Unmanned aerial vehicle countering method and system

Abstract

The invention discloses an unmanned aerial vehicle countering method and system, and the method comprises the steps: obtaining the behavior data of an unmanned aerial vehicle, carrying out the first registration processing of the behavior data of the unmanned aerial vehicle, and obtaining the first behavior data of the unmanned aerial vehicle; identifying the first unmanned aerial vehicle behavior data through a first neural network to obtain an identification result; and combining the identification result with a first countering strategy to generate a countering instruction, and determining the opportunity of countering action to execute the countering instruction so as to realize accurate positioning of the unmanned aerial vehicle.

Description

Technical Field

[0001] The present invention relates to the technical field of drone countermeasures, and in particular to a drone countermeasure method and system. Background Art

[0002] With the rapid development of drone technology, drones are increasingly used in military and civilian fields, but they also bring security threats. The intrusion and abuse of illegal drones have become a serious problem, and effective countermeasures are needed to protect the basic public safety and national privacy of power facilities.

[0003] Traditional countermeasures mainly rely on radar, optical sensors and other equipment, but the accuracy and real-time performance of these methods in complex environments need to be improved. Traditional radar and optical sensors have limitations in data processing and response speed, and cannot respond quickly, resulting in a delay in the implementation of countermeasures. Insufficient real-time performance may cause drones to cause damage or security threats before countermeasures are deployed. Summary of the invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a drone countermeasure method and system to solve the problem that traditional drone countermeasure methods usually rely on radar or visual recognition technology, and have inaccurate recognition or slow response speed in complex environments.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a method for countering a drone, comprising:

[0008] Acquire UAV behavior data, and perform a first registration process on the UAV behavior data to obtain first UAV behavior data;

[0009] Identify the first UAV behavior data using a first neural network to obtain an identification result;

[0010] The identification result is combined with the first countermeasure strategy to generate a countermeasure instruction, and the timing of the countermeasure action is determined to execute the countermeasure instruction.

[0011] As a preferred solution of the drone countermeasure method of the present invention, the identifying of the first drone behavior data includes:

[0012] Training the first drone behavior data using a first machine learning algorithm to obtain a first machine learning model;

[0013] The first neural network is used to learn the characteristic pattern of the UAV to obtain the first feature.

[0014] As a preferred solution of the drone countermeasure method of the present invention, it also includes:

[0015] Inputting the first feature into a first machine learning model to perform real-time identification of the drone;

[0016] The identified UAV is tracked using the first tracking algorithm to obtain abnormal behavior.

[0017] As a preferred solution of the drone countermeasure method of the present invention, determining the timing of the countermeasure action includes:

[0018] When the UAV's flight speed is greater than the first threshold and the distance from the monitoring point is less than the second threshold, the countermeasure command is executed.

[0019] As a preferred solution of the drone countermeasure method of the present invention, the first registration processing of the drone behavior data includes:

[0020] Use an external clock signal to synchronize the triggering of all sensors;

[0021] Each sensor adds a timestamp when collecting data, and adjusts the time difference through multiple fitting algorithms;

[0022] Measure the time delay between sensors and make time adjustments to the measured data;

[0023] When the temporal resolution of sensor data is inconsistent, the sensor data is interpolated using an interpolation method to match the temporal resolution of the sensor data.

[0024] As a preferred solution of the drone countermeasure method of the present invention, it also includes:

[0025] Transform data from different sensors into a common global coordinate system;

[0026] Use the rotation matrix to spatially transform the sensor data and correspond the data of different sensors in the same coordinate system;

[0027] Match the feature points between different images to obtain the spatial transformation relationship between images;

[0028] Use optimization to find the best spatial transformation parameters;

[0029] Each sensor is calibrated independently to obtain the internal and external parameters of each sensor, and the internal and external parameters are spatially aligned;

[0030] The result of the first registration is optimized through iteration until the predetermined progress requirement is met.

[0031] As a preferred solution of the drone countermeasure method of the present invention, the first countermeasure strategy includes:

[0032] Interference strategies, interception strategies, and capture strategies.

[0033] In a second aspect, the present invention provides a drone countermeasure system, comprising:

[0034] A registration processing module, used for acquiring UAV behavior data, performing a first registration processing on the UAV behavior data, and obtaining first UAV behavior data;

[0035] an identification module, configured to identify the first UAV behavior data through a first neural network to obtain an identification result;

[0036] The execution instruction module is used to generate a counter-instruction by combining the recognition result with the first counter-measure strategy, and determine the timing of the counter-measure action to execute the counter-instruction.

[0037] In a third aspect, the present invention provides a computing device, comprising:

[0038] Memory and processor;

[0039] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the drone countermeasure method are implemented.

[0040] In a fourth aspect, the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the drone countermeasure method.

[0041] Compared with the prior art, the present invention has the following beneficial effects: the present invention utilizes the spatiotemporal registration algorithm to accurately calculate the position and speed of the drone, and realizes the precise positioning of the drone by analyzing the spatiotemporal characteristics of the drone signal. This high-precision positioning capability makes countermeasures more accurate, reduces the possibility of accidental injuries, and can process and analyze drone signals in real time, and quickly respond to dynamic changes of drones. In the event of an emergency of drone intrusion, countermeasures can be quickly initiated to effectively curb potential threats. Compared with traditional drone countermeasures equipment, the system based on the spatiotemporal registration algorithm has simpler hardware requirements and mainly relies on efficient algorithm processing, thereby reducing the overall cost of the system. At the same time, due to its high performance and low maintenance requirements, the long-term operating costs are also relatively low. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0043] Figure 1 The figure is a logical schematic diagram of the overall process of a drone countermeasure method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0045] Example 1

[0046] Reference Figure 1 , as an embodiment of the present invention, provides a method for countering a drone, comprising:

[0047] S100: Acquire drone behavior data, perform a first registration process on the drone behavior data, and obtain first drone behavior data;

[0048] S200: Identifying the behavior data of the first drone through a first neural network to obtain an identification result;

[0049] S300: Combining the recognition result with the first countermeasure strategy to generate a countermeasure instruction, and determining the timing of the countermeasure action to execute the countermeasure instruction.

[0050] It should be noted that through precise space-time registration technology, illegal intrusion drones can be quickly located and effectively countered, which improves the accuracy and response speed of the counter-measure system. The space-time registration algorithm can reduce the risk of accidental injury and avoid damage to non-target objects through precise space-time positioning, ensuring the accuracy and safety of counter-measure actions. It can adapt to different environmental conditions, improve the versatility of counter-measure methods and their adaptability to complex environments. Combined with machine learning technology, it can realize automatic identification and counter-measures of drones, reduce human intervention, and improve the intelligence level of the counter-measure system.

[0051] In the embodiment of the present application, the above step S100 includes the following sub-steps A1-A10;

[0052] In A1: Use an external clock signal to synchronize all sensors.

[0053] In A2: each sensor adds a timestamp when collecting data, and the time difference is adjusted through a multi-fitting algorithm;

[0054] In A3: measuring the time delay between sensors and performing time adjustment on the measured data;

[0055] In A4: when the temporal resolution of the sensor data is inconsistent, the sensor data is interpolated using an interpolation method to match the temporal resolution of the sensor data;

[0056] In A5: transform the data from different sensors into a common global coordinate system;

[0057] In A6: Use the rotation matrix to perform spatial transformation on sensor data, so that data from different sensors correspond to each other in the same coordinate system;

[0058] In A7: Match feature points between different images to obtain the spatial transformation relationship between images;

[0059] In A8: Use optimization to find the best spatial transformation parameters;

[0060] In A9: each sensor is calibrated independently to obtain the internal and external parameters of each sensor, and the internal and external parameters are spatially aligned;

[0061] In A10: the result of the first registration is optimized through iteration until a predetermined progress requirement is met.

[0062] In an optional embodiment, obtaining drone behavior data may include infrared sensors to detect drones at night or in low visibility conditions. The engine and battery of a drone will generate heat. Acoustic sensors may also be used to capture the sound produced by a drone during flight to detect and identify the drone, which is particularly suitable for close-range detection.

[0063] In the embodiment of the present application, data collection is performed through multi-source sensors. The radar sensor is used to transmit radio waves and receive their reflected signals to determine the position and speed of the drone. The radar system can work in various weather conditions and provide high-precision distance and speed information. At the same time, the radar can also detect the slight movement of the drone, thereby updating its flight trajectory in real time;

[0064] Use optical cameras to capture visual images of the drone to determine its position and flight status. Modern camera systems are usually equipped with high-resolution sensors and advanced image processing algorithms, capable of working in both day and night conditions. By analyzing the features in the image, the drone's position and flight direction can be accurately calculated;

[0065] The radio spectrum analyzer is used to monitor the spectrum characteristics of drone communication signals. By analyzing these signals, the type, location and flight mode of the drone can be inferred. At the same time, spectrum analysis can also help detect and identify potential interference sources to ensure the stability and security of drone communications.

[0066] The data collected by multiple source sensors are integrated, and Kalman filtering and particle filtering algorithms are used to fuse the data of different sensors to improve the accuracy of position and speed estimation. By processing this data in real time, a detailed flight trajectory of the drone can be generated and any abnormal behavior can be detected in time.

[0067] It should be noted that by integrating data from sensors such as radar, optical cameras, and radio spectrum analyzers, more comprehensive information on UAV behavior can be provided, including location, speed, flight trajectory, and communication characteristics. Real-time processing of multi-source sensor data can quickly identify abnormal behavior of UAVs, such as illegal intrusions or malicious attacks, so that countermeasures can be taken in a timely manner. The fusion of multi-sensor data can reduce false alarms and missed alarms that may be generated by a single sensor, thereby improving the overall reliability of the system.

[0068] In an optional embodiment, the first registration process may include modality registration, which registers sensor data of different modalities to ensure the consistency of different types of sensor data in terms of features and information; and may also include feature registration, which improves the accuracy of registration by identifying and matching key feature points in images or other data;

[0069] In an optional embodiment, the first registration process may also include intensity registration, adjusting the brightness and contrast of different sensors or images to reduce intensity inconsistencies caused by illumination changes or sensor differences.

[0070] In the embodiment of the present application, the first registration process includes:

[0071] Time registration aligns the data of different sensors on the time axis, completes the correction of timestamps, estimation of time delays and data interpolation, thereby ensuring synchronous analysis. An external clock signal is used to synchronously trigger all sensors to ensure that data is collected at the same time. Each sensor adds a timestamp when collecting data, and the time difference is adjusted through a polynomial fitting algorithm. Time registration is achieved by measuring the time delay between sensors and making corresponding time adjustments to the data. When the time resolution of sensor data is inconsistent, the Lagrange interpolation method is used to interpolate the data to match their time resolution.

[0072] In the embodiment of the present application, the first registration process further includes:

[0073] The spatial alignment principle is to align the data in the spatial coordinate system, complete the coordinate transformation, geometric correction, and image registration, so as to ensure seamless fusion and analysis. Spatial alignment is achieved by converting the coordinate systems of different sensors into the global coordinate system; the sensor data is spatially transformed using a rotation matrix so that the data of different sensors correspond in the same coordinate system; in image processing, the spatial transformation relationship between images is estimated by matching the feature points (such as corner points, edge points, etc.) between different images; the optimization theory is used to find the best spatial transformation parameters to minimize the error between data; each sensor is independently calibrated to obtain its internal and external parameters, and then spatial alignment is performed based on these parameters. Through an iterative approach, the results of temporal and spatial alignment are continuously optimized until the predetermined accuracy requirements are met.

[0074] It should be noted that time registration ensures the time synchronization of data from different sensors by correcting timestamps, estimating time delays and data interpolation, providing a basis for real-time analysis and processing. Spatial registration ensures the precise alignment of data in the spatial coordinate system through coordinate transformation, geometric correction and image registration, providing conditions for seamless fusion and analysis. It can adapt to different environmental conditions and task requirements, and improves the adaptability and flexibility of the system.

[0075] In the embodiment of the present application, the above step S200 includes the following sub-steps B1-B2;

[0076] In B1: the first UAV behavior data is trained by a first machine learning algorithm to obtain a first machine learning model;

[0077] In B2: the first neural network is used to learn the characteristic pattern of the UAV to obtain the first feature.

[0078] In an optional embodiment, the first machine learning algorithm may include using a recurrent neural network or a long short-term memory network to analyze the behavior pattern of the drone and identify abnormal or suspicious activities; it may also include using an integrated learning method such as a random forest or a gradient boosting machine to fuse data from multiple sensors to improve the accuracy of decision making;

[0079] In an optional embodiment, the first machine learning algorithm may also include using a single-class support vector machine or an isolation forest to detect abnormal behavior of the drone;

[0080] In this embodiment of the present application, the first machine learning algorithm includes:

[0081] Deep learning algorithms can automatically learn and extract features from drone images to identify the type, status, and behavior of drones, which is crucial for distinguishing different types of drones and identifying potential threats. Deep learning algorithms analyze the flight trajectory, speed changes, and other dynamic behavior patterns of drones. They can identify normal and abnormal behaviors to predict the intentions of drones, process and fuse data from multiple sensors, and quickly analyze the behavior of drones in real-time monitoring. They are trained to identify normal behavior patterns of drones to detect abnormal behaviors that deviate from these patterns.

[0082] It should be noted that deep learning algorithms excel in automatically extracting complex features from raw data, reducing reliance on manual feature engineering. In contrast, although recurrent neural networks and long short-term memory networks can process sequence data, they may not be as automatic and effective as deep learning algorithms in feature extraction. Deep learning algorithms support end-to-end learning and can directly map from input to output without complex preprocessing or feature design. Ensemble learning methods usually require a combination of multiple models, requiring more parameter adjustments and model selection. Deep learning models, especially those trained with a large amount of data, usually have good generalization capabilities and can handle new, unseen data. Single-class support vector machines and isolation forests are mainly used for anomaly detection and may perform poorly when generalizing to new types of drone behaviors.

[0083] In an optional embodiment, the first neural network may include a convolutional neural network for identifying the drone from an image captured by an optical camera, which has good adaptability to translation invariance in the image. No matter where the drone is in the image, the convolutional neural network can effectively identify it; it may also include a recurrent neural network for analyzing the flight trajectory and behavior pattern of the drone, which has a significant effect on understanding the behavior pattern of the drone and predicting its future movement;

[0084] In an optional embodiment, the first neural network may also include a long short-term memory network for predicting the long-term behavior and trajectory of the drone, and performs well in processing long time series data;

[0085] In an embodiment of the present application, the first neural network includes:

[0086] The structure of the BP neural network includes the number of neurons in the input layer, hidden layer, and output layer. The activation function is selected to introduce nonlinearity so that the network can learn complex patterns. The drone behavior data is forward propagated through each layer of the network until the output layer. In each layer, the neuron calculates the sum of the weighted input and generates output through the activation function. The loss function is used to evaluate the difference between the network output and the drone's behavior category. The loss function provides a quantitative indicator of network performance and is used to guide the adjustment of network weights. The gradient of the loss function with respect to the network weights is calculated through the back-propagation algorithm. Starting from the output layer, the error is back-propagated along each layer of the network until the input layer. The gradient descent algorithm is used to update the network weights according to the calculated gradient.

[0087] It should be noted that compared with other neural networks, the structure of the BP neural network is relatively simple, usually consisting of an input layer, one or more hidden layers and an output layer. The simple structure makes the network easy to understand and implement, especially when the characteristics of the drone behavior data are relatively clear and do not need to process time series data. For drone behavior recognition tasks that do not need to consider time series dependencies, BP neural networks usually require fewer computing resources, and the training and inference speeds may be faster than long short-term memory networks and recurrent neural networks because they do not need to process the dependencies between time steps. The training process of the BP neural network is relatively intuitive, using the standard back-propagation algorithm and gradient descent optimization method. The training process may be easier to converge, especially when the data set is not particularly large or the time series features are not particularly critical. If the drone behavior data can be regarded as static patterns, the BP neural network can effectively identify these patterns without considering time series information.

[0088] In the embodiment of the present application, after completing steps B1-B2 in the above step S200, the following steps B3-B4 are also included;

[0089] In B3: the first feature is input into the first machine learning model to perform real-time identification of the drone;

[0090] In B4: the identified UAV is tracked using the first tracking algorithm to obtain abnormal behavior;

[0091] Specifically, the Feishu information of illegal drones is unknown, and threatening behaviors such as malicious attacks are considered abnormal behaviors;

[0092] In an optional embodiment, the first tracking algorithm may include a Kalman filter, which is an effective recursive filter used for state estimation of linear dynamic systems. In drone tracking, the Kalman filter can be used to predict the future position of the drone and update the prediction based on new observation data; it may also include a particle filter, which represents the probability distribution through a set of random samples (particles) and can be used to process nonlinear motion models of drones;

[0093] In an optional embodiment, the first tracking algorithm may also include a multi-target tracking algorithm. When there are multiple drones in the scene, the MOT algorithm may track multiple targets simultaneously and use data association technology to solve the multi-target tracking problem;

[0094] In the embodiment of the present application, the first tracking algorithm includes:

[0095] Collect drone images for annotation, determine the position of the drone in the image, build a training data set, and use the annotated training data set to train the deep learning model. During the training process, the model learns to identify the features of the drone from the image. The model parameters are adjusted through the back propagation algorithm to minimize the difference between the predicted position and the actual annotated position. The trained model is deployed to the real-time video stream, and each frame of the image is processed to identify and track the drone. The model is used to predict the position of the drone in each frame, and the tracking status of the drone is updated based on the information of consecutive frames.

[0096] It should be noted that through deep learning algorithms and BP neural networks, the system can automatically learn and extract the characteristics of drones, thereby improving the recognition accuracy of drone types, states and behaviors. Deep learning algorithms can analyze the flight trajectory, speed changes and other dynamic behavior patterns of drones, effectively identify normal and abnormal behaviors, and predict the intentions of drones. Real-time recognition and tracking algorithms enable the system to quickly respond to dynamic changes in drones and promptly detect and deal with potential threats.

[0097] In the embodiment of the present application, the above step S300 includes the following sub-step C1:

[0098] In C1: When the UAV flight speed is greater than the first threshold and the distance from the monitoring point is less than the second threshold, the countermeasure instruction is executed.

[0099] In an optional embodiment, the first countermeasure strategy may include deception, sending false GPS signals or communication signals to mislead the drone to fly along a preset wrong path, and inducing the drone to land or return by simulating drone control signals; it may also include directed energy weapons, using lasers or high-energy microwave weapons to directly destroy the drone's electronic equipment or structure;

[0100] In an optional embodiment, the first countermeasure strategy may also include a network attack, invading the control system of the drone through hacking techniques, gaining control or rendering it ineffective, and utilizing software vulnerabilities or backdoor programs to interfere with or control the drone.

[0101] In the embodiment of the present application, the construction of the first countermeasure strategy includes:

[0102] jamming, interception, and capture strategies;

[0103] Interference is to destroy the attacker's communication link or computing resources by sending interference signals or data packets, making it impossible for the attacker to effectively execute the attack;

[0104] Interception: Prevent the spread and implementation of attacks through technical means before or in the early stages of an attack.

[0105] Capture: After an attack occurs, the attacker or his key resources can be captured through tracking and analysis to provide support for subsequent legal actions or technical counterattacks.

[0106] Specifically, generate specific countermeasure instructions based on the identification results and countermeasure strategies, clearly identify the countermeasure targets, including specific IP addresses, domain names or physical locations, select the most appropriate countermeasures based on the nature and severity of the threat, and ensure that there are sufficient resources and technical support to execute the countermeasure instructions, including manpower, equipment and funds;

[0107] Specifically, through expulsion countermeasures, countermeasures are initiated when the drone flies too fast or is too close, that is, the first threshold is 10m / s and the second threshold is 100m. When the flight speed is greater than 10m / s and the distance to the monitoring point is less than 100m, countermeasures are initiated to maximize the effect and minimize potential risks.

[0108] It should be noted that interference, interception and capture strategies usually do not involve physical destruction of drones, which helps to counter without damaging drones, so that drones can be retained as evidence or for further analysis when necessary. In contrast, directed energy weapons can cause permanent damage to drones, which may not only lead to loss of evidence but also cause more serious security issues. Interference strategies can interrupt the communication link of the drone by sending specific signals. This strategy is reversible and the interference can be stopped immediately once the threat is lifted. Deception and cyber attacks can cause the drone control system to be permanently tampered with or damaged, which is not the best option in some cases.

[0109] It should also be noted that the combination of jamming, interception and capture strategies has formed a multi-level, all-round defense system that can effectively respond to drone threats at different stages and of different types. Through precise identification and positioning technology, it ensures that countermeasures are targeted at specific threats, reduces the impact on non-target objects, and reduces the risk of accidental injuries.

[0110] The above is a schematic scheme of a drone countermeasure method of this embodiment. It should be noted that the technical scheme of the drone countermeasure system and the technical scheme of the drone countermeasure method described above are of the same concept, and the details of the technical scheme of the drone countermeasure system in this embodiment that are not described in detail can all be referred to the description of the technical scheme of the drone countermeasure method described above.

[0111] Example 2

[0112] This embodiment is different from the first embodiment and provides an application embodiment of a drone countermeasure method to verify and illustrate the technical effects used in this method.

[0113] Take the environment range within a radius of 1000m as an example;

[0114] Multi-source sensors, including radar, optical cameras, and radio spectrum analyzers, are deployed around the environment to collect the drone's position, speed, and flight trajectory data in real time.

[0115] The collected data is processed using a spatiotemporal registration algorithm to ensure that the data from different sensors are aligned in time and space; an external clock signal is used to synchronize all sensors to ensure that the data is collected at the same time. The time difference is adjusted using a polynomial fitting algorithm to achieve timestamp correction; the data from different sensors is converted to a common global coordinate system, and the sensor data is spatially transformed using a rotation matrix to ensure accurate alignment of the data in the spatial coordinate system.

[0116] Collect a large amount of drone flight data, including normal and abnormal behaviors, and train the BP neural network model to enable it to identify and predict the behavior patterns of drones. Once an illegal drone is detected, countermeasures are immediately initiated. The most appropriate countermeasure strategy is selected based on the drone's behavior and threat level.

[0117] When an illegal drone is identified approaching the environment, the jamming strategy is activated to send jamming signals to destroy the GPS signal and communication link of the drone, making it unable to perform the intended mission;

[0118] When the jamming strategy fails to deflect the drone from its route, the interception strategy is initiated, using physical means such as drone interceptors or net guns to directly intercept illegal drones;

[0119] For drones that have been interfered with or intercepted, a capture strategy is initiated to capture the drone or its key resources through tracking and analysis.

[0120] Through accurate drone identification and tracking, illegal drones can be discovered and dealt with in a timely manner, thereby protecting environmental safety and avoiding possible mid-air collisions or malicious surveillance. Illegal drones are used to infringe on personal privacy, and can effectively prevent the intrusion of these drones. For public places, the privacy of tourists can be protected. The high-precision positioning capability of the spatiotemporal registration algorithm reduces accidental injuries to non-target objects and ensures the accuracy and safety of countermeasures. It can process and analyze drone signals in real time, quickly respond to the dynamic changes of drones, and effectively curb potential threats. It can adapt to different environmental conditions, such as different weather and lighting conditions, which improves the versatility of countermeasures and adaptability to complex environments.

[0121] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

[0122] Example 3

[0123] The drone countermeasure system in this embodiment includes:

[0124] A registration processing module, used for acquiring UAV behavior data, performing a first registration processing on the UAV behavior data, and obtaining first UAV behavior data;

[0125] an identification module, configured to identify the first UAV behavior data through a first neural network to obtain an identification result;

[0126] The execution instruction module is used to generate a counter-instruction by combining the recognition result with the first counter-measure strategy, and determine the timing of the counter-measure action to execute the counter-instruction.

[0127] This embodiment also provides a computing device, which is applicable to the countermeasure of drones, including:

[0128] Memory and processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the drone countermeasure method proposed in the above embodiment.

[0129] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for countering a drone as proposed in the above embodiment is implemented.

[0130] The storage medium proposed in this embodiment and the method for implementing drone countermeasures proposed in the above embodiment belong to the same inventive concept. The technical details not fully described in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0131] Through the above description of the implementation methods, the technicians in the relevant field can clearly understand that the present invention can be implemented by means of software and necessary general hardware, and of course can also be implemented by hardware. Based on such an understanding, the technical solution of the present invention 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 a computer floppy disk, read-only memory (ReadOnly, Memory, ROM), random access memory (Random Access Memory, RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform the methods of various embodiments of the present invention.

Claims

1. A method for countering a drone, characterized in that: include: Acquire UAV behavior data, and perform a first registration process on the UAV behavior data to obtain first UAV behavior data; Identify the first UAV behavior data using a first neural network to obtain an identification result; The identification result is combined with the first countermeasure strategy to generate a countermeasure instruction, and the timing of the countermeasure action is determined to execute the countermeasure instruction.

2. The method for countering a drone as claimed in claim 1, characterized in that: Identifying the first drone behavior data includes: Training the first drone behavior data using a first machine learning algorithm to obtain a first machine learning model; The first neural network is used to learn the characteristic pattern of the UAV to obtain the first feature.

3. The drone countermeasure method according to claim 2, characterized in that: Also includes: Inputting the first feature into a first machine learning model to perform real-time identification of the drone; The identified UAV is tracked using the first tracking algorithm to obtain abnormal behavior.

4. The drone countermeasure method according to claim 1 or 3, characterized in that: Determining when to take countermeasures includes: When the UAV's flight speed is greater than the first threshold and the distance from the monitoring point is less than the second threshold, the countermeasure command is executed.

5. The method for countering a drone as claimed in claim 4, characterized in that: Performing a first registration process on the drone behavior data includes: Use an external clock signal to synchronize the triggering of all sensors; Each sensor adds a timestamp when collecting data, and adjusts the time difference through multiple fitting algorithms; Measure the time delay between sensors and make time adjustments to the measured data; When the temporal resolution of sensor data is inconsistent, the sensor data is interpolated using an interpolation method to match the temporal resolution of the sensor data.

6. The method for countering a drone as claimed in claim 5, characterized in that: Also includes: Transform data from different sensors into a common global coordinate system; Use the rotation matrix to spatially transform the sensor data and correspond the data of different sensors in the same coordinate system; Match the feature points between different images to obtain the spatial transformation relationship between images; Use optimization to find the best spatial transformation parameters; Each sensor is calibrated independently to obtain the internal and external parameters of each sensor, and the internal and external parameters are spatially aligned; The result of the first registration is optimized through iteration until the predetermined progress requirement is met.

7. The method for countering a drone as claimed in claim 6, characterized in that: The first countermeasure strategy includes: Interference strategies, interception strategies, and capture strategies.

8. A system using the drone countermeasure method according to any one of claims 1 to 7, characterized in that: include: A registration processing module, used for acquiring UAV behavior data, performing a first registration processing on the UAV behavior data, and obtaining first UAV behavior data; an identification module, configured to identify the first UAV behavior data through a first neural network to obtain an identification result; The execution instruction module is used to generate a counter-instruction by combining the recognition result with the first counter-measure strategy, and determine the timing of the counter-measure action to execute the counter-instruction.

9. An electronic device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the drone countermeasure method described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the drone countermeasure method described in any one of claims 1 to 7.

Images