A radar and audio-based anti-UAV defense method and system

Abstract

The present invention provides an anti-UAV defense method and system based on radar and audio, wherein the present invention constructs a three-dimensional airspace monitoring network by deploying radar sensors and audio capture devices at different heights and positions, implements time-space synchronized multimodal information fusion, obtains signal audio alignment data, generates a real-time behavior trajectory of a target in three-dimensional space by a spatial positioning algorithm, obtains a behavior profile with unique identification by combining a behavior profile modeling technology with a machine learning algorithm, and analyzes the behavior profile to obtain an analysis result, constructs a threat level determination model according to the analysis result, and uses the threat level determination model to evaluate the intention and risk level of the abnormal behavior of the UAV to generate a threat assessment result, and uses an adaptive threat assessment engine to generate a target defense plan; the technical solution provided by the present invention improves the accuracy and reliability of target positioning and the level of intelligence, and enhances the ability to cope with complex threats.

Description

Technical Field

[0001] The embodiments of the present invention relate to the technical field of unmanned aerial vehicles (UAVs), and in particular to an anti-UAV defense method based on radar and audio. Background Art

[0002] With the rapid development of drone technology, the widespread use of drones in civil and military fields has brought new security challenges, especially in sensitive areas such as critical infrastructure, airports, border areas, and large-scale event venues. Illegal or unauthorized drone activities may pose a serious security risk. In order to effectively deal with this threat, an anti-drone defense system is needed that can monitor, identify and evaluate drone behavior in real time and respond quickly;

[0003] Existing solutions mainly include single sensor (such as radar or camera) monitoring and rule-based threat assessment. These methods usually use radar sensors to transmit and receive reflected signal data to detect objects in the airspace, provide target location information, and record surrounding environmental sounds through audio capture devices to assist in determining the presence of drones and their approximate location.

[0004] However, these solutions have significant limitations: relying on a single type of sensor cannot provide enough information to accurately distinguish between legal and illegal drones, and is prone to false alarms or missed alarms; existing rule-based threat assessment models lack flexibility, have difficulty adapting to the ever-changing threat environment, and are slow to respond to new or unforeseen behavior patterns; they fail to fully integrate multimodal information (such as radar reflection signals and audio data), resulting in an incomplete understanding of target behavior and affecting the quality of the final decision; the lack of intelligent predictive models and adaptive engines that support rapid adjustments leads to delayed responses when facing complex and rapidly changing threats. Summary of the invention

[0005] The embodiments of the present invention provide a radar and audio-based anti-UAV defense method and system to solve the problems in the prior art that false alarms or missed alarms are prone to occur, the system is slow to respond to new or unforeseen behavior patterns, and the system fails to fully integrate multimodal information (such as radar reflection signals and audio data), resulting in an incomplete understanding of target behavior. The system lacks an intelligent prediction model and an adaptive engine that supports rapid adjustment, resulting in a delayed response when facing complex and rapidly changing threats.

[0006] In a first aspect, an embodiment of the present invention provides an anti-UAV defense method based on radar and audio, comprising:

[0007] A three-dimensional airspace monitoring network is constructed by deploying radar sensors and audio capture devices at different heights and locations, wherein the radar sensors are used to transmit and receive reflected signal data to detect objects in the airspace, and the audio capture devices are used to capture real-time audio data to record ambient sound;

[0008] Utilizing the three-dimensional airspace monitoring network, implementing spatiotemporal synchronized multimodal information fusion, aligning the reflected signal data from the radar sensor and the real-time audio data acquired by the audio capture device on the time axis to obtain signal audio alignment data, and associating the target dynamics captured by each radar sensor through a spatial positioning algorithm to generate a real-time behavior trajectory of the target in the three-dimensional space;

[0009] Based on the real-time behavior trajectory, the behavior profile modeling technology is combined with the machine learning algorithm to extract the UAV flight mode characteristics from the signal audio alignment data, and the UAV flight mode characteristics are compared with the preset UAV behavior pattern library in real time to obtain a behavior profile with unique identification, and the behavior profile with unique identification is analyzed to obtain an analysis result;

[0010] A threat level determination model is constructed based on the analysis results, and the threat level determination model is used to evaluate the intention and risk level of the abnormal behavior of the drone to generate a threat assessment result. An adaptive threat assessment engine is used to select defense measures corresponding to the threat level based on the threat assessment result, and the action trend of the drone is predicted in combination with a preset prediction model to generate a target defense plan.

[0011] Optionally, based on the real-time behavior trajectory, the behavior profile modeling technology is combined with a machine learning algorithm to extract the UAV flight mode features from the signal audio alignment data, and the UAV flight mode features are compared with a preset UAV behavior pattern library in real time to obtain a behavior profile with a unique identifier, and the behavior profile with a unique identifier is parsed to obtain a parsing result, including:

[0012] The real-time behavior trajectory is analyzed and processed by using the behavior profile modeling technology combined with the machine learning algorithm, the UAV flight mode characteristics are extracted, and a UAV flight mode feature set is generated, wherein the UAV flight mode characteristics include: flight speed, acceleration, altitude change rate, hovering time, and steering angle and frequency;

[0013] Compare the UAV flight mode feature set with a preset UAV behavior pattern library in real time, identify the behavior pattern matching the preset value, and mark it as a preliminary matching result, and calculate the similarity score of each behavior pattern to obtain a preliminary matching list;

[0014] Based on the preliminary matching list, a deep learning algorithm is used for analysis, time series analysis is introduced to evaluate the temporal consistency of the behavior pattern, and a multi-scale feature detection method is used to capture the behavior characteristics of the UAV at different scales, optimize the matching accuracy of the UAV flight mode feature set and the preset UAV behavior pattern library, determine a uniquely identifiable behavior profile, and generate an optimized behavior profile;

[0015] Based on the optimized behavior profile, the action intention of the UAV is analyzed in combination with the historical behavior record and the prediction model to generate a behavior analysis report, wherein the reported behavior analysis includes flight path planning, hovering purpose, and intention to approach the target;

[0016] Combined with the information in the behavior analysis report, the assessment rules of the adaptive threat assessment engine are adjusted to obtain an adjusted adaptive threat assessment engine, and the behavior pattern of the drone is analyzed again using the adjusted adaptive threat assessment engine to generate a target analysis result.

[0017] Optionally, based on the preliminary matching list, a deep learning algorithm is used for analysis, time series analysis is introduced to evaluate the temporal consistency of the behavior pattern, and a multi-scale feature detection method is used to capture the behavior characteristics of the drone at different scales, optimize the matching accuracy of the drone flight mode feature set and the preset drone behavior pattern library, determine a uniquely identifiable behavior profile, and generate an optimized behavior profile, including:

[0018] Using deep learning algorithms, we analyze the behavior patterns in the preliminary matching list, introduce time series analysis to evaluate the temporal consistency of each behavior pattern, and obtain a temporal consistency evaluation report.

[0019] According to the temporal consistency assessment report, a multi-scale feature detection method is used to capture the behavioral characteristics of the drone at different scales under the same behavior mode, and the frequency components in the behavior mode are extracted by combining frequency domain analysis to generate a multi-scale behavior feature description;

[0020] Based on the multi-scale behavior feature description, a comparative optimization algorithm is used to optimize the matching process between the UAV flight mode feature set and the preset UAV behavior pattern library, and the optimal matching parameters are calculated by genetic algorithm and particle swarm optimization technology, and the similarity score is adjusted to obtain the optimized matching result;

[0021] Based on the optimized matching results, cluster analysis and anomaly detection algorithms are applied to identify and determine initial behavior profiles with unique identification, and a unique identifier is assigned to each initial behavior profile to generate an optimized behavior profile.

[0022] Optionally, based on the multi-scale behavior feature description, a comparative optimization algorithm is used to optimize the matching process between the UAV flight mode feature set and the preset UAV behavior pattern library, and the optimal matching parameters are calculated by genetic algorithm and particle swarm optimization technology, and the similarity score is adjusted to obtain an optimized matching result, including:

[0023] Using the comparative optimization algorithm, based on the multi-scale behavior feature description, the UAV flight mode feature set is preliminarily compared with the preset UAV behavior pattern library to obtain the initial similarity score matrix, and the time consistency evaluation score of each behavior pattern is recorded;

[0024] According to the initial similarity score matrix, a genetic algorithm is applied to search for the best matching parameter combination between the UAV flight mode feature set and the preset UAV behavior pattern library, and a cross-validation mechanism is introduced to generate a candidate matching parameter set;

[0025] Based on the candidate matching parameter set, the particle swarm optimization technology is used to optimize the matching parameters, the fitness of each candidate matching parameter is evaluated by iterative calculation, and the optimal matching parameters are updated in combination with the time consistency evaluation score, and a local search strategy is implemented to adjust the initial similarity scoring matrix to obtain an optimized similarity scoring matrix;

[0026] According to the optimized similarity scoring matrix, the behavior patterns that meet the preset UAV behavior pattern library are screened, and the hierarchical clustering analysis is applied to classify the behavior patterns that meet the preset UAV behavior pattern library into the same category, and the behavior patterns that meet the preset matching accuracy are determined, the optimized matching results are generated, and a unique identifier is assigned to each optimized matching result.

[0027] Optionally, a three-dimensional airspace monitoring network is constructed by deploying radar sensors and audio capture devices at different heights and positions, wherein the radar sensors are used to transmit and receive reflected signal data to detect objects in the airspace, and the audio capture devices are used to capture real-time audio data to record ambient sound, including:

[0028] Deploy radar sensors at different heights and positions in a preset airspace to generate a spatial distribution of radar sensors, wherein each radar sensor is used to transmit and receive reflection signal data to detect objects in the airspace, thereby obtaining a radar reflection signal data set;

[0029] According to the spatial distribution of the radar sensors, audio capture devices are deployed at locations coordinated with the radar sensors, and the sounds of the surrounding environment are recorded in real time to obtain a real-time audio data set;

[0030] Based on the radar reflection signal dataset and the real-time audio dataset, high-precision clock synchronization technology and distributed computing architecture are used to perform spatiotemporal synchronization processing on the radar reflection signal dataset and the real-time audio dataset to obtain signal audio alignment data;

[0031] By applying spatial positioning algorithms, combining geographic information system data and three-dimensional terrain models, the reflected signal data of each radar sensor and the real-time audio data of each audio capture device are correlated and integrated to build a three-dimensional airspace monitoring network.

[0032] Optionally, the three-dimensional airspace monitoring network is used to implement time-space synchronized multimodal information fusion, align the reflected signal data from the radar sensor and the real-time audio data acquired by the audio capture device on the time axis to obtain signal audio alignment data, and associate the target dynamics captured by each radar sensor through a spatial positioning algorithm to generate a real-time behavior trajectory of the target in the three-dimensional space, including:

[0033] Using high-precision clock synchronization technology and distributed computing architecture, the reflected signal data from the radar sensor and the real-time audio data obtained by the audio capture device are aligned on the time axis to obtain initial signal audio alignment data;

[0034] According to the signal audio alignment data, multi-level preprocessing is performed, and the reflected signal data of each radar sensor and the real-time audio data acquired by each audio capture device are calibrated through a calibration program to generate optimized signal audio alignment data, wherein the multi-level preprocessing includes: noise filtering, data smoothing, outlier detection and frequency domain analysis;

[0035] Based on the optimized signal audio alignment data, a spatial positioning algorithm is applied, combined with geographic information system data, three-dimensional terrain model and environmental meteorological data, the target dynamic information captured by each radar sensor is associated to generate a preliminary positioning result of the target in three-dimensional space. According to the preliminary positioning result, an advanced trajectory tracking algorithm is applied to process and generate a real-time behavior trajectory of the target in three-dimensional space.

[0036] Optionally, a threat level determination model is constructed according to the analysis result, and the intention and risk level of the abnormal behavior of the drone is evaluated using the threat level determination model to generate a threat assessment result. An adaptive threat assessment engine is used to select defense measures corresponding to the threat level according to the threat assessment result, and the action trend of the drone is predicted in combination with a preset prediction model to generate a target defense plan, including:

[0037] Using the analysis results, the flight mode characteristics and historical behavior data of drones are integrated, real-time environmental parameters and context information are introduced, multiple threat levels and corresponding evaluation rules are defined, and a dynamic threat level determination model is obtained;

[0038] According to the dynamic threat level determination model, a multi-dimensional comprehensive evaluation is performed on the abnormal behavior of the drone identified in the analysis results. By comparing the behavior pattern of the drone with the preset drone behavior pattern library and combining the situational awareness algorithm, the intention and risk level of the abnormal behavior of the drone are determined from multiple angles to generate a preliminary threat assessment result.

[0039] Based on the preliminary threat assessment results, an intelligent adaptive threat assessment engine is activated to provide threat assessment suggestions in combination with current assessment results, historical assessment records and preset prediction models to generate target threat assessment results;

[0040] Using the target threat assessment results, select multi-level defense measures that match the threat level and generate a priority ranking of defense measures;

[0041] Using the preset prediction model, combined with the drone's historical behavior, real-time behavior trajectory and external intelligence information, the drone's action trend is predicted to obtain the prediction results. Based on the prediction results and the priority of defense measures, a target defense plan is formulated.

[0042] In a second aspect, an embodiment of the present invention provides an anti-UAV defense system based on radar and audio, comprising:

[0043] A construction module is used to construct a three-dimensional airspace monitoring network by deploying radar sensors and audio capture devices at different heights and positions, wherein the radar sensors are used to transmit and receive reflected signal data to detect objects in the airspace, and the audio capture devices are used to capture real-time audio data to record ambient sound;

[0044] An alignment module is used to utilize the three-dimensional airspace monitoring network to implement time-space synchronized multimodal information fusion, align the reflected signal data from the radar sensor and the real-time audio data acquired by the audio capture device on the time axis to obtain signal audio alignment data, and associate the target dynamics captured by each radar sensor through a spatial positioning algorithm to generate a real-time behavior trajectory of the target in the three-dimensional space;

[0045] A comparison module is used to extract the UAV flight mode features from the signal audio alignment data based on the real-time behavior trajectory by using the behavior profile modeling technology combined with the machine learning algorithm, and to compare the UAV flight mode features with a preset UAV behavior pattern library in real time to obtain a behavior profile with a unique identifier, and to parse the behavior profile with a unique identifier to obtain a parsing result;

[0046] An assessment module is used to construct a threat level determination model based on the analysis results, and use the threat level determination model to evaluate the intention and risk level of the abnormal behavior of the drone to generate a threat assessment result. An adaptive threat assessment engine is used to select defense measures corresponding to the threat level based on the threat assessment result, and a preset prediction model is combined to predict the action trend of the drone to generate a target defense plan.

[0047] In a third aspect, an embodiment of the present invention provides a computing device, comprising a processor and a memory, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute a radar and audio-based anti-UAV defense method as described in any one of the first aspects.

[0048] In a fourth aspect, an embodiment of the present invention provides a computer storage medium having computer program instructions stored thereon, wherein the computer program instructions, when executed by a processor, implement a radar and audio-based anti-UAV defense method as described in any one of the first aspects.

[0049] In the embodiment of the present invention, a three-dimensional airspace monitoring network is constructed by deploying radar sensors and audio capture devices at different heights and positions. The radar sensor transmits and receives reflected signal data to detect objects in the airspace, while the audio capture device records real-time audio data to reflect the surrounding environment sound. Using this three-dimensional airspace monitoring network, the system implements time-space synchronous multimodal information fusion, aligns the radar reflected signal data and audio data on the time axis, generates signal audio alignment data, and associates the target dynamics captured by each radar sensor through a spatial positioning algorithm to generate the target's real-time behavior trajectory in three-dimensional space; based on these real-time behavior trajectories, the system uses behavior profile modeling technology and machine learning algorithms to extract drone flight mode features from the signal audio alignment data, and compares it with the preset drone behavior pattern library in real time to obtain a uniquely identifiable behavior profile. By analyzing these behavior profiles, the system generates detailed analysis results; finally, a threat level determination model is constructed based on the analysis results to evaluate the intention and risk level of the drone's abnormal behavior and generate a threat assessment result. The adaptive threat assessment engine selects corresponding defense measures based on these assessment results, and combines the prediction model to predict the action trend of the drone, and finally generates a comprehensive target defense plan; the technical solution provided by the present invention realizes comprehensive monitoring and precise positioning, ensures all-round and multi-dimensional monitoring of drone activities, improves the accuracy and reliability of target positioning, improves the intelligence level of the system, and enhances its ability to deal with complex threats;

[0050] The behavior profile modeling technology is further combined with the machine learning algorithm to analyze and process the real-time behavior trajectory, extract the UAV flight mode features including flight speed, acceleration, altitude change rate, hovering time, steering angle and frequency, and generate the UAV flight mode feature set. This detailed feature extraction provides a solid foundation for subsequent behavior pattern recognition; the UAV flight mode feature set is compared with the preset behavior pattern library in real time, the behavior pattern matching the preset value is identified and marked as the preliminary matching result. At the same time, the similarity score of each behavior pattern is calculated to obtain a preliminary matching list. This process ensures the rapid identification and classification of UAV behavior patterns and improves the response speed of the system; based on the preliminary matching list, the deep learning algorithm is used for analysis, time series analysis is introduced to evaluate the temporal consistency of the behavior pattern, and a multi-scale feature detection method is used to capture the behavior characteristics of the UAV at different scales, optimize the matching accuracy of the UAV flight mode feature set and the preset behavior pattern library, determine the behavior profile with unique identification, and generate the optimized behavior profile. This method significantly improves the accuracy and robustness of behavior pattern recognition; based on the optimized behavior profile, the drone's action intention is analyzed in combination with historical behavior records and prediction models, and a behavior analysis report is generated that includes flight path planning, hovering purpose, and target approach intention. In addition, the evaluation rules of the adaptive threat assessment engine are adjusted to analyze the drone's behavior pattern again and generate target analysis results. This enables the system to understand the drone's intention more accurately, thereby taking more targeted defense measures.

[0051] Furthermore, using the comparative optimization algorithm, based on the multi-scale behavior feature description, the UAV flight mode feature set is preliminarily compared with the preset behavior pattern library to obtain the initial similarity score matrix and record the time consistency evaluation score of each behavior pattern. This step ensures the initial accuracy of the matching process and provides basic data for subsequent optimization.

[0052] Efficient parameter search mechanism: According to the initial similarity score matrix, the genetic algorithm is applied to search for the best matching parameter combination between the UAV flight mode feature set and the preset behavior pattern library, and a cross-validation mechanism is introduced to generate a candidate matching parameter set. The application of the genetic algorithm improves the search efficiency, ensures that the optimal parameter combination is found, and enhances the generalization ability of the system; based on the candidate matching parameter set, the particle swarm optimization technology is used to optimize the matching parameters, iteratively calculate and evaluate the fitness of each candidate matching parameter, and update the optimal matching parameters in combination with the time consistency evaluation score. At the same time, the local search strategy is implemented to adjust the initial similarity score matrix to obtain the optimized similarity score matrix. This process ensures the refined optimization of the matching parameters and further improves the matching accuracy; according to the optimized similarity score matrix, the behavior patterns that meet the preset behavior pattern library are screened, and the hierarchical clustering analysis is applied to classify the behavior patterns that meet the preset behavior pattern library into the same category, and the behavior patterns that meet the preset matching accuracy are determined, the optimized matching results are generated, and a unique identifier is assigned to each optimized matching result. This method not only ensures the reliability of the matching results, but also facilitates subsequent tracking and management, and improves the overall performance of the system.

[0053] These and other aspects of the present invention will become more apparent from the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0055] Figure 1 A flowchart of an anti-UAV defense method based on radar and audio provided by an embodiment of the present invention;

[0056] Figure 2 A schematic diagram of the structure of an anti-UAV defense system based on radar and audio provided by an embodiment of the present invention;

[0057] Figure 3 A schematic diagram of the structure of a computing device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0058] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0059] In some of the processes described in the specification and claims of the present invention and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0060] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in 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 embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0061] Figure 1 A flowchart of an anti-UAV defense method based on radar and audio is provided for an embodiment of the present invention, such as Figure 1 As shown, the method includes:

[0062] Step 101: constructing a three-dimensional airspace monitoring network by deploying radar sensors and audio capture devices at different heights and positions, wherein the radar sensors are used to transmit and receive reflected signal data to detect objects in the airspace, and the audio capture devices are used to capture real-time audio data to record ambient sound;

[0063] In this step, a three-dimensional airspace monitoring network is constructed by deploying radar sensors and audio capture devices at different heights and locations. Radar sensors are used to transmit and receive reflected signal data to detect objects in the airspace, including the distance, speed, and angle information of the target; audio capture devices are used to capture real-time audio data and record the sound characteristics of the surrounding environment, such as the sound frequency and intensity of drone propellers. This step ensures comprehensive coverage and multi-dimensional perception of all potential threats in the airspace;

[0064] First, multiple radar sensors and audio capture devices are deployed at different heights and locations according to the preset geographical area and security requirements. Each radar sensor is responsible for emitting electromagnetic waves and receiving reflected signals to generate a reflected signal data set; each audio capture device continuously records the audio information of the surrounding environment to form a real-time audio data set. By rationally arranging these sensors, a three-dimensional airspace monitoring network that can cover the entire monitoring area can be constructed to ensure that there are no blind spots in monitoring;

[0065] For example, in a security protection case around an airport, in order to protect the airport from unauthorized drone interference, the system deployed multiple radar sensors at high points (such as towers, building tops) and low points (such as ground control stations) around the airport, and set up audio capture devices at key locations. The radar sensors not only cover core areas such as runways and aprons, but also extend to a certain range outside the airport to ensure all-round monitoring. The audio capture device especially strengthens the collection of acoustic information at both ends of the runway and near the terminal building to ensure that any approaching drones can be detected in time. In this way, the system builds a strict three-dimensional airspace monitoring network, providing a solid data foundation for subsequent behavioral analysis.

[0066] Step 102: using the three-dimensional airspace monitoring network, implementing time-space synchronized multimodal information fusion, aligning the reflected signal data from the radar sensor and the real-time audio data acquired by the audio capture device on the time axis to obtain signal audio alignment data, and associating the target dynamics captured by each radar sensor through a spatial positioning algorithm to generate a real-time behavior trajectory of the target in the three-dimensional space;

[0067] In this step, the three-dimensional airspace monitoring network is used to implement time-space synchronized multimodal information fusion. This step aligns the reflected signal data from the radar sensor and the real-time audio data obtained by the audio capture device on the time axis to obtain signal audio alignment data. The spatial positioning algorithm is used to associate the target dynamics captured by each radar sensor to generate the real-time behavior trajectory of the target in three-dimensional space. This step ensures that the data obtained from different sensors can accurately correspond in time and space, enhancing the perception accuracy of the system;

[0068] The system first uses high-precision clock synchronization technology and distributed computing architecture to ensure the precise alignment of data from different sensors on the time axis. Then, it applies spatial positioning algorithms, combines geographic information system (GIS) data and three-dimensional terrain models, integrates the target dynamic information captured by each radar sensor, and generates a continuous behavior trajectory of the target in three-dimensional space. This process not only improves the accuracy of target positioning, but also provides a detailed description of the target's motion path;

[0069] Continuing with the airport case above, the system uses high-precision clock synchronization technology to ensure the precise alignment of radar reflection signal data and audio data on the time axis. Subsequently, a complex spatial positioning algorithm is applied, combined with the airport's GIS data and three-dimensional terrain model, to integrate the data from radar sensors and audio capture devices at different locations to generate the real-time behavior trajectory of the drone in three-dimensional space. For example, when an unauthorized drone approaches an airport, the system can not only determine its specific location, but also track its flight path, providing detailed information for subsequent behavior analysis.

[0070] Step 103: Based on the real-time behavior trajectory, the behavior profile modeling technology is combined with the machine learning algorithm to extract the UAV flight mode characteristics from the signal audio alignment data, and the UAV flight mode characteristics are compared with the preset UAV behavior pattern library in real time to obtain a behavior profile with unique identification, and the behavior profile with unique identification is analyzed to obtain an analysis result;

[0071] In this step, based on the real-time behavior trajectory, the behavior profile modeling technology combined with the machine learning algorithm is used to extract the UAV flight mode features from the signal audio alignment data. These features include flight speed, acceleration, altitude change rate, hovering duration, steering angle and frequency, etc. The system compares the extracted UAV flight mode features with the preset UAV behavior pattern library in real time, identifies the behavior profiles with unique identifiers, and analyzes these behavior profiles to generate detailed analysis results. This step ensures a deep understanding and classification of UAV behavior;

[0072] The system first uses behavioral profile modeling technology combined with machine learning algorithms to analyze and process real-time behavioral trajectories, extract detailed drone flight pattern features, and generate drone flight pattern feature sets. These feature sets are then compared in real time with the preset drone behavior pattern library to identify behavioral patterns that match preset values ​​and mark them as preliminary matching results. Next, time series analysis is introduced through deep learning algorithms to evaluate the temporal consistency of behavioral patterns, and multi-scale feature detection methods are used to capture the behavioral characteristics of drones at different scales, optimize matching accuracy, and ultimately determine a uniquely identifiable behavioral profile and generate a behavior analysis report;

[0073] Continuing with the previous airport case, the system extracts the drone's flight speed, acceleration, altitude change rate, hovering duration, and steering angle and frequency by analyzing the real-time behavior trajectory. After comparing these features with the preset drone behavior pattern library in real time, the system identifies that the drone may be conducting reconnaissance activities and generates a detailed analysis report. The report points out that the drone has hovered in a specific area many times and exhibited an irregular flight path, indicating that it may be trying to obtain sensitive information. This detailed behavior analysis provides an important basis for subsequent threat assessment.

[0074] Step 104: construct a threat level determination model according to the analysis result, and use the threat level determination model to evaluate the intention and risk level of the abnormal behavior of the drone to generate a threat assessment result, use an adaptive threat assessment engine to select defense measures corresponding to the threat level according to the threat assessment result, and combine the preset prediction model to predict the action trend of the drone to generate a target defense plan;

[0075] In this step, a threat level determination model is constructed based on the analysis results, and the model is used to evaluate the intention and risk level of the drone's abnormal behavior to generate a threat assessment result. The adaptive threat assessment engine selects defense measures corresponding to the threat level based on the threat assessment results, and combines the prediction model to predict the drone's action trend and generate a target defense plan. This step ensures that the system can respond quickly to potential threats and take appropriate defense measures to minimize losses;

[0076] The system first builds a threat level determination model based on the behavior analysis report, evaluates the intention and risk level of the drone's abnormal behavior, and generates a threat assessment result. The adaptive threat assessment engine selects corresponding defense measures based on the evaluation results, such as non-lethal means such as electromagnetic interference, directed energy weapons, or safety net capture. At the same time, the system combines the prediction model to predict the future action trends of the drone and develop a detailed comprehensive defense plan. The defense measure priority sorting mechanism ensures the flexibility and effectiveness of the response strategy;

[0077] Continuing with the previous airport case, based on the information in the behavior analysis report, the system built a threat level determination model, evaluated the drone's intentions and risk level, and generated a threat assessment result. The adaptive threat assessment engine selects defense measures that match the threat level, such as activating electromagnetic interference equipment to prevent the drone from moving forward. At the same time, the system combines the prediction model to predict that the drone may try to detour to another undisturbed direction. Based on this prediction, the system deploys additional safety net capture devices in advance to ensure that the drone can be effectively intercepted once it deviates from the original path. This forward-looking defense solution significantly improves the system's reaction speed and response effect.

[0078] Through the above steps, the present invention realizes a complete process from comprehensive monitoring to intelligent analysis, and then to accurate evaluation and rapid response; Step 101 ensures comprehensive coverage of all potential threats in the airspace by constructing a three-dimensional airspace monitoring network; Step 102 improves the accuracy and reliability of target positioning through multimodal information fusion with time and space synchronization; Step 103 achieves in-depth understanding and classification of drone behavior through behavioral profile modeling and machine learning algorithms; Step 104 ensures that the system can quickly respond to potential threats and minimize losses by constructing a threat level determination model and selecting appropriate defense measures. Overall, this method not only has efficient monitoring capabilities and intelligent analysis functions, but also performs well in behavioral pattern recognition and matching optimization, effectively improving the system's ability to deal with complex threats.

[0079] In order to solve the problems of inaccurate drone behavior recognition and insufficient analysis in the prior art, in some embodiments, as described in step 103, based on the real-time behavior trajectory, the behavior profile modeling technology is combined with the machine learning algorithm to extract the drone flight mode characteristics from the signal audio alignment data, and the drone flight mode characteristics are compared with the preset drone behavior pattern library in real time to obtain a uniquely identifiable behavior profile, and the behavior profile with the unique identification is analyzed to obtain the analysis result, which specifically includes:

[0080] The real-time behavior trajectory is analyzed and processed by combining the behavior profile modeling technology with the machine learning algorithm, the UAV flight mode features are extracted, and the UAV flight mode feature set is generated, wherein the UAV flight mode features include: flight speed, acceleration, altitude change rate, hovering time, and steering angle and frequency; the UAV flight mode feature set is compared with the preset UAV behavior pattern library in real time, the behavior pattern matching the preset value is identified and marked as the preliminary matching result, and the similarity score of each behavior pattern is calculated to obtain a preliminary matching list; based on the preliminary matching list, the deep learning algorithm is used for analysis, time series analysis is introduced to evaluate the temporal consistency of the behavior pattern, and multi-scale feature detection is used to detect the behavior pattern matching. The detection method captures the behavioral characteristics of the UAV at different scales, optimizes the matching accuracy of the UAV flight mode feature set and the preset UAV behavior pattern library, determines a behavior profile with unique identification, and generates an optimized behavior profile; based on the optimized behavior profile, the UAV's action intention is analyzed in combination with historical behavior records and prediction models to generate a behavior analysis report, wherein the reported behavior analysis includes flight path planning, hovering purpose, and intention to approach a target; combined with the information in the behavior analysis report, the evaluation rules of the adaptive threat assessment engine are adjusted to obtain an adjusted adaptive threat assessment engine, and the UAV's behavior pattern is analyzed again using the adjusted adaptive threat assessment engine to generate a target analysis result;

[0081] In this embodiment, the UAV flight mode features include detailed dynamic parameters such as flight speed, acceleration, altitude change rate, hovering time, and steering angle and frequency. These feature data are used to describe the motion characteristics of the UAV and provide detailed information about its behavior. The behavior profile modeling technology combined with the machine learning algorithm can not only extract the above flight mode features and generate a UAV flight mode feature set by analyzing and processing the real-time behavior trajectory, but also combine historical behavior records to provide a more comprehensive behavior description. The preset UAV behavior pattern library is a database containing a variety of known UAV behavior patterns, each of which has specific flight feature parameters. The system compares the extracted flight mode feature set with the patterns in the library in real time, identifies the behavior pattern that matches the preset value, and marks it as a preliminary matching result, and calculates the similarity score of each behavior pattern to form a preliminary matching list. The deep learning algorithm further optimizes the matching process, introduces time series analysis to evaluate the temporal consistency of the behavior pattern, and uses a multi-scale feature detection method to capture the behavior characteristics of the UAV at different scales, ensuring the improvement of matching accuracy, and finally determines a uniquely identifiable behavior profile to generate an optimized behavior profile. Based on the optimized behavior profile, the system analyzes the drone's action intentions in combination with historical behavior records and prediction models, and generates a behavior analysis report. The report includes flight path planning, hovering purpose, and intention to approach the target, etc., which provides an important basis for subsequent threat assessment. Finally, the adaptive threat assessment engine adjusts the assessment rules based on the behavior analysis report, analyzes the drone's behavior pattern again, and generates the final target analysis results. This iterative optimization ensures the accuracy of the assessment and the timeliness of the response.

[0082] In an embodiment of the present application, first, the system uses behavioral profile modeling technology combined with a machine learning algorithm to analyze and process real-time behavioral trajectories, extract detailed drone flight mode features, and generate a drone flight mode feature set. Then, these feature sets are compared with the preset drone behavior pattern library in real time to identify behavioral patterns that match preset values, and calculate similarity scores to form a preliminary match list. Next, based on the preliminary match list, the system uses a deep learning algorithm for further analysis, introduces time series analysis and multi-scale feature detection methods, optimizes matching accuracy, and determines a uniquely identifiable behavioral profile. Subsequently, the system combines historical behavior records and prediction models to parse the drone's action intentions and generate a behavior analysis report. Finally, combined with the information in the report, the evaluation rules of the adaptive threat assessment engine are adjusted to generate the final target analysis results;

[0083] For example, in a security protection case around an airport, in order to protect the airport from unauthorized drone interference, after receiving the real-time behavior trajectory, the system first uses the behavior profile modeling technology combined with the machine learning algorithm to extract the flight speed, acceleration, altitude change rate, hovering time, turning angle and frequency of the drone, and generates the drone flight mode feature set. Then, the system compares these feature sets with the preset drone behavior pattern library in real time, and finds that a drone hovers near the runway many times and shows an irregular flight path. The system marks it as a preliminary match result and calculates the similarity score to form a preliminary match list; to further optimize the matching accuracy, the system uses deep learning algorithms for analysis, introduces time series analysis to evaluate the temporal consistency of the behavior pattern, and uses multi-scale feature detection methods to capture the behavior characteristics of the drone at different scales. After this optimization process, the system determines that the drone has a unique and identifiable behavior profile, indicating that it may be conducting reconnaissance activities; based on the optimized behavior profile, the system combines historical behavior records and prediction models to analyze the drone's action intentions and generate a behavior analysis report. The report pointed out that the drone had hovered in a specific area several times and exhibited an irregular flight path, indicating that it might be trying to obtain sensitive information. In addition, the report predicted that the drone might try to detour to another unmonitored direction; finally, the system combined the information in the behavior analysis report, adjusted the evaluation rules of the adaptive threat assessment engine, analyzed the behavior pattern of the drone again, and generated the final target analysis result. The system decided to activate the electromagnetic interference equipment to prevent the drone from moving forward, and deployed additional safety net capture devices in advance to ensure that the drone can be effectively intercepted once it deviates from the original path. This forward-looking defense solution significantly improves the system's reaction speed and response effect.

[0084] In order to improve the accuracy and robustness of drone behavior pattern recognition, according to the previous embodiment, based on the preliminary matching list, a deep learning algorithm is used for analysis, time series analysis is introduced to evaluate the temporal consistency of the behavior pattern, and a multi-scale feature detection method is used to capture the behavior characteristics of the drone at different scales, optimize the matching accuracy of the drone flight mode feature set and the preset drone behavior pattern library, determine a uniquely identifiable behavior profile, and generate an optimized behavior profile, specifically including:

[0085] Using a deep learning algorithm, the behavior patterns in the preliminary matching list are analyzed, and time series analysis is introduced to evaluate the time consistency of each behavior pattern to obtain a time consistency evaluation report; based on the time consistency evaluation report, a multi-scale feature detection method is used to capture the behavior characteristics of the drone at different scales under the same behavior pattern, and the frequency components in the behavior pattern are extracted in combination with frequency domain analysis to generate a multi-scale behavior feature description; based on the multi-scale behavior feature description, a comparative optimization algorithm is used to optimize the matching process between the drone flight mode feature set and the preset drone behavior pattern library, and the optimal matching parameters are calculated by genetic algorithm and particle swarm optimization technology, and the similarity score is adjusted to obtain an optimized matching result; based on the optimized matching result, cluster analysis and anomaly detection algorithms are applied to identify and determine an initial behavior profile with unique identification, and a unique identifier is assigned to each initial behavior profile to generate an optimized behavior profile;

[0086] In this embodiment, a deep learning algorithm is used to conduct an in-depth analysis of the behavior patterns in the preliminary matching list, and automatically learn and extract complex behavior features through a neural network model. Time series analysis is used to evaluate the temporal consistency of each behavior pattern, ensure the stability and coherence of the behavior pattern in the time dimension, and obtain a temporal consistency evaluation report. The multi-scale feature detection method is combined with frequency domain analysis to capture the behavioral characteristics of the drone at different scales under the same behavior pattern, and extract the frequency components in the behavior pattern to generate a multi-scale behavior feature description. The comparative optimization algorithm calculates the best matching parameters through genetic algorithms and particle swarm optimization techniques, adjusts the similarity score, and optimizes the matching process between the drone flight mode feature set and the preset drone behavior pattern library. Finally, cluster analysis and anomaly detection algorithms are used to identify and determine the initial behavior profile with unique identification, assign a unique identifier to each initial behavior profile, and generate an optimized behavior profile;

[0087] In an embodiment of the present application, the system first uses a deep learning algorithm to analyze the behavior patterns in the preliminary matching list, introduces time series analysis to evaluate the time consistency of each behavior pattern, and obtains a time consistency evaluation report. Then, based on the report, a multi-scale feature detection method is used to capture the behavioral characteristics of the drone at different scales under the same behavior pattern, and the frequency components in the behavior pattern are extracted in combination with frequency domain analysis to generate a multi-scale behavior feature description. Next, based on the multi-scale behavior feature description, the system uses a comparative optimization algorithm to optimize the matching process between the drone flight mode feature set and the preset drone behavior pattern library, calculates the best matching parameters through genetic algorithms and particle swarm optimization technology, and adjusts the similarity score to obtain the optimized matching results. Finally, based on the optimized matching results, cluster analysis and anomaly detection algorithms are applied to identify and determine the initial behavior profile with unique identification, assign a unique identifier to each initial behavior profile, and generate an optimized behavior profile, thereby ensuring the accuracy and robustness of behavior pattern recognition;

[0088] For example, in a critical infrastructure security protection scenario, after receiving the preliminary matching list, the system first uses a deep learning algorithm to analyze the behavior patterns in the list, introduces time series analysis to evaluate the temporal consistency of each behavior pattern, and finds that a drone hovers multiple times in a specific area and exhibits an irregular flight path. The system generates a temporal consistency assessment report, indicating that the temporal stability of the drone's behavior pattern is high. Then, based on the report, the system uses a multi-scale feature detection method to capture the behavioral characteristics of the drone at different scales under the same behavior pattern, and extracts the frequency components in the behavior pattern in combination with frequency domain analysis to generate a multi-scale behavioral feature description. These feature descriptions reveal the behavioral characteristics of the drone at different heights and speeds. Subsequently, based on the multi-scale behavioral feature description, the system uses a comparative optimization algorithm to optimize the matching process between the drone flight mode feature set and the preset drone behavior pattern library, calculates the best matching parameters through genetic algorithms and particle swarm optimization techniques, and adjusts the similarity score to obtain the optimized matching results. Finally, based on the optimized matching results, the system applies clustering analysis and anomaly detection algorithms to identify and determine the initial behavior profile with unique identification, assigns a unique identifier to each initial behavior profile, and generates an optimized behavior profile. This process not only improves the accuracy of behavioral pattern recognition, but also provides a solid foundation for subsequent threat assessment and ensures the system's efficient defense capabilities.

[0089] In order to improve the accuracy of drone behavior pattern recognition and matching precision, according to the previous embodiment, based on the multi-scale behavior feature description, a comparative optimization algorithm is used to optimize the matching process between the drone flight pattern feature set and the preset drone behavior pattern library, and the optimal matching parameters are calculated by genetic algorithm and particle swarm optimization technology, and the similarity score is adjusted to obtain the optimized matching results, which specifically include:

[0090] Using a comparative optimization algorithm, based on multi-scale behavioral feature descriptions, a preliminary comparison is made between the UAV flight mode feature set and the preset UAV behavior pattern library to obtain an initial similarity scoring matrix, and the time consistency evaluation score of each behavior pattern is recorded; based on the initial similarity scoring matrix, a genetic algorithm is applied to search for the best matching parameter combination between the UAV flight mode feature set and the preset UAV behavior pattern library, and a cross-validation mechanism is introduced to generate a candidate matching parameter set; based on the candidate matching parameter set, a particle swarm optimization technique is used to optimize the matching parameters, the fitness of each candidate matching parameter is iteratively calculated and evaluated, and the optimal matching parameters are updated in combination with the time consistency evaluation score, and a local search strategy is implemented to adjust the initial similarity scoring matrix to obtain an optimized similarity scoring matrix; based on the optimized similarity scoring matrix, the behavior patterns that meet the preset UAV behavior pattern library are screened, and the behavior patterns that meet the preset UAV behavior pattern library are classified into the same category by applying hierarchical clustering analysis, and the behavior patterns that meet the preset matching accuracy are determined, an optimized matching result is generated, and a unique identifier is assigned to each optimized matching result;

[0091] In this embodiment, the comparative optimization algorithm is used to perform a preliminary comparison between the UAV flight mode feature set and the preset UAV behavior pattern library, generate an initial similarity score matrix, and record the time consistency evaluation score of each behavior pattern. Genetic algorithm is a search algorithm based on natural selection and genetic mechanisms, which is used to search for the best matching parameter combination between the UAV flight mode feature set and the preset UAV behavior pattern library, and introduce a cross-validation mechanism to ensure the generalization ability of the model to generate a set of candidate matching parameters. Particle swarm optimization technology is an optimization algorithm based on swarm intelligence. It simulates the behavior of flocks of birds or schools of fish, iteratively calculates and evaluates the fitness of each candidate matching parameter, and updates the optimal matching parameter based on the time consistency evaluation score. Hierarchical clustering analysis is used to classify behavior patterns that conform to the preset UAV behavior pattern library into the same category, determine behavior patterns that meet the preset matching accuracy, and assign a unique identifier to each optimized matching result;

[0092] In the embodiment of the present application, the system first uses a comparative optimization algorithm to perform a preliminary comparison between the UAV flight mode feature set and the preset UAV behavior pattern library based on a multi-scale behavior feature description, generates an initial similarity scoring matrix, and records the time consistency evaluation score of each behavior pattern. Then, based on the initial similarity scoring matrix, a genetic algorithm is applied to search for the best matching parameter combination between the UAV flight mode feature set and the preset UAV behavior pattern library, and a cross-validation mechanism is introduced to generate a candidate matching parameter set. Next, based on the candidate matching parameter set, the particle swarm optimization technology is used to optimize the matching parameters, the fitness of each candidate matching parameter is iteratively calculated and evaluated, and the optimal matching parameters are updated in combination with the time consistency evaluation score. At the same time, the system implements a local search strategy, adjusts the initial similarity scoring matrix, and obtains an optimized similarity scoring matrix. Finally, based on the optimized similarity scoring matrix, the behavior patterns that meet the preset UAV behavior pattern library are screened, and the hierarchical clustering analysis is applied to classify these behavior patterns into the same category, and the behavior patterns that meet the preset matching accuracy are determined, and the optimized matching results are generated, and a unique identifier is assigned to each optimized matching result, ensuring the accuracy and reliability of the matching results;

[0093] For example, in a critical infrastructure security protection scenario, after receiving a multi-scale behavioral feature description, the system first uses a comparative optimization algorithm to perform a preliminary comparison between the UAV flight mode feature set and the preset UAV behavior pattern library, generates an initial similarity score matrix, and records the time consistency evaluation score of each behavior pattern. The system found that a UAV hovered multiple times in a specific area and exhibited an irregular flight path, and its time consistency score was high. Then, based on the initial similarity score matrix, the system applied a genetic algorithm to search for the best matching parameter combination between the UAV flight mode feature set and the preset UAV behavior pattern library, and introduced a cross-validation mechanism to generate a candidate matching parameter set. These parameter combinations have been verified to effectively capture the behavioral characteristics of the UAV. Subsequently, based on the candidate matching parameter set, the system uses particle swarm optimization technology to optimize the matching parameters, iteratively calculates and evaluates the fitness of each candidate matching parameter, and combines the time consistency evaluation score to update the optimal matching parameters. At the same time, the system implements a local search strategy to adjust the initial similarity score matrix to obtain an optimized similarity score matrix. Finally, based on the optimized similarity score matrix, the system screens the behavior patterns that meet the preset drone behavior pattern library, applies hierarchical clustering analysis to classify these behavior patterns into the same category, and determines the behavior patterns that meet the preset matching accuracy, generates optimized matching results, and assigns a unique identifier to each optimized matching result. This process not only improves the accuracy of behavior pattern recognition, but also provides a solid foundation for subsequent threat assessment, ensuring the system's efficient defense capabilities.

[0094] In order to solve the problems of limited monitoring range, difficult data synchronization and insufficient positioning accuracy in the prior art, as another embodiment, according to step 101, a three-dimensional airspace monitoring network is constructed by deploying radar sensors and audio capture devices at different heights and positions, wherein the radar sensor is used to transmit and receive reflected signal data to detect objects in the airspace, and the audio capture device is used to capture real-time audio data to record the surrounding environment sound, specifically including:

[0095] Deploy radar sensors at different heights and positions in a preset airspace to generate a spatial distribution of radar sensors, wherein each radar sensor is used to transmit and receive reflected signal data to detect objects in the airspace, and obtain a radar reflected signal data set; deploy audio capture devices according to the spatial distribution of the radar sensors, distributed at positions coordinated with the radar sensors, and record the sounds of the surrounding environment in real time to obtain a real-time audio data set; based on the radar reflected signal data set and the real-time audio data set, use high-precision clock synchronization technology and distributed computing architecture to perform spatiotemporal synchronization processing on the radar reflected signal data set and the real-time audio data set to obtain signal audio alignment data; apply a spatial positioning algorithm, combine geographic information system data and a three-dimensional terrain model, and associate and integrate the reflected signal data of each radar sensor and the real-time audio data of each audio capture device to build a three-dimensional airspace monitoring network;

[0096] In this embodiment, the spatial distribution of radar sensors refers to the deployment of radar sensors at different heights and positions in a preset airspace. Each radar sensor is responsible for transmitting and receiving reflected signal data to detect objects in the airspace and generate radar reflected signal data sets. These data sets contain the distance, speed and angle information of the target, which are used to accurately identify and track objects in the airspace. The audio capture devices are distributed at locations coordinated with the radar sensors to record the sounds of the surrounding environment in real time to obtain real-time audio data sets. These data sets not only contain the sound frequency and intensity of the drone propellers, but may also capture other background noises, which are helpful in assisting in determining the existence of the drone and its approximate location. High-precision clock synchronization technology ensures the precise alignment of data from different sensors on the timeline, while the distributed computing architecture supports large-scale data processing and ensures the efficient operation of the system. The spatial positioning algorithm combines geographic information system (GIS) data and three-dimensional terrain models to correlate and integrate the reflected signal data of each radar sensor and the real-time audio data of each audio capture device to build a comprehensive three-dimensional airspace monitoring network;

[0097] In an embodiment of the present application, the system first deploys radar sensors at different heights and positions in a preset airspace, generates a spatial distribution of radar sensors, ensures that each radar sensor can effectively transmit and receive reflected signal data, detects objects in the airspace, and obtains a radar reflected signal data set. Next, according to the spatial distribution of radar sensors, the system deploys audio capture devices, which are distributed at positions coordinated with the radar sensors, and records the sounds of the surrounding environment in real time to obtain a real-time audio data set. In order to achieve time and space synchronization, the system uses high-precision clock synchronization technology and distributed computing architecture to process the radar reflection signal data set and the real-time audio data set to ensure the consistency of the two in time and space, and obtain signal audio alignment data. Finally, the spatial positioning algorithm is applied, combined with geographic information system data and three-dimensional terrain models, to correlate and integrate the reflected signal data of each radar sensor and the real-time audio data of each audio capture device, and to construct a three-dimensional airspace monitoring network to ensure comprehensive coverage and multi-dimensional perception of all potential threats in the airspace;

[0098] For example, in a security protection scenario around an airport, in order to protect the airport from unauthorized drone interference, the system first made a detailed plan for the airspace around the airport, determined the best solution for deploying radar sensors at different heights and locations, and generated the spatial distribution of radar sensors. Each radar sensor not only covers core areas such as runways and aprons, but also extends to a certain range outside the airport to ensure all-round monitoring. Subsequently, the system deployed audio capture devices based on the spatial distribution of radar sensors, especially strengthening the collection of acoustic information at both ends of the runway and near the terminal building to ensure that any approaching drones can be discovered in time. In order to ensure data synchronization, the system uses high-precision clock synchronization technology to ensure the precise alignment of radar reflection signal data sets and real-time audio data sets on the time axis; at the same time, a distributed computing architecture is used to process large amounts of data to ensure the efficient operation of the system. Next, the system applies a spatial positioning algorithm, combines the airport's GIS data and three-dimensional terrain model, associates and integrates the reflection signal data of each radar sensor and the real-time audio data of each audio capture device, and builds a strict three-dimensional airspace monitoring network. This process not only improves the accuracy of target positioning, but also enhances the overall defense capability of the system, ensuring that any abnormal behavior of drones can be quickly identified and responded to.

[0099] In order to solve the problems of insufficient multimodal information fusion, low data synchronization accuracy and insufficient target dynamic association in the prior art, as another embodiment, according to step 102, the three-dimensional airspace monitoring network is used to implement spatiotemporal synchronized multimodal information fusion, and the reflected signal data from the radar sensor and the real-time audio data obtained by the audio capture device are aligned on the time axis to obtain signal audio alignment data, and the target dynamics captured by each radar sensor are associated through a spatial positioning algorithm to generate a real-time behavior trajectory of the target in the three-dimensional space, specifically including:

[0100] Using high-precision clock synchronization technology and distributed computing architecture, the reflected signal data from the radar sensor and the real-time audio data obtained by the audio capture device are aligned on the time axis to obtain initial signal audio alignment data; based on the signal audio alignment data, multi-level preprocessing is performed, and the reflected signal data of each radar sensor and the real-time audio data obtained by each audio capture device are calibrated through a calibration program to generate optimized signal audio alignment data, wherein the multi-level preprocessing includes: noise filtering, data smoothing, outlier detection and frequency domain analysis; based on the optimized signal audio alignment data, a spatial positioning algorithm is applied, combined with geographic information system data, a three-dimensional terrain model and environmental meteorological data, the target dynamic information captured by each radar sensor is associated, and a preliminary positioning result of the target in the three-dimensional space is generated; based on the preliminary positioning result, an advanced trajectory tracking algorithm is applied to process and generate a real-time behavior trajectory of the target in the three-dimensional space;

[0101] In this embodiment, high-precision clock synchronization technology ensures the precise alignment of data from different sensors on the time axis, eliminating the time deviation problem caused by clock asynchrony. The distributed computing architecture supports large-scale data processing and improves the response speed and efficiency of the system. Multi-level preprocessing covers steps such as noise filtering, data smoothing, outlier detection, and frequency domain analysis, which are used to purify and optimize the original data and reduce the impact of interference factors. The calibration program is used to calibrate the reflected signal data of each radar sensor and the real-time audio data obtained by each audio capture device to ensure the consistency and accuracy of the data. The spatial positioning algorithm combines geographic information system (GIS) data, three-dimensional terrain models, and environmental meteorological data to associate the target dynamic information captured by each radar sensor and generate a preliminary positioning result of the target in three-dimensional space. The advanced trajectory tracking algorithm is used to process and generate the real-time behavior trajectory of the target in three-dimensional space, providing a more accurate and continuous description of the target motion;

[0102] In the embodiment of the present application, the system first uses high-precision clock synchronization technology and distributed computing architecture to align the reflected signal data from the radar sensor and the real-time audio data obtained by the audio capture device on the time axis to obtain initial signal audio alignment data. Then, based on these alignment data, the system performs multi-level preprocessing, and calibrates the reflected signal data of each radar sensor and the real-time audio data obtained by each audio capture device through a calibration program to generate optimized signal audio alignment data. The multi-level preprocessing includes noise filtering, data smoothing, outlier detection and frequency domain analysis to ensure the quality and reliability of the data. Based on the preprocessed signal audio alignment data, the system applies a spatial positioning algorithm, combines geographic information system data, three-dimensional terrain models, and environmental meteorological data, associates the target dynamic information captured by each radar sensor, and generates a preliminary positioning result of the target in three-dimensional space. Finally, based on the preliminary positioning results, the system applies an advanced trajectory tracking algorithm to process and generate a real-time behavior trajectory of the target in three-dimensional space, ensuring accurate tracking and description of the target dynamics;

[0103] For example, in a security protection scenario around an airport, in order to protect the airport from unauthorized drone interference, the system first uses high-precision clock synchronization technology and distributed computing architecture to align the reflected signal data from the radar sensor and the real-time audio data obtained by the audio capture device on the time axis to obtain the initial signal audio alignment data. When the system finds a drone approaching the airport, it immediately starts multi-level preprocessing, including noise filtering, data smoothing, outlier detection and frequency domain analysis, to ensure the purity and consistency of the data. Subsequently, the system calibrates the reflected signal data of each radar sensor and the real-time audio data obtained by each audio capture device through a calibration program to generate optimized signal audio alignment data, further improving the accuracy and reliability of the data. Next, based on the preprocessed signal audio alignment data, the system applies a spatial positioning algorithm, combines the airport's geographic information system data, the three-dimensional terrain model, and the environmental meteorological data at the time, associates the target dynamic information captured by each radar sensor, and generates a preliminary positioning result of the target in three-dimensional space. Finally, based on the preliminary positioning results, the system applies an advanced trajectory tracking algorithm to process and generate the real-time behavior trajectory of the target in three-dimensional space, accurately depicting the flight path of the drone. This process not only improves the accuracy of target positioning, but also enhances the overall defense capability of the system, ensuring that any abnormal behavior of the drone can be quickly identified and responded to.

[0104] In order to solve the problems of inaccurate threat assessment and lack of flexibility in defense measures selection in the prior art and to improve the accuracy and response speed of abnormal behavior identification of drones, as another embodiment, according to step 104, a threat level determination model is constructed according to the analysis result, and the intention and risk level of the abnormal behavior of the drone are evaluated by using the threat level determination model to generate a threat assessment result, and an adaptive threat assessment engine is used to select defense measures corresponding to the threat level according to the threat assessment result, and the action trend of the drone is predicted in combination with a preset prediction model to generate a target defense plan, which specifically includes:

[0105] Using the analysis results, the UAV flight mode characteristics and historical behavior data are integrated, real-time environmental parameters and context information are introduced, multiple threat levels and corresponding evaluation rules for multiple threat levels are defined, and a dynamic threat level determination model is obtained; according to the dynamic threat level determination model, a multi-dimensional comprehensive evaluation is performed on the abnormal behavior of the UAV identified in the analysis results, and by comparing the behavior pattern of the UAV with the preset UAV behavior pattern library, combined with the situational awareness algorithm, the intention and risk level of the abnormal behavior of the UAV are determined from multiple angles to generate a preliminary threat assessment result; based on the preliminary threat assessment result, the intelligent adaptive threat assessment engine is activated, and threat assessment suggestions are provided in combination with the current assessment results, historical assessment records and preset prediction models to generate a target threat assessment result; using the target threat assessment result, multi-level defense measures that match the threat level are selected to generate a priority ranking of defense measures; using the preset prediction model, combined with the historical behavior, real-time behavior trajectory and external intelligence information of the UAV, the action trend of the UAV is predicted to obtain a prediction result, and a target defense plan is formulated based on the prediction result and the priority ranking of defense measures;

[0106] In this embodiment, the analysis results integrate the UAV flight mode characteristics and historical behavior data, and introduce real-time environmental parameters and contextual information to define multiple threat levels and corresponding evaluation rules, and build a dynamic threat level determination model. The multi-dimensional comprehensive evaluation determines the intention and risk level of the UAV's abnormal behavior from multiple angles by comparing the UAV's behavior pattern with the preset UAV behavior pattern library, combined with the situational awareness algorithm. The intelligent adaptive threat assessment engine can provide threat assessment recommendations based on the current assessment results and historical assessment records to ensure dynamic adjustment and continuous optimization of the assessment. Multi-level defense measures refer to selecting matching defense strategies according to different threat levels, forming a priority ranking of defense measures, and ensuring the effectiveness and timeliness of response measures. The preset prediction model combines the UAV's historical behavior, real-time behavior trajectory, and external intelligence information to predict the UAV's action trend and support the formulation of more accurate target defense plans;

[0107] In the embodiment of the present application, the system first uses the analysis results to integrate the flight mode characteristics and historical behavior data of the drone, introduces real-time environmental parameters and context information, defines multiple threat levels and corresponding evaluation rules, and obtains a dynamic threat level determination model. Then, according to the dynamic threat level determination model, the system performs a multi-dimensional comprehensive evaluation of the abnormal behavior of the drone identified in the analysis results, and determines the intention and risk level of the abnormal behavior of the drone from multiple angles by comparing the behavior pattern of the drone with the preset drone behavior pattern library, combined with the situational awareness algorithm, and generates a preliminary threat assessment result. Based on the preliminary threat assessment results, the intelligent adaptive threat assessment engine is activated, and threat assessment suggestions are provided in combination with the current assessment results, historical assessment records and preset prediction models to generate target threat assessment results. Next, using the target threat assessment results, the system selects multi-level defense measures that match the threat level, generates a priority ranking of defense measures, and ensures the effectiveness and pertinence of the defense measures. Finally, using the preset prediction model, combined with the historical behavior, real-time behavior trajectory and external intelligence information of the drone, the action trend of the drone is predicted to obtain the prediction results, and based on the prediction results and the priority ranking of defense measures, a target defense plan is formulated to ensure the responsiveness and defense effect of the system;

[0108] For example, in a security protection scenario around an airport, in order to protect the airport from unauthorized drone interference, the system first uses the analysis results to integrate the drone flight pattern characteristics and historical behavior data, introduces real-time environmental parameters such as weather conditions, time and other factors, as well as contextual information such as the location of key facilities approached by the drone, defines multiple threat levels and corresponding evaluation rules, and builds a dynamic threat level determination model. The system finds that a drone hovers multiple times in a specific area and exhibits an irregular flight path. It immediately initiates a multi-dimensional comprehensive evaluation, compares the drone's behavior pattern with the preset drone behavior pattern library, and combines the situational awareness algorithm to determine the intention and risk level of the drone's abnormal behavior from multiple angles, and generates a preliminary threat assessment result. The preliminary results show that the drone may have reconnaissance intentions and has a high risk level. Based on this result, the system activates the intelligent adaptive threat assessment engine, combines the current assessment results with historical assessment records and preset prediction models, provides threat assessment suggestions, generates target threat assessment results, and confirms that the drone's behavior constitutes a high-risk threat. Subsequently, the system uses the target threat assessment results to select multi-level defense measures that match the threat level, such as activating electromagnetic interference equipment to prevent the drone from moving forward, and deploying additional safety net capture devices in advance, generating defense measures priority rankings, and ensuring the effectiveness and pertinence of defense measures. Finally, using the preset prediction model, combined with the drone's historical behavior, real-time behavior trajectory, and external intelligence information, the drone's action trend is predicted to obtain the prediction results. Based on the prediction results and defense measures priority rankings, a target defense plan is formulated to ensure the system's responsiveness and defense effectiveness. This process not only improves the accuracy of threat assessment, but also enhances the system's overall defense capabilities, ensuring that any abnormal behavior of the drone can be quickly identified and effectively responded to.

[0109] Figure 2 A schematic diagram of the structure of an anti-UAV defense system based on radar and audio is provided for an embodiment of the present invention, such as Figure 2 As shown, the system includes:

[0110] A construction module 21 is used to construct a three-dimensional airspace monitoring network by deploying radar sensors and audio capture devices at different heights and positions, wherein the radar sensors are used to transmit and receive reflected signal data to detect objects in the airspace, and the audio capture devices are used to capture real-time audio data to record ambient sound;

[0111] An alignment module 22 is used to utilize the three-dimensional airspace monitoring network to implement spatiotemporal synchronized multimodal information fusion, align the reflected signal data from the radar sensor and the real-time audio data acquired by the audio capture device on the time axis, obtain signal audio alignment data, and associate the target dynamics captured by each radar sensor through a spatial positioning algorithm to generate a real-time behavior trajectory of the target in the three-dimensional space;

[0112] A comparison module 23 is used to extract the UAV flight mode features from the signal audio alignment data based on the real-time behavior trajectory by using the behavior profile modeling technology combined with the machine learning algorithm, and to compare the UAV flight mode features with the preset UAV behavior pattern library in real time to obtain a behavior profile with unique identification, and to analyze the behavior profile with unique identification to obtain an analysis result;

[0113] The evaluation module 24 is used to construct a threat level determination model based on the analysis results, and use the threat level determination model to evaluate the intention and risk level of the abnormal behavior of the drone to generate a threat assessment result, and use an adaptive threat assessment engine to select defense measures corresponding to the threat level based on the threat assessment result, and combine the preset prediction model to predict the action trend of the drone to generate a target defense plan.

[0114] Figure 2 The radar and audio based anti-drone defense system can perform Figure 1 The implementation principle and technical effect of the radar and audio-based anti-UAV defense method described in the illustrated embodiment will not be described in detail. The specific manner in which each module and unit performs operations in the radar and audio-based anti-UAV defense system in the above embodiment has been described in detail in the embodiment of the method, and will not be described in detail here.

[0115] In one possible design, Figure 2 A radar and audio-based anti-drone defense system of the illustrated embodiment may be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32;

[0116] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0117] The processing component 32 is used to: construct a three-dimensional airspace monitoring network by deploying radar sensors and audio capture devices at different heights and positions, wherein the radar sensor is used to transmit and receive reflected signal data to detect objects in the airspace, and the audio capture device is used to capture real-time audio data to record ambient sound; utilize the three-dimensional airspace monitoring network to implement spatiotemporal synchronized multimodal information fusion, align the reflected signal data from the radar sensor and the real-time audio data obtained by the audio capture device on the time axis to obtain signal audio alignment data, and associate the target dynamics captured by each radar sensor through a spatial positioning algorithm to generate a real-time behavior trajectory of the target in the three-dimensional space; based on the real-time behavior trajectory, The behavior profile modeling technology is combined with a machine learning algorithm to extract the UAV flight mode characteristics from the signal audio alignment data, and the UAV flight mode characteristics are compared with a preset UAV behavior pattern library in real time to obtain a behavior profile with a unique identifier, and the behavior profile with a unique identifier is analyzed to obtain a parsing result; a threat level determination model is constructed according to the parsing result, and the threat level determination model is used to evaluate the intention and risk level of the abnormal behavior of the UAV to generate a threat assessment result, and an adaptive threat assessment engine is used to select defense measures corresponding to the threat level according to the threat assessment result, and the action trend of the UAV is predicted in combination with a preset prediction model to generate a target defense plan.

[0118] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above method.

[0119] The storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0120] Of course, the computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0121] The input / output interface provides an interface between the processing component and the peripheral interface module, which may be an output device, an input device, etc.

[0122] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0123] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0124] The embodiment of the present invention further provides a computer storage medium storing a computer program, which can achieve the above-mentioned Figure 1 A radar and audio based anti-drone defense method of the illustrated embodiment.

[0125] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0126] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0127] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0128] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A radar and audio based anti-UAV defense method, characterized in that: include: A three-dimensional airspace monitoring network is constructed by deploying radar sensors and audio capture devices at different heights and locations, wherein the radar sensors are used to transmit and receive reflected signal data to detect objects in the airspace, and the audio capture devices are used to capture real-time audio data to record ambient sound; Utilizing the three-dimensional airspace monitoring network, implementing spatiotemporal synchronized multimodal information fusion, aligning the reflected signal data from the radar sensor and the real-time audio data acquired by the audio capture device on the time axis to obtain signal audio alignment data, and associating the target dynamics captured by each radar sensor through a spatial positioning algorithm to generate a real-time behavior trajectory of the target in the three-dimensional space; Based on the real-time behavior trajectory, the behavior profile modeling technology is combined with the machine learning algorithm to extract the UAV flight mode characteristics from the signal audio alignment data, and the UAV flight mode characteristics are compared with the preset UAV behavior pattern library in real time to obtain a behavior profile with unique identification, and the behavior profile with unique identification is analyzed to obtain an analysis result; A threat level determination model is constructed according to the analysis results, and the intention and risk level of the abnormal behavior of the drone is evaluated by using the threat level determination model to generate a threat assessment result, and a defense measure corresponding to the threat level is selected according to the threat assessment result by using an adaptive threat assessment engine, and the action trend of the drone is predicted in combination with a preset prediction model to generate a target defense plan; The step of comparing the UAV flight mode characteristics with a preset UAV behavior pattern library in real time to obtain a uniquely identifiable behavior profile includes: Utilizing behavioral profile modeling technology combined with machine learning algorithms, the real-time behavioral trajectory is analyzed and processed, the UAV flight mode features are extracted, and a UAV flight mode feature set is generated; Compare the UAV flight mode feature set with a preset UAV behavior pattern library in real time, identify the behavior pattern matching the preset value, and mark it as a preliminary matching result, and calculate the similarity score of each behavior pattern to obtain a preliminary matching list; Based on the preliminary matching list, deep learning algorithms are used for analysis, time series analysis is introduced to evaluate the temporal consistency of behavior patterns, and a multi-scale feature detection method is used to capture the behavioral characteristics of drones at different scales, optimize the matching accuracy between the drone flight mode feature set and the preset drone behavior pattern library, and determine a uniquely identifiable behavior profile.

2. The method according to claim 1, characterized in that The analysis is performed based on the uniquely identifiable behavior profile to obtain analysis results, including: Based on the optimized behavior profile, the action intention of the UAV is analyzed in combination with the historical behavior record and the prediction model to generate a behavior analysis report, wherein the reported behavior analysis includes flight path planning, hovering purpose, and intention to approach the target; Combined with the information in the behavior analysis report, the assessment rules of the adaptive threat assessment engine are adjusted to obtain an adjusted adaptive threat assessment engine, and the behavior pattern of the drone is analyzed again using the adjusted adaptive threat assessment engine to generate a target analysis result.

3. The method according to claim 2, characterized in that Based on the preliminary matching list, a deep learning algorithm is used for analysis, time series analysis is introduced to evaluate the temporal consistency of the behavior pattern, and a multi-scale feature detection method is used to capture the behavior characteristics of the drone at different scales, optimize the matching accuracy of the drone flight mode feature set and the preset drone behavior pattern library, determine the behavior profile with unique identification, and generate an optimized behavior profile, including: Using deep learning algorithms, we analyze the behavior patterns in the preliminary matching list, introduce time series analysis to evaluate the temporal consistency of each behavior pattern, and obtain a temporal consistency evaluation report. According to the temporal consistency assessment report, a multi-scale feature detection method is used to capture the behavioral characteristics of the drone at different scales under the same behavior mode, and the frequency components in the behavior mode are extracted by combining frequency domain analysis to generate a multi-scale behavior feature description; Based on the multi-scale behavior feature description, a comparative optimization algorithm is used to optimize the matching process between the UAV flight mode feature set and the preset UAV behavior pattern library, and the optimal matching parameters are calculated by genetic algorithm and particle swarm optimization technology, and the similarity score is adjusted to obtain the optimized matching result; Based on the optimized matching results, cluster analysis and anomaly detection algorithms are applied to identify and determine initial behavior profiles with unique identification, and a unique identifier is assigned to each initial behavior profile to generate an optimized behavior profile.

4. The method according to claim 3, characterized in that Based on the multi-scale behavior feature description, the matching process between the UAV flight mode feature set and the preset UAV behavior pattern library is optimized by using the comparative optimization algorithm, the optimal matching parameters are calculated by using the genetic algorithm and particle swarm optimization technology, and the similarity score is adjusted to obtain the optimized matching results, including: Using the comparative optimization algorithm, based on the multi-scale behavior feature description, the UAV flight mode feature set is preliminarily compared with the preset UAV behavior pattern library to obtain the initial similarity score matrix, and the time consistency evaluation score of each behavior pattern is recorded; According to the initial similarity score matrix, a genetic algorithm is applied to search for the best matching parameter combination between the UAV flight mode feature set and the preset UAV behavior pattern library, and a cross-validation mechanism is introduced to generate a candidate matching parameter set; Based on the candidate matching parameter set, the particle swarm optimization technology is used to optimize the matching parameters, the fitness of each candidate matching parameter is evaluated by iterative calculation, and the optimal matching parameters are updated in combination with the time consistency evaluation score, and a local search strategy is implemented to adjust the initial similarity scoring matrix to obtain an optimized similarity scoring matrix; According to the optimized similarity scoring matrix, the behavior patterns that meet the preset UAV behavior pattern library are screened, and the hierarchical clustering analysis is applied to classify the behavior patterns that meet the preset UAV behavior pattern library into the same category, and the behavior patterns that meet the preset matching accuracy are determined, the optimized matching results are generated, and a unique identifier is assigned to each optimized matching result.

5. The method according to claim 1, characterized in that A three-dimensional airspace monitoring network is constructed by deploying radar sensors and audio capture devices at different heights and positions, wherein the radar sensors are used to transmit and receive reflected signal data to detect objects in the airspace, and the audio capture devices are used to capture real-time audio data to record ambient sound, including: Deploy radar sensors at different heights and positions in a preset airspace to generate a spatial distribution of radar sensors, wherein each radar sensor is used to transmit and receive reflection signal data to detect objects in the airspace, thereby obtaining a radar reflection signal data set; According to the spatial distribution of the radar sensors, audio capture devices are deployed at locations coordinated with the radar sensors, and the sounds of the surrounding environment are recorded in real time to obtain a real-time audio data set; Based on the radar reflection signal dataset and the real-time audio dataset, high-precision clock synchronization technology and distributed computing architecture are used to perform spatiotemporal synchronization processing on the radar reflection signal dataset and the real-time audio dataset to obtain signal audio alignment data; By applying spatial positioning algorithms, combining geographic information system data and three-dimensional terrain models, the reflected signal data of each radar sensor and the real-time audio data of each audio capture device are correlated and integrated to build a three-dimensional airspace monitoring network.

6. The method according to claim 1, characterized in that The three-dimensional airspace monitoring network is used to implement time-space synchronized multimodal information fusion, align the reflected signal data from the radar sensor and the real-time audio data obtained by the audio capture device on the time axis to obtain signal audio alignment data, and associate the target dynamics captured by each radar sensor through a spatial positioning algorithm to generate a real-time behavior trajectory of the target in the three-dimensional space, including: Using high-precision clock synchronization technology and distributed computing architecture, the reflected signal data from the radar sensor and the real-time audio data obtained by the audio capture device are aligned on the time axis to obtain initial signal audio alignment data; According to the signal audio alignment data, multi-level preprocessing is performed, and the reflected signal data of each radar sensor and the real-time audio data acquired by each audio capture device are calibrated through a calibration program to generate optimized signal audio alignment data, wherein the multi-level preprocessing includes: noise filtering, data smoothing, outlier detection and frequency domain analysis; Based on the optimized signal audio alignment data, a spatial positioning algorithm is applied, combined with geographic information system data, three-dimensional terrain model and environmental meteorological data, the target dynamic information captured by each radar sensor is associated to generate a preliminary positioning result of the target in three-dimensional space. According to the preliminary positioning result, an advanced trajectory tracking algorithm is applied to process and generate a real-time behavior trajectory of the target in three-dimensional space.

7. The method according to claim 1, characterized in that A threat level determination model is constructed according to the analysis results, and the intention and risk level of the abnormal behavior of the drone is evaluated by using the threat level determination model to generate a threat assessment result. An adaptive threat assessment engine is used to select defense measures corresponding to the threat level according to the threat assessment result, and the action trend of the drone is predicted in combination with a preset prediction model to generate a target defense plan, including: Using the analysis results, the flight mode characteristics and historical behavior data of drones are integrated, real-time environmental parameters and context information are introduced, multiple threat levels and corresponding evaluation rules are defined, and a dynamic threat level determination model is obtained; According to the dynamic threat level determination model, a multi-dimensional comprehensive evaluation is performed on the abnormal behavior of the drone identified in the analysis results. By comparing the behavior pattern of the drone with the preset drone behavior pattern library and combining the situational awareness algorithm, the intention and risk level of the abnormal behavior of the drone are determined from multiple angles to generate a preliminary threat assessment result. Based on the preliminary threat assessment results, an intelligent adaptive threat assessment engine is activated to provide threat assessment suggestions in combination with current assessment results, historical assessment records and preset prediction models to generate target threat assessment results; Using the target threat assessment results, select multi-level defense measures that match the threat level and generate a priority ranking of defense measures; Using the preset prediction model, combined with the drone's historical behavior, real-time behavior trajectory and external intelligence information, the drone's action trend is predicted to obtain the prediction results. Based on the prediction results and the priority of defense measures, a target defense plan is formulated.

8. A radar and audio based anti-UAV defense system, characterized in that: include: A construction module is used to construct a three-dimensional airspace monitoring network by deploying radar sensors and audio capture devices at different heights and positions, wherein the radar sensors are used to transmit and receive reflected signal data to detect objects in the airspace, and the audio capture devices are used to capture real-time audio data to record ambient sound; An alignment module is used to utilize the three-dimensional airspace monitoring network to implement time-space synchronized multimodal information fusion, align the reflected signal data from the radar sensor and the real-time audio data acquired by the audio capture device on the time axis to obtain signal audio alignment data, and associate the target dynamics captured by each radar sensor through a spatial positioning algorithm to generate a real-time behavior trajectory of the target in the three-dimensional space; A comparison module is used to extract the UAV flight mode features from the signal audio alignment data based on the real-time behavior trajectory by using the behavior profile modeling technology combined with the machine learning algorithm, and to compare the UAV flight mode features with a preset UAV behavior pattern library in real time to obtain a behavior profile with a unique identifier, and to parse the behavior profile with a unique identifier to obtain a parsing result; An assessment module, used to construct a threat level determination model according to the analysis result, and use the threat level determination model to evaluate the intention and risk level of the abnormal behavior of the drone to generate a threat assessment result, and use an adaptive threat assessment engine to select defense measures corresponding to the threat level according to the threat assessment result, and predict the action trend of the drone in combination with a preset prediction model to generate a target defense plan; The step of comparing the UAV flight mode characteristics with a preset UAV behavior pattern library in real time to obtain a uniquely identifiable behavior profile includes: Utilizing behavioral profile modeling technology combined with machine learning algorithms, the real-time behavioral trajectory is analyzed and processed, the UAV flight mode features are extracted, and a UAV flight mode feature set is generated; Compare the UAV flight mode feature set with a preset UAV behavior pattern library in real time, identify the behavior pattern matching the preset value, and mark it as a preliminary matching result, and calculate the similarity score of each behavior pattern to obtain a preliminary matching list; Based on the preliminary matching list, deep learning algorithms are used for analysis, time series analysis is introduced to evaluate the temporal consistency of behavior patterns, and a multi-scale feature detection method is used to capture the behavioral characteristics of drones at different scales, optimize the matching accuracy between the drone flight mode feature set and the preset drone behavior pattern library, determine a uniquely identifiable behavior profile, and generate an optimized behavior profile.

9. A computing device, characterized in that It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a radar and audio-based anti-UAV defense method as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a computer, an anti-UAV defense method based on radar and audio is implemented as described in any one of claims 1 to 7.

Images