UAV detection method, device and system integrating multimodal data

Abstract

The present invention relates to the field of drone positioning technology, and in particular to a drone detection method, device and system that integrates multimodal data. The method obtains radar signals and sound wave signals of a microphone in an overhead power transmission line area; obtains the target speed of a suspected drone and the target distance between it and the radar based on the radar signal; obtains the target time point when the suspected drone is detected by the microphone based on the target distance and the time point corresponding to the suspected drone; obtains the reference speed of the suspected drone and the reference distance between it and the radar based on the spectrum diagram of the window corresponding to the sound wave signal at the target time point and the spectrum diagram of the preset neighborhood window; detects whether the suspected drone is real based on the difference between the target speed and the reference speed, and the difference between the target distance and the reference distance. The present invention integrates radar signals and sound wave signals to detect drones, improves the detection accuracy of drones, and accurately detects drones in the overhead power transmission line area.

Description

Technical Field

[0001] The present invention relates to the field of drone positioning technology, and in particular to a drone detection method, device and system integrating multimodal data. Background Art

[0002] With the rapid development of drone technology and the continuous expansion of its application fields, drones are playing an increasingly important role in multiple industries such as agriculture, logistics, and monitoring. However, the popularity of drones has also brought a series of security challenges, especially in the protection of overhead power transmission lines. Due to the easy accessibility of drones and the concealed nature of their operation, illegal intrusions and collisions with overhead power transmission lines have occurred frequently, resulting in large-scale power outages, casualties, and huge economic losses. Therefore, it is particularly urgent to accurately detect drones around overhead power transmission lines, and then interfere with and counter the accurately detected drones to ensure power grid security and public safety.

[0003] Existing drone detection technologies mainly include radar and sound waves. Among them, radar detection relies on electromagnetic waves to detect drones. The radar system can transmit wireless signals. When the wireless signal interacts with the drone, part of the signal will be reflected back by the drone. The radar system will receive and identify the reflected signal, and then calculate the drone's position, speed, distance and other related information. Radar can achieve remote detection of drones. At the same time, there is less interference from weather and natural environment, so high-precision drone position and speed information can be provided. Sound wave detection relies on sound wave signals to detect drones. When the motor and rotor of the drone rotate, a certain amount of noise will be generated. The frequency distribution range is usually 0.2-20kHz. By analyzing the sound wave characteristics emitted by the drone and using audio recognition and sound wave array receiving technology, technicians can accurately determine the type and location of the drone.

[0004] When using radar for drone detection in actual situations, since some drones are now made of light materials, the high wave transmittance of light materials and the small radar scattering area will lead to inaccurate radar detection results, especially for small, low-speed and low-altitude flying drones, which may not be detected, and then the radar cannot accurately identify drones around overhead power transmission lines. During the operation of overhead power transmission lines, there is a corona discharge phenomenon, which generates high-frequency noise. When using sound waves to detect drones, the high-frequency noise of the corona discharge phenomenon will interfere with the noise detection of drones, resulting in the inability to accurately detect drones around overhead power transmission lines through sound waves. Summary of the invention

[0005] In order to solve the technical problem that the existing methods cannot accurately detect drones, the purpose of the present invention is to provide a drone detection method, device and system that integrates multimodal data. The technical solutions adopted are as follows:

[0006] In a first aspect, an embodiment of the present invention provides a method for detecting a drone by fusing multimodal data, the method comprising the following steps:

[0007] Acquire radar signals and acoustic wave signals of each microphone in the area of ​​overhead power transmission lines;

[0008] Based on the radar signal, a suspected drone signal segment is obtained, and according to the transmission signal and the reception signal in each suspected drone signal segment, the target speed of the suspected drone corresponding to each suspected drone signal segment and the target distance between the suspected drone and the radar are obtained;

[0009] Based on the target distance and the time point corresponding to each suspected drone signal segment, the target time point at which the suspected drone corresponding to each suspected drone signal segment is detected by the microphone is obtained; based on the spectrum diagram of the window corresponding to the sound wave signal of each microphone at the target time point and the spectrum diagram of the preset neighborhood window, the reference speed of the suspected drone and the reference distance between the suspected drone and the radar corresponding to each suspected drone signal segment obtained by the microphone at the target time point are obtained;

[0010] Based on the difference between the target speed and the reference speed of the suspected drone corresponding to each suspected drone signal segment, and the difference between the target distance and the reference distance between the suspected drone and the radar corresponding to each suspected drone signal segment, it is detected whether the suspected drone corresponding to each suspected drone signal segment is real.

[0011] Furthermore, the method for obtaining the suspected drone signal segment is:

[0012] Divide the radar signal according to the time of transmitting pulses to obtain radar signal segments;

[0013] The power of each radar signal segment is obtained through the radar equation as the target power of each radar signal segment;

[0014] The minimum detectable power of the radar signal is obtained. When the target power is greater than the minimum detectable power, the corresponding radar signal segment is regarded as a suspected drone signal segment.

[0015] Furthermore, the method for obtaining the target speed of the suspected drone corresponding to each suspected drone signal segment and the target distance between the suspected drone and the radar is as follows:

[0016] For any suspected drone signal segment, according to the frequency of the transmitted signal and the received signal in the suspected drone signal segment, and the propagation speed of the radar signal in the atmosphere, through the Doppler effect, according to the formula , obtain the target speed of the suspected drone corresponding to the suspected drone signal segment; where, is the frequency of the received signal in the xth suspected drone signal segment; is the frequency of the transmitted signal in the xth suspected drone signal segment; c is the propagation speed of the radar signal in the atmosphere; is the target speed of the suspected drone corresponding to the xth suspected drone signal segment;

[0017] Obtaining a time difference between a transmitted signal and a received signal in the suspected drone signal segment as a first time difference;

[0018] Half of the result of multiplying the first time difference by the propagation speed of the radar signal in the atmosphere is taken as the target distance between the suspected drone and the radar corresponding to the suspected drone signal segment.

[0019] Furthermore, the method for obtaining the target time point is:

[0020] For any suspected drone signal segment, the target distance between the suspected drone and the radar corresponding to the suspected drone signal segment is taken as the first distance;

[0021] taking the ratio of the first distance to the propagation speed of the radar signal in the atmosphere as the first time period;

[0022] The ratio of the first distance to the propagation speed of the sound wave in the atmosphere is used as the second time period;

[0023] The absolute value of the difference between the first time period and the second time period is used as the reference time difference when the suspected drone corresponding to the suspected drone signal segment is detected by the radar and the microphone;

[0024] The time at the specified position in the suspected drone signal segment is used as the time point of the suspected drone signal segment;

[0025] The sum of the time point of the suspected drone signal segment and the reference time difference is used as the target time point when the suspected drone corresponding to the suspected drone signal segment is detected by the microphone.

[0026] Furthermore, the method for obtaining the reference speed of the suspected drone and the reference distance between the suspected drone and the radar corresponding to each suspected drone signal segment is as follows:

[0027] For any suspected drone signal segment, the target time point at which the suspected drone corresponding to the suspected drone signal segment is detected by the microphone is used as the analysis time point;

[0028] The sound wave signal of each microphone is divided into preset windows, and the spectrum of each window is obtained by Fourier transform;

[0029] For any microphone, the window corresponding to the sound wave signal of the microphone at the analysis time point is used as the analysis window;

[0030] The average spectrum of the spectrum of the analysis window and its preset neighboring window is used as the overall spectrum of the analysis window;

[0031] For any frequency in the overall spectrum, the result of negative correlation and normalization between the absolute value of the difference between the frequency and the drone's designated sound wave frequency is used as the drone sound wave weight of the frequency;

[0032] The power corresponding to the frequency in the overall spectrum is negatively correlated and normalized, and the result is used as the noise suppression weight of the frequency;

[0033] Obtain the product of the drone sound wave weight and noise suppression weight of the frequency and the power corresponding to the frequency in the spectrum diagram of the analysis window as the corrected power corresponding to the frequency in the spectrum diagram of the analysis window;

[0034] Obtain the corrected power corresponding to each frequency in the spectrum diagram of the analysis window, and use the maximum corrected power as the representative power of the microphone at the analysis time point; use the frequency corresponding to the representative power as the representative frequency of the microphone at the analysis time point;

[0035] According to the characteristic power of each microphone at the analysis time point and the spherical attenuation model of its sound wave intensity in the atmosphere, the reference distance between the suspected drone and the radar corresponding to the suspected drone signal segment obtained by the microphone at the analysis time point is obtained;

[0036] According to the characteristic frequency of each microphone at the analysis time point, the propagation speed of sound waves in the atmosphere and the sound wave frequency of the drone flight, the reference speed of the suspected drone corresponding to the suspected drone signal segment obtained by the microphone at the analysis time point is obtained.

[0037] Furthermore, the reference distance is obtained by:

[0038] For the w-th microphone, the spherical attenuation model of the sound wave intensity of the w-th microphone in the atmosphere is: ; The reference function corresponding to the w-th microphone at the analysis time point is: ; In the formula, is the attenuation of the sound wave signal intensity for the wth microphone; is the distance between the suspected drone corresponding to the suspected drone signal segment and the wth microphone; is the ratio of pi; is a logarithmic function with base 2; is the characteristic power of the w-th microphone at the analysis time point; is the sound wave power of the drone;

[0039] The spherical attenuation model of the sound wave intensity of each microphone in the atmosphere and the reference function corresponding to each microphone at the analysis time point are combined, and the optimal solution is obtained through the optimization algorithm. The distance in the optimal solution at this time is used as the analysis distance between the suspected drone and the microphone corresponding to the suspected drone signal segment obtained at the analysis time point;

[0040] Taking the position of each microphone as the center point, make the spherical equation of each microphone in the spatial coordinate system with the analysis distance as the radius;

[0041] The spherical equations of each microphone are solved simultaneously, and the distance between the solved coordinate value and the radar coordinate value is used as the reference distance between the suspected drone and the radar corresponding to the suspected drone signal segment obtained by the microphone at the analysis time point.

[0042] Furthermore, the reference speed is obtained by:

[0043] For the w-th microphone, the objective function corresponding to the w-th microphone at the analysis time point is obtained through the Doppler effect: ; In the formula, is the characteristic frequency of the w-th microphone at the analysis time point; is the sound wave frequency of the UAV flight; s is the propagation speed of the sound wave in the atmosphere; is the speed of the suspected drone corresponding to the suspected drone signal segment obtained by the wth microphone at the analysis time point;

[0044] The objective functions corresponding to each microphone at the analysis time point are combined, and the optimal solution is obtained through the optimization algorithm. The speed in the optimal solution at this time is used as the reference speed of the suspected drone corresponding to the suspected drone signal segment obtained by the microphone at the analysis time point.

[0045] Furthermore, the method for detecting whether the suspected drone corresponding to each suspected drone signal segment is real is:

[0046] For any suspected drone signal segment, the calculation formula for obtaining the authenticity of the suspected drone corresponding to the suspected drone signal segment is: ; In the formula, is the authenticity of the suspected drone corresponding to the xth suspected drone signal segment; is the target distance between the suspected UAV and the radar corresponding to the xth suspected UAV signal segment; is the reference distance between the suspected UAV and the radar corresponding to the xth suspected UAV signal segment; is the target speed of the suspected drone corresponding to the xth suspected drone signal segment; is the reference speed of the suspected drone corresponding to the xth suspected drone signal segment; is the absolute value function; exp is the exponential function with a natural constant as the base;

[0047] When the authenticity is greater than the preset authenticity threshold, the suspected drone corresponding to the suspected drone signal segment is authentic;

[0048] When the authenticity is less than or equal to the preset authenticity threshold, the suspected drone corresponding to the suspected drone signal segment is not authentic.

[0049] In a second aspect, another embodiment of the present invention provides a drone detection device that integrates multimodal data, the device comprising: a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, the steps of any one of the above methods are implemented.

[0050] In a third aspect, another embodiment of the present invention provides a drone detection system integrating multimodal data, the system comprising:

[0051] An acquisition module, used for acquiring radar signals and acoustic wave signals of each microphone in the overhead power transmission line area;

[0052] A radar signal processing module is used to obtain a suspected drone signal segment based on the radar signal, and obtain the target speed of the suspected drone corresponding to each suspected drone signal segment and the target distance between the suspected drone and the radar according to the transmission signal and the reception signal in each suspected drone signal segment;

[0053] The acoustic signal processing module is used to obtain the target time point at which the suspected drone corresponding to each suspected drone signal segment is detected by the microphone based on the target distance and the time point corresponding to each suspected drone signal segment; based on the spectrum diagram of the window corresponding to the acoustic signal of each microphone at the target time point and the spectrum diagram of the preset neighborhood window, obtain the reference speed of the suspected drone and the reference distance between the suspected drone and the radar corresponding to each suspected drone signal segment obtained by the microphone at the target time point;

[0054] The detection module is used to detect whether the suspected drone corresponding to each suspected drone signal segment is real based on the difference between the target speed of the suspected drone corresponding to each suspected drone signal segment and the reference speed, and the difference between the target distance and the reference distance between the suspected drone corresponding to each suspected drone signal segment and the radar.

[0055] The present invention has the following beneficial effects:

[0056] The present invention obtains suspected drone signal segments based on radar signals, and preliminarily determines the drones that may exist in the overhead power line area; in order to determine the position and speed of the drone, timely countermeasures are taken against the drones in the overhead power line area to avoid damage to the overhead power line, and then the target speed of the suspected drone corresponding to each suspected drone signal segment and the target distance between it and the radar are obtained according to the transmitted signal and the received signal in each suspected drone signal segment, and the speed of each suspected drone obtained through the radar signal and the distance between it and the radar are determined; considering that the detection effect of radar signals on drones made of modern lightweight materials is inaccurate, and then based on the spectrum diagram of the window corresponding to the sound wave signal of each microphone at the target time point and the spectrum diagram of the preset neighborhood window, the reference speed of the suspected drone corresponding to each suspected drone signal segment obtained by the microphone at the target time point and the reference speed of the suspected drone and the distance between it and the radar are obtained. The reference distance between the two is determined to determine the speed of each suspected drone obtained by the sound wave signal and the distance between it and the radar; in order to accurately analyze whether the suspected drone corresponding to each suspected drone signal segment is real, accurately detect the drones in the overhead transmission line area, and then according to the difference between the target speed of the suspected drone corresponding to each suspected drone signal segment and the reference speed, and the difference between the target distance and the reference distance between the suspected drone corresponding to each suspected drone signal segment and the radar, accurately analyze the possibility that the suspected drone corresponding to each suspected drone signal segment is real, and then accurately detect whether the suspected drone corresponding to each suspected drone signal segment is real, and accurately detect the drones in the overhead transmission line area, which is conducive to accurate and efficient counter-interference of drones in the overhead transmission line area, avoids damage to overhead transmission lines by drones, and improves the safety of overhead transmission lines. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0058] Figure 1 A schematic flow chart of a drone detection method integrating multi-modal data provided by one embodiment of the present invention;

[0059] Figure 2 A structural diagram of a drone detection system integrating multimodal data provided by an embodiment of the present invention;

[0060] Figure 3A schematic diagram of a computer device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0061] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the drone detection method, device and system for integrating multimodal data proposed by the present invention, its specific implementation, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0062] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0063] The specific scheme of the drone detection method, device and system integrating multimodal data provided by the present invention is described in detail below with reference to the accompanying drawings.

[0064] Embodiment 1:

[0065] The specific scenario of this embodiment is: when using a radar system to detect drones, since modern drones are often made of lightweight materials, the high wave transmittance and small radar scattering area of ​​lightweight materials will affect the radar detection results, especially for the detection of small, low-speed and low-altitude drones, the radar will have a poor recognition effect on such drones due to the weak performance of radar reflection signals. Considering that drones have specific sound wave frequencies when flying, this embodiment integrates the sound wave signals and radar signals of drones to detect drones, improve the detection accuracy of drones, accurately detect drones in the overhead power line area, and then accurately counter-interfere with drones, thereby improving the safety of the overhead power line area.

[0066] This paper proposes a drone detection method that integrates multimodal data. Figure 1 , which shows a schematic flow chart of a drone detection method integrating multimodal data provided by an embodiment of the present invention, the method comprising the following steps:

[0067] Step S1: Acquire the radar signal and the sound wave signal of each microphone in the overhead transmission line area.

[0068] Specifically, in order to reduce the damage caused by drones to overhead power transmission lines, it is necessary to accurately detect drones close to overhead power transmission lines. When drones appear in the overhead power transmission line area, it is necessary to counter-interfere with the drones to eliminate the harmful effects of drones on overhead power transmission lines. The overhead power transmission line area can be set according to actual conditions and is not limited here. In order to accurately identify drones in the overhead power transmission line area, this embodiment sets a radar module and a sound wave detection module in the overhead power transmission line area. The radar in the radar module covers the entire overhead power transmission line area, and the microphone in the sound wave detection module detects sound waves in the entire overhead power transmission line area.

[0069] The radar module consists of a transmitter and a receiver, which are installed around the overhead power line to cover the entire overhead power line area. The transmitter is responsible for transmitting the radar signal, and the receiver is responsible for receiving the reflected radar signal. In order to improve the accuracy and coverage of detection, the radar module can be installed at a higher position, such as on the tower of the overhead power line or on the ground at a certain angle upward to monitor low-flying drones. The radar module transmits the radar signal through the transmitter. When the transmitted radar signal encounters an object, it will be reflected back and captured by the receiver. The radar module converts the received radar signal into an electrical signal and sends it to the processing unit for analysis. Since the radar has multiple transmitters in different directions, the different transmission directions can be distinguished by the received signal, so as to locate the object. It should be noted that the subsequent radar signals are all converted electrical signals. Continuously obtain radar signals in the overhead power line area.

[0070] The sound wave detection module usually includes one or more microphone arrays for capturing sound wave signals in the overhead power transmission line area, wherein a microphone array includes multiple microphones. In this embodiment, a microphone array is set to include 10 microphones, and the distance between two adjacent microphone installation positions is 5 meters. The implementer can set the number of microphones contained in a microphone array and the distance between two adjacent microphone installation positions according to actual conditions, which are not limited here. Among them, the microphone needs to be installed around the overhead power transmission line to ensure coverage of the entire overhead power transmission line area. At the same time, the installation position of the microphone should be selected in a place without a lot of background noise interference, and the microphone array should be installed perpendicular to the ground to capture sound wave signals from the air. Each microphone in the sound wave detection module will continuously monitor the sound in the overhead power transmission line area and convert it into an electrical signal. It should be noted that the subsequent sound wave signals are all converted electrical signals. The sound wave signals of each microphone in the overhead power transmission line area are continuously obtained.

[0071] Step S2: Obtain suspected drone signal segments based on radar signals, and obtain the target speed of the suspected drone and the target distance between the suspected drone and the radar corresponding to each suspected drone signal segment according to the transmitted signal and received signal in each suspected drone signal segment.

[0072] In actual situations, radar detects objects by identifying them through the power of reflected waves. When the power of reflected waves exceeds a certain threshold, it is considered that an object has been detected in the corresponding direction. Therefore, this embodiment first obtains a suspected drone signal segment based on the radar signal, where one suspected drone signal segment corresponds to one suspected drone. In order to determine the position and speed of each suspected drone, so as to accurately analyze whether each suspected drone is real, the target speed of the suspected drone corresponding to each suspected drone signal segment and the target distance between it and the radar are obtained based on the transmitted signal and received signal in each suspected drone signal segment.

[0073] Preferably, in a method that can be implemented in this embodiment, the method for obtaining the suspected drone signal segment is: first, the radar signal is divided according to the time of the transmitted pulse to obtain the radar signal segment; that is, the radar system will set one or more time windows according to the time of the transmitted pulse, and wait for the reception of the transmitted signal in these windows. Each time window corresponds to a specific distance range. Since the propagation speed of electromagnetic waves is known, the time when the signal may return can be determined according to the time window. Therefore, it can be inferred that a radar signal segment in this embodiment only includes one signal transmission and one signal reception. It is known that the radar detection object is identified by the power of the reflected wave. Therefore, this embodiment obtains the power of each radar signal segment through the radar equation as the target power of each radar signal segment; at the same time, the minimum detectable power of the radar signal is obtained. It is known that the radar signal segment with a power greater than the minimum detectable power is more likely to be a radar signal segment that detects the drone, and then when the target power is greater than the minimum detectable power, the corresponding radar signal segment is used as a suspected drone signal segment. Among them, the radar equation and the method for obtaining the minimum detectable power are both well-known technologies and will not be repeated.

[0074] Preferably, in an achievable manner of this embodiment, the method for calculating the target speed of the suspected drone corresponding to each suspected drone signal segment and the target distance between the suspected drone and the radar is as follows: for any suspected drone signal segment, according to the frequency of the transmitted signal and the frequency of the received signal in the suspected drone signal segment, and the propagation speed of the radar signal in the atmosphere, through the Doppler effect, according to the formula , obtain the target speed of the suspected drone corresponding to the suspected drone signal segment; where, is the frequency of the received signal in the xth suspected drone signal segment; is the frequency of the transmitted signal in the xth suspected drone signal segment; c is the propagation speed of the radar signal in the atmosphere; is the target speed of the suspected drone corresponding to the xth suspected drone signal segment; the propagation speed of the radar signal in the atmosphere is the speed of light; the frequency of the received signal and the frequency of the transmitted signal in the xth suspected drone signal segment, as well as the propagation speed of the radar signal in the atmosphere are all known, therefore, through the formula The target speed of the suspected drone corresponding to the xth suspected drone signal segment can be obtained Among them, the formula It is obtained through the Doppler effect, which is a well-known technology and will not be described in detail;

[0075] The time difference between the transmitted signal and the received signal in the suspected drone signal segment is obtained as the first time difference; half of the result of multiplying the first time difference by the propagation speed of the radar signal in the atmosphere is used as the target distance between the suspected drone and the radar corresponding to the suspected drone signal segment. It should be noted that the method for obtaining the target distance between the suspected drone and the radar corresponding to the suspected drone signal segment is also based on the Doppler effect.

[0076] At this point, the target speed of the suspected drone corresponding to each suspected drone signal segment and the target distance between it and the radar are obtained.

[0077] Step S3: Based on the target distance and the time point corresponding to each suspected drone signal segment, obtain the target time point at which the suspected drone corresponding to each suspected drone signal segment is detected by the microphone; based on the spectrum diagram of the window corresponding to the sound wave signal of each microphone at the target time point and the spectrum diagram of the preset neighborhood window, obtain the reference speed of the suspected drone and the reference distance between the suspected drone and the radar corresponding to each suspected drone signal segment obtained by the microphone at the target time point.

[0078] It is known that when a drone is flying, due to the interaction between its motor and rotor and the air, the drone will generate a sound wave signal of a specific frequency. The sound wave signal of a specific frequency can be used to detect the drone. Therefore, for each suspected drone signal segment corresponding to the suspected drone, the sound wave signal obtained by the microphone can be used to further analyze whether the suspected drone corresponding to each suspected drone signal segment is real. Considering that the propagation speeds of radar signals and sound wave signals in the atmosphere are different, the time when the same suspected drone is detected by the radar signal and the sound wave signal is not the same. In order to determine the corresponding time when the suspected drone corresponding to each suspected drone signal segment is detected by the sound wave signal, this embodiment obtains the target time point when the suspected drone corresponding to each suspected drone signal segment is detected by the microphone based on the target distance and the time point corresponding to each suspected drone signal segment.

[0079] Select any suspected drone signal segment, and take the target time point when the suspected drone corresponding to the suspected drone signal segment is detected by the microphone as the reference time point, and then analyze the spectrum of the sound wave signal of each microphone in the window corresponding to the reference time point to determine the speed of the suspected drone corresponding to the suspected drone signal segment detected by the sound wave signal, and the distance between the suspected drone corresponding to the suspected drone signal segment and the radar, and then compare with the target speed and target distance of the suspected drone corresponding to the suspected drone signal segment obtained by the radar signal to determine whether the suspected drone corresponding to the suspected drone signal segment is real. Among them, the method for obtaining the spectrum is a well-known technology and will not be repeated. In this embodiment, the horizontal axis of the spectrum is frequency and the vertical axis is power. In actual situations, the sound wave signal detected by the microphone will be affected by the noise interference of the corona discharge phenomenon in the overhead transmission line itself, resulting in errors in the obtained spectrum diagram. In order to reduce the error influence of the sound wave signal caused by the noise interference of the corona discharge phenomenon, this embodiment is based on the spectrum diagram of the window corresponding to the sound wave signal of each microphone at the target time point and the spectrum diagram of the preset neighborhood window, to obtain the reference speed of the suspected drone and the reference distance between it and the radar corresponding to each suspected drone signal segment obtained by the microphone at the target time point.

[0080] Preferably, in a method that can be implemented in this embodiment, the method for obtaining the target time point is: for any suspected drone signal segment, the target distance between the suspected drone and the radar corresponding to the suspected drone signal segment is taken as the first distance; the ratio of the first distance to the propagation speed of the radar signal in the atmosphere is taken as the first time period; the ratio of the first distance to the propagation speed of the sound wave in the atmosphere is taken as the second time period; wherein the propagation speed of the sound wave in the atmosphere is the speed of sound. The absolute value of the difference between the first time period and the second time period is taken as the reference time difference when the suspected drone corresponding to the suspected drone signal segment is detected by the radar and the microphone; the moment at the specified position in the suspected drone signal segment is taken as the time point of the suspected drone signal segment; in this embodiment, the specified position is set as the starting position of the suspected drone signal segment, and the implementer can set the specified position according to the actual situation, which is not limited here. The addition result of the time point of the suspected drone signal segment and the reference time difference is taken as the target time point when the suspected drone corresponding to the suspected drone signal segment is detected by the microphone.

[0081] At this point, the target time point at which the suspected drone is detected by the microphone corresponding to each suspected drone signal segment is obtained.

[0082] Preferably, in a method that can be implemented in this embodiment, the method for obtaining the reference speed of the suspected drone corresponding to each suspected drone signal segment and the reference distance between it and the radar is as follows: for any suspected drone signal segment, the target time point when the suspected drone corresponding to the suspected drone signal segment is detected by the microphone is used as the analysis time point; the sound wave signal of each microphone is divided according to a preset window, and the frequency spectrum of each window is obtained by Fourier transform; this embodiment sets the duration of the preset window to 10ms, and the implementer can set the duration of the preset window according to actual conditions, which is not limited here. Among them, Fourier transform is a well-known technology and will not be described in detail. It should be noted that the number of windows corresponding to the sound wave signal of each microphone is the same, and the time period corresponding to the window at the same position of each microphone is the same;

[0083] For any microphone, the window corresponding to the acoustic wave signal of the microphone at the analysis time point is used as the analysis window, wherein the analysis time point must be located in the analysis window. In actual situations, the corona discharge phenomenon of overhead transmission lines occurs randomly, that is, the noise generated by overhead transmission lines is random noise. Therefore, this embodiment can reduce the random performance of noise in the spectrum diagram by averaging the spectrum diagram over a long time series. Furthermore, this embodiment uses the 30,000 windows closest to the analysis window before and after as preset neighborhood windows of the analysis window. Among them, the analysis window and its preset neighborhood windows are windows corresponding to the acoustic wave signal of the microphone. The implementer can set the number of preset neighborhood windows of the analysis window according to actual conditions, which is not limited here. It should be noted that if the number of windows before or after the analysis window is less than 30,000, only the existing windows are used as preset neighborhood windows of the analysis window; then the average spectrum diagram of the spectrum diagram of the analysis window and its preset neighborhood window is obtained as the overall spectrum diagram of the analysis window, which effectively reduces the noise interference of the overhead transmission line on the acoustic wave signal in the analysis window;

[0084] According to research, the sound wave frequency of drones in flight is mainly concentrated around 10,000 Hz. Therefore, this embodiment sets 10,000 Hz as the drone's designated sound wave frequency. The closer a frequency in the overall spectrum graph is to 10,000 kHz, the more likely it is that the frequency corresponds to a drone. If a drone exists, the power corresponding to the frequency in the spectrum graph of the analysis window is larger and more accurate. For any frequency in the overall spectrum graph, the absolute value of the difference between the frequency and the drone's designated sound wave frequency is negatively correlated and normalized as the drone's sound wave weight for that frequency. The larger the drone's sound wave weight, the more accurate the power corresponding to the frequency in the spectrum graph of the analysis window. The calculation formula for the drone's sound wave weight is: ; In the formula, is the drone sound wave weight of the ath frequency in the overall spectrum; is the ath frequency in the overall spectrum; is the absolute value function; e is a natural constant;

[0085] The greater the power corresponding to the frequency in the overall spectrum diagram, the stronger the noise performance of the frequency in the sound wave signal of the analysis window, that is, the stronger the distribution of the noise generated by the overhead transmission line at the frequency, the less accurate the power corresponding to the frequency in the spectrum diagram of the analysis window, and the more it needs to be modified. Therefore, this embodiment negatively correlates and normalizes the power corresponding to the frequency in the overall spectrum diagram as the noise suppression weight of the frequency; the greater the noise suppression weight, the smaller the power corresponding to the frequency in the overall spectrum diagram, which indirectly indicates that the power corresponding to the frequency in the spectrum diagram of the analysis window is more accurate. This embodiment uses The power corresponding to the frequency in the overall spectrum is negatively correlated and normalized, where x represents the power corresponding to the frequency in the overall spectrum; exp is an exponential function with a natural constant as the base;

[0086] In order to accurately correct the power of the frequency in the spectrum diagram of the analysis window, and then obtain the product of the drone sound wave weight and noise suppression weight of the frequency, and the power corresponding to the frequency in the spectrum diagram of the analysis window, as the corrected power corresponding to the frequency in the spectrum diagram of the analysis window; the larger the correction power, the more likely the sound wave signal in the analysis window is generated by the drone; in order to accurately detect the location information of the suspected drone corresponding to the suspected drone signal segment through the sound wave signal, and then obtain the corrected power corresponding to each frequency in the spectrum diagram of the analysis window, considering that the power corresponding to the drone sound wave frequency in the spectrum diagram is the largest in the absence of noise interference, therefore, in this embodiment, the maximum correction power is used as the characterization power of the microphone at the analysis time point; at the same time, the frequency corresponding to the characterization power is used as the characterization frequency of the microphone at the analysis time point;

[0087] In actual situations, the strength of the sound wave signal will be partially lost in the atmosphere. In this embodiment, the position of the suspected drone corresponding to the suspected drone signal segment can be calculated by using the spherical attenuation model of the sound wave signal strength in the atmosphere and the microphone array. Furthermore, this embodiment obtains the reference distance between the suspected drone and the radar corresponding to the suspected drone signal segment obtained by the microphone at the analysis time point based on the characteristic power of each microphone at the analysis time point and the spherical attenuation model of its sound wave strength in the atmosphere; wherein, the spherical attenuation model of the sound wave signal strength in the atmosphere is well known and will not be described in detail;

[0088] Since the frequency of the sound waves emitted by the drone is relatively fixed when it is flying, but in actual situations, the distance between the drone and the microphone changes, which causes the frequency of the sound waves received by the microphone to change accordingly. Therefore, this embodiment calculates the speed of the drone by analyzing the characteristic frequency at the time point. Furthermore, this embodiment obtains the reference speed of the suspected drone corresponding to the suspected drone signal segment obtained by the microphone at the analysis time point based on the characteristic frequency of each microphone at the analysis time point, the propagation speed of the sound wave in the atmosphere, and the sound wave frequency of the drone flight.

[0089] Preferably, in an achievable manner of this embodiment, the reference distance is obtained by: for the w-th microphone, the spherical attenuation model of the sound wave intensity of the w-th microphone in the atmosphere is: ; The reference function corresponding to the w-th microphone at the analysis time point is: ; In the formula, is the attenuation of the sound wave signal intensity for the wth microphone; is the distance between the suspected drone corresponding to the suspected drone signal segment and the wth microphone; is the ratio of pi; is a logarithmic function with base 2; is the characteristic power of the w-th microphone at the analysis time point; is the sound wave power of the drone;

[0090] In order to obtain the distance between the suspected drone and the microphone corresponding to the suspected drone signal segment, the spherical attenuation model of the sound wave intensity of each microphone in the atmosphere and the reference function corresponding to each microphone at the analysis time point are combined, and the optimal solution is obtained by the optimization algorithm. The distance in the optimal solution at this time is used as the analysis distance between the suspected drone and the microphone corresponding to the suspected drone signal segment obtained at the analysis time point; wherein the optimization algorithm is a well-known technology and will not be described in detail;

[0091] In order to analyze the distance between the suspected drone and the radar corresponding to the suspected drone signal segment obtained at the analysis time point, a spherical equation of each microphone in the spatial coordinate system with the analysis distance as the radius is made with the position of each microphone as the center point; the spherical equations of each microphone are solved simultaneously, and the distance between the solved coordinate value and the radar coordinate value is used as the reference distance between the suspected drone and the radar corresponding to the suspected drone signal segment obtained through the microphone at the analysis time point.

[0092] At this point, the reference distance between the suspected drone and the radar corresponding to each suspected drone signal segment obtained by the microphone at the target time point is obtained.

[0093] Preferably, in an achievable manner of this embodiment, the reference speed is obtained by: for the w-th microphone, obtaining the target function corresponding to the w-th microphone at the analysis time point through the Doppler effect: ; In the formula, is the characteristic frequency of the w-th microphone at the analysis time point; is the sound wave frequency of the UAV flight; s is the propagation speed of the sound wave in the atmosphere; is the speed of the suspected drone corresponding to the suspected drone signal segment obtained by the wth microphone at the analysis time point; wherein the propagation speed of sound waves in the atmosphere is known, and the frequency of the sound waves of the drone flying is unknown;

[0094] In order to obtain the speed of the suspected drone corresponding to the suspected drone signal segment obtained through the sound wave signal, the objective functions corresponding to each microphone at the analysis time point are combined, and the optimal solution is obtained through the optimization algorithm. The speed in the optimal solution at this time is used as the reference speed of the suspected drone corresponding to the suspected drone signal segment obtained through the microphone at the analysis time point.

[0095] At this point, the reference speed of the suspected drone corresponding to each suspected drone signal segment obtained by the microphone at the target time point is obtained.

[0096] Step S4: Detect whether the suspected drone corresponding to each suspected drone signal segment is real based on the difference between the target speed and the reference speed of the suspected drone corresponding to each suspected drone signal segment, and the difference between the target distance and the reference distance between the suspected drone and the radar corresponding to each suspected drone signal segment.

[0097] Specifically, when the target speed of the suspected drone corresponding to a suspected drone signal segment is more similar to the reference speed, and the target distance and reference distance between the suspected drone corresponding to the suspected drone signal segment and the radar are more equal, it means that the position and speed performance of the suspected drone corresponding to the suspected drone signal segment analyzed by the radar signal and the sound wave signal are more consistent, which indirectly reflects that the suspected drone corresponding to the suspected drone signal segment is more real. Furthermore, this embodiment detects whether the suspected drone corresponding to each suspected drone signal segment is real based on the difference between the target speed and the reference speed of the suspected drone corresponding to each suspected drone signal segment, and the difference between the target distance and the reference distance between the suspected drone corresponding to each suspected drone signal segment and the radar.

[0098] Preferably, in one possible implementation of this embodiment, the method for detecting whether the suspected drone corresponding to each suspected drone signal segment is real is: for any suspected drone signal segment, the calculation formula for obtaining the authenticity of the suspected drone corresponding to the suspected drone signal segment is: ; In the formula, is the authenticity of the suspected drone corresponding to the xth suspected drone signal segment; is the target distance between the suspected UAV and the radar corresponding to the xth suspected UAV signal segment; is the reference distance between the suspected UAV and the radar corresponding to the xth suspected UAV signal segment; is the target speed of the suspected drone corresponding to the xth suspected drone signal segment; is the reference speed of the suspected drone corresponding to the xth suspected drone signal segment; is the absolute value function; exp is the exponential function with a natural constant as the base. The larger the value is, the more real the suspected drone corresponding to the xth suspected drone signal segment is.

[0099] Furthermore, this embodiment sets the preset authenticity threshold as 0.8. The implementer can set the preset authenticity threshold according to the actual situation, and it is not limited here. When the authenticity is greater than the preset authenticity threshold, the suspected drone corresponding to the suspected drone signal segment is real; when the authenticity is less than or equal to the preset authenticity threshold, the suspected drone corresponding to the suspected drone signal segment is not real.

[0100] At this point, the drones in the overhead power line area are accurately analyzed.

[0101] It should be noted that although there is a time difference in detecting the same drone by radar signals and sound wave signals, since the propagation speed of radar signals and sound wave signals in the atmosphere is much greater than the flight speed of the drone, a drone can be detected in advance before it damages the overhead power transmission lines.

[0102] In order to prevent the detected drones in the overhead power transmission line area from damaging the overhead power transmission line, this embodiment uses suppressive interference countermeasures or deceptive interference countermeasures to counter the drones in the overhead power transmission line area. Among them, suppressive interference countermeasures are radio countermeasures that use special equipment to transmit radio signals of specific frequencies to interfere with the frequency of the drone's flight control system, thereby causing the drone to lose control or be unable to fly normally, causing the drone to be forced to land or be driven away; deceptive interference countermeasures are to obtain the drone's position, speed, heading and other status information by analyzing the communication protocols such as commands, controls and sensor data used by the drone, and then transmit intermittent or continuous signals modulated with phase and frequency, so that the drone that needs to be countered makes wrong judgments or decisions, causing it to lose its way or return to the base. It effectively avoids the damage of drones to overhead power transmission lines and improves the safety of overhead power transmission lines.

[0103] In summary, this embodiment obtains the radar signal and the sound wave signal of the microphone in the overhead power transmission line area; obtains the target speed of the suspected drone and the target distance between it and the radar based on the radar signal; obtains the target time point when the suspected drone is detected by the microphone based on the target distance and the time point corresponding to the suspected drone; obtains the reference speed of the suspected drone and the reference distance between it and the radar based on the spectrum diagram of the window corresponding to the sound wave signal at the target time point and the spectrum diagram of the preset neighborhood window; detects whether the suspected drone is real based on the difference between the target speed and the reference speed, and the difference between the target distance and the reference distance. The present invention combines radar signals and sound wave signals to detect drones, improves the detection accuracy of drones, and accurately detects drones in the overhead power transmission line area.

[0104] Embodiment 2:

[0105] The present invention also proposes a drone detection system that integrates multimodal data, see Figure 2 , which shows a structural diagram of a drone detection system integrating multimodal data provided by an embodiment of the present invention. The system includes: an acquisition module 10, a radar signal processing module 20, an acoustic wave signal processing module 30 and a detection module 40.

[0106] The acquisition module 10 is used to acquire the radar signal and the sound wave signal of each microphone in the overhead transmission line area.

[0107] The radar signal processing module 20 is used to obtain suspected drone signal segments based on radar signals, and obtain the target speed of the suspected drone corresponding to each suspected drone signal segment and the target distance between the suspected drone and the radar according to the transmitted signal and the received signal in each suspected drone signal segment.

[0108] The acoustic wave signal processing module 30 is used to obtain the target time point at which the suspected drone corresponding to each suspected drone signal segment is detected by the microphone based on the target distance and the time point corresponding to each suspected drone signal segment; based on the spectrum diagram of the window corresponding to the acoustic wave signal of each microphone at the target time point and the spectrum diagram of the preset neighborhood window, obtain the reference speed of the suspected drone and the reference distance between the suspected drone and the radar corresponding to each suspected drone signal segment obtained by the microphone at the target time point.

[0109] The detection module 40 is used to detect whether the suspected drone corresponding to each suspected drone signal segment is real based on the difference between the target speed of the suspected drone corresponding to each suspected drone signal segment and the reference speed, and the difference between the target distance and the reference distance between the suspected drone corresponding to each suspected drone signal segment and the radar.

[0110] It should be noted that the system provided in the above embodiment is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the computer device is divided into different functional modules to complete all or part of the functions described above. In addition, the drone detection system integrating multimodal data and the drone detection method integrating multimodal data provided in the above embodiment belong to the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0111] Embodiment 3:

[0112] The present invention also proposes a drone detection device that integrates multimodal data, the device comprising a memory and a processor, wherein the memory stores an executable program code, and the processor is used to call and execute the executable program code to execute a drone detection method that integrates multimodal data provided in an embodiment of the present application. The device can be specifically a chip, a component or a module, and the chip may include a connected processor and a memory; wherein the memory is used to store instructions, and when the processor calls and executes the instructions, the chip can execute a drone detection method that integrates multimodal data provided in the above embodiment.

[0113] In addition, the present application embodiment also protects a computer device, see Figure 3 The computer device includes a memory 401, a processor 402, and a computer program 403 stored in the memory 401 and running on the processor 402, wherein when the processor 402 executes the computer program 403, the computer device can execute any of the drone detection methods for fusing multimodal data introduced above.

[0114] Embodiment 4:

[0115] This embodiment also provides a computer-readable storage medium, which stores computer program code. When the computer program code runs on a computer, the computer executes the above-mentioned related method steps to implement a drone detection method that integrates multi-modal data provided in the above embodiment.

[0116] Embodiment 5:

[0117] This embodiment also provides a computer program product. When the computer program product runs on a computer, it enables the computer to execute the above-mentioned related steps to implement a drone detection method that integrates multi-modal data provided by the above embodiment.

[0118] Among them, the device, computer-readable storage medium, computer program product or chip provided in this embodiment is used to execute the corresponding method provided above. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding method provided above, and will not be repeated here.

[0119] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0120] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A drone detection method integrating multimodal data, characterized in that: The method comprises the following steps: Acquire radar signals and acoustic wave signals of each microphone in the area of ​​overhead power transmission lines; Based on the radar signal, a suspected drone signal segment is obtained, and according to the transmission signal and the reception signal in each suspected drone signal segment, the target speed of the suspected drone corresponding to each suspected drone signal segment and the target distance between the suspected drone and the radar are obtained; Based on the target distance and the time point corresponding to each suspected drone signal segment, the target time point at which the suspected drone corresponding to each suspected drone signal segment is detected by the microphone is obtained; based on the spectrum diagram of the window corresponding to the sound wave signal of each microphone at the target time point and the spectrum diagram of the preset neighborhood window, the reference speed of the suspected drone and the reference distance between the suspected drone and the radar corresponding to each suspected drone signal segment obtained by the microphone at the target time point are obtained; Based on the difference between the target speed and the reference speed of the suspected drone corresponding to each suspected drone signal segment, and the difference between the target distance and the reference distance between the suspected drone and the radar corresponding to each suspected drone signal segment, it is detected whether the suspected drone corresponding to each suspected drone signal segment is real.

2. The method for detecting drones by fusing multimodal data according to claim 1, characterized in that: The method for obtaining the suspected drone signal segment is: Divide the radar signal according to the time of transmitting pulses to obtain radar signal segments; The power of each radar signal segment is obtained through the radar equation as the target power of each radar signal segment; The minimum detectable power of the radar signal is obtained. When the target power is greater than the minimum detectable power, the corresponding radar signal segment is regarded as a suspected drone signal segment.

3. The method for detecting drones by fusing multimodal data according to claim 1, characterized in that: The method for obtaining the target speed of the suspected drone corresponding to each suspected drone signal segment and the target distance between the suspected drone and the radar is as follows: For any suspected drone signal segment, according to the frequency of the transmitted signal and the received signal in the suspected drone signal segment, and the propagation speed of the radar signal in the atmosphere, through the Doppler effect, according to the formula , obtain the target speed of the suspected drone corresponding to the suspected drone signal segment; where, is the frequency of the received signal in the xth suspected drone signal segment; is the frequency of the transmitted signal in the xth suspected drone signal segment; c is the propagation speed of the radar signal in the atmosphere; is the target speed of the suspected drone corresponding to the xth suspected drone signal segment; Obtaining a time difference between a transmitted signal and a received signal in the suspected drone signal segment as a first time difference; Half of the result of multiplying the first time difference by the propagation speed of the radar signal in the atmosphere is taken as the target distance between the suspected drone and the radar corresponding to the suspected drone signal segment.

4. The method for detecting drones by fusing multimodal data according to claim 1, characterized in that: The method for obtaining the target time point is: For any suspected drone signal segment, the target distance between the suspected drone and the radar corresponding to the suspected drone signal segment is taken as the first distance; taking the ratio of the first distance to the propagation speed of the radar signal in the atmosphere as the first time period; The ratio of the first distance to the propagation speed of the sound wave in the atmosphere is used as the second time period; The absolute value of the difference between the first time period and the second time period is used as the reference time difference when the suspected drone corresponding to the suspected drone signal segment is detected by the radar and the microphone; The time at the specified position in the suspected drone signal segment is used as the time point of the suspected drone signal segment; The sum of the time point of the suspected drone signal segment and the reference time difference is used as the target time point when the suspected drone corresponding to the suspected drone signal segment is detected by the microphone.

5. The method for detecting drones by fusing multimodal data according to claim 1, characterized in that: The method for obtaining the reference speed of the suspected drone and the reference distance between the suspected drone and the radar corresponding to each suspected drone signal segment is as follows: For any suspected drone signal segment, the target time point at which the suspected drone corresponding to the suspected drone signal segment is detected by the microphone is used as the analysis time point; The sound wave signal of each microphone is divided into preset windows, and the spectrum of each window is obtained by Fourier transform; For any microphone, the window corresponding to the sound wave signal of the microphone at the analysis time point is used as the analysis window; The average spectrum of the spectrum of the analysis window and its preset neighboring window is used as the overall spectrum of the analysis window; For any frequency in the overall spectrum, the result of negative correlation and normalization between the absolute value of the difference between the frequency and the drone's designated sound wave frequency is used as the drone sound wave weight of the frequency; The power corresponding to the frequency in the overall spectrum is negatively correlated and normalized, and the result is used as the noise suppression weight of the frequency; Obtain the product of the drone sound wave weight and noise suppression weight of the frequency and the power corresponding to the frequency in the spectrum diagram of the analysis window as the corrected power corresponding to the frequency in the spectrum diagram of the analysis window; Obtain the corrected power corresponding to each frequency in the spectrum diagram of the analysis window, and use the maximum corrected power as the representative power of the microphone at the analysis time point; use the frequency corresponding to the representative power as the representative frequency of the microphone at the analysis time point; According to the characteristic power of each microphone at the analysis time point and the spherical attenuation model of its sound wave intensity in the atmosphere, the reference distance between the suspected drone and the radar corresponding to the suspected drone signal segment obtained by the microphone at the analysis time point is obtained; According to the characteristic frequency of each microphone at the analysis time point, the propagation speed of sound waves in the atmosphere and the sound wave frequency of the drone flight, the reference speed of the suspected drone corresponding to the suspected drone signal segment obtained by the microphone at the analysis time point is obtained.

6. The method for detecting drones by fusing multimodal data according to claim 5, characterized in that: The method for obtaining the reference distance is: For the w-th microphone, the spherical attenuation model of the sound wave intensity of the w-th microphone in the atmosphere is: ; The reference function corresponding to the w-th microphone at the analysis time point is: ; In the formula, is the attenuation of the sound wave signal intensity for the wth microphone; is the distance between the suspected drone corresponding to the suspected drone signal segment and the wth microphone; is the ratio of pi; is a logarithmic function with base 2; is the characteristic power of the w-th microphone at the analysis time point; is the sound wave power of the drone; The spherical attenuation model of the sound wave intensity of each microphone in the atmosphere and the reference function corresponding to each microphone at the analysis time point are combined, and the optimal solution is obtained through the optimization algorithm. The distance in the optimal solution at this time is used as the analysis distance between the suspected drone and the microphone corresponding to the suspected drone signal segment obtained at the analysis time point; Taking the position of each microphone as the center point, make the spherical equation of each microphone in the spatial coordinate system with the analysis distance as the radius; The spherical equations of each microphone are solved simultaneously, and the distance between the solved coordinate value and the radar coordinate value is used as the reference distance between the suspected drone and the radar corresponding to the suspected drone signal segment obtained by the microphone at the analysis time point.

7. The method for detecting drones by fusing multimodal data according to claim 6, characterized in that: The method for obtaining the reference speed is: For the w-th microphone, the objective function corresponding to the w-th microphone at the analysis time point is obtained through the Doppler effect: ; In the formula, is the characteristic frequency of the w-th microphone at the analysis time point; is the sound wave frequency of the UAV flight; s is the propagation speed of the sound wave in the atmosphere; is the speed of the suspected drone corresponding to the suspected drone signal segment obtained by the wth microphone at the analysis time point; The objective functions corresponding to each microphone at the analysis time point are combined, and the optimal solution is obtained through the optimization algorithm. The speed in the optimal solution at this time is used as the reference speed of the suspected drone corresponding to the suspected drone signal segment obtained by the microphone at the analysis time point.

8. The method for detecting drones by fusing multimodal data according to claim 1, characterized in that: The method for detecting whether the suspected drone corresponding to each suspected drone signal segment is real is: For any suspected drone signal segment, the calculation formula for obtaining the authenticity of the suspected drone corresponding to the suspected drone signal segment is: ; In the formula, is the authenticity of the suspected drone corresponding to the xth suspected drone signal segment; is the target distance between the suspected UAV and the radar corresponding to the xth suspected UAV signal segment; is the reference distance between the suspected UAV and the radar corresponding to the xth suspected UAV signal segment; is the target speed of the suspected drone corresponding to the xth suspected drone signal segment; is the reference speed of the suspected drone corresponding to the xth suspected drone signal segment; is the absolute value function; exp is the exponential function with a natural constant as the base; When the authenticity is greater than the preset authenticity threshold, the suspected drone corresponding to the suspected drone signal segment is authentic; When the authenticity is less than or equal to the preset authenticity threshold, the suspected drone corresponding to the suspected drone signal segment is not authentic.

9. A drone detection device integrating multimodal data, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the processor implements the steps of a drone detection method integrating multimodal data as described in any one of claims 1 to 8.

10. A drone detection system integrating multimodal data, characterized in that: The system comprises: An acquisition module, used for acquiring radar signals and acoustic wave signals of each microphone in the overhead power transmission line area; A radar signal processing module is used to obtain a suspected drone signal segment based on the radar signal, and obtain the target speed of the suspected drone corresponding to each suspected drone signal segment and the target distance between the suspected drone and the radar according to the transmission signal and the reception signal in each suspected drone signal segment; The acoustic signal processing module is used to obtain the target time point at which the suspected drone corresponding to each suspected drone signal segment is detected by the microphone based on the target distance and the time point corresponding to each suspected drone signal segment; based on the spectrum diagram of the window corresponding to the acoustic signal of each microphone at the target time point and the spectrum diagram of the preset neighborhood window, obtain the reference speed of the suspected drone and the reference distance between the suspected drone and the radar corresponding to each suspected drone signal segment obtained by the microphone at the target time point; The detection module is used to detect whether the suspected drone corresponding to each suspected drone signal segment is real based on the difference between the target speed of the suspected drone corresponding to each suspected drone signal segment and the reference speed, and the difference between the target distance and the reference distance between the suspected drone corresponding to each suspected drone signal segment and the radar.

Images