spectrum deviation suppression weight factor, which is used to suppress the impact of single base station measurement errors on the fusion results; represents the radial distance parameter estimation value and radial velocity parameter estimation value of the target UAV obtained by fusion estimation; The output interface unit is used to output the radial distance parameters and radial velocity parameters of the target UAV obtained by fusion estimation.

5. The integrated unmanned aerial vehicle monitoring system according to claim 1, characterized in that: The capacity management module includes: The drone number calculation unit is used to calculate the maximum number of drones N that can be supported based on the dynamic estimation of the base station's carrying capacity. max , the calculation formula is: Where: γ avail To effectively perceive the signal-to-noise ratio, is the ambient noise power, γ min is the minimum detection threshold; Control unit, used to calculate the maximum number of drones N max , perform real-time traffic control, when the number of drones approaches N max When the delay or guidance control is performed.

6. The integrated unmanned aerial vehicle monitoring system according to claim 1, characterized in that: The flight identification module includes: A track feature extraction unit is used to extract the speed change rate, track curvature and flight altitude change rate as track feature parameters based on the historical track data of the UAV target; a classifier unit, configured to perform category determination on the track characteristic parameters based on a trained machine learning model, wherein the machine learning model includes a support vector machine (SVM) model or a convolutional neural network (CNN) model; A data comparison unit is used to compare the identification result of the target with the flight permission information and no-fly zone information in the supervision database to verify the target's ID number, flight permission status and the legitimacy of the flight zone; The output unit is used to output the legality classification results and flight category labels of the target drone.

7. The integrated unmanned aerial vehicle monitoring system according to claim 1, characterized in that: The scheduling control module includes: The illegal flight emergency response unit is used to automatically generate flight trajectory optimization instructions when illegal or abnormal drones approach, and control the drone to adjust its speed and heading to avoid entering restricted areas; Illegal flight warning unit, used to compare flight targets with the geographical restriction database in real time, trigger illegal flight warnings and initiate relevant handling mechanisms; The path adjustment unit activates alternative flight path planning or adjusts the scheduling order of flight missions to avoid congested areas when the flight density is high.

8. The integrated unmanned aerial vehicle monitoring system according to claim 1, characterized in that: The intelligent analysis module includes: The data processing unit is used to receive multimodal sensor data and fuse the different sensor data using a weighted decision algorithm. The calculation result formula is as follows: Where: w i is the feature weight of each sensing modality, f i (I) is the modal characteristic function, and the anomaly score D is calculated by weighted fusion; N is the total number of sensing modalities involved in weighted fusion; The dynamic evaluation unit is used to set a dynamic threshold according to the fusion evaluation value and perform anomaly detection on the sensor data. The exception handling unit is used to initiate predetermined emergency measures and feedback alarm information when a system abnormality is detected.

9. The integrated unmanned aerial vehicle monitoring system according to claim 8, characterized in that: The intelligent analysis module also includes: The data parsing unit is used to perform structured analysis on the mission payload data transmitted back by the target UAV and extract a multi-dimensional feature set; the dynamic feature selection unit is used to dynamically determine the feature subset for anomaly detection based on the real-time statistical characteristics of the extracted multi-dimensional feature set; An adaptive threshold setting unit, used to dynamically adjust the decision threshold of anomaly detection based on the distribution characteristics of feature subsets; The anomaly recognition unit is used to perform anomaly detection on the mission payload data and generate anomaly recognition results based on the results of dynamic feature selection and adaptive threshold setting.

10. An integrated comprehensive monitoring method for unmanned aerial vehicles, using an integrated comprehensive monitoring system for unmanned aerial vehicles according to any one of claims 1 to 9, characterized in that: The following steps are involved: Step 1: Control the base station perception module to synchronously broadcast the standard 5G synchronization signal block (SSB), receive the target echo signal in the airspace, and generate a local detection spectrum; Step 2: Receive the local detection spectra output by multiple base station sensing modules, perform symbol-level fusion processing on the local detection spectra, and use the improved spatial smoothing multiple signal classification (MUSIC) algorithm to minimize the joint spectrum deviation to extract the position and velocity parameters of the target UAV; Step 3: Based on the position and speed parameters, combined with the flight permission information and no-fly zone information stored in the regulatory database, the target drone is identified for legitimacy and its flight category is classified; Step 4: Based on the local detection spectrum, dynamically estimate the base station perception capacity and generate traffic scheduling instructions when the number of detected drones approaches or exceeds the capacity threshold; Step 5: Generate flight instructions or security linkage measures based on the target drone classification information and traffic scheduling instructions, and send the instructions to the ground control equipment or security system; Step 6: Receive the mission payload data sent back by the target UAV, perform data structure analysis and abnormal feature extraction, and generate anomaly detection results for subsequent regulatory processing.

Images