A high-altitude projectile detection and discrimination method based on millimeter-wave radar data fitting

Abstract

The present invention discloses a method for detecting and discriminating high-altitude objects based on millimeter-wave radar data fitting. This method uses a millimeter-wave radar system to monitor suspected object-dropping areas near high-rise buildings in real time, obtaining the range-Doppler detection trajectory of moving targets. The range-Doppler characteristic curve of the moving target detected by radar is then used to discriminate against high-altitude object dropping. By fitting the clutter-suppressed range-Doppler trajectory of the moving target with the simulated characteristics of classical parabolic motion, the presence of high-altitude object dropping is determined. This invention effectively provides a method for determining high-altitude object dropping trajectory using millimeter-wave radar, significantly improving the feasibility and reliability of assessing high-altitude object dropping.

Description

Technical Field

[0001] The present invention relates to the fields of millimeter wave radar detection and urban safety, and in particular to a method for detecting and distinguishing high-altitude projectiles based on millimeter wave radar data fitting. Background Art

[0002] Monitoring objects dropped from high places is key to preventing and deterring such behavior, and both academia and industry have conducted relevant research on this issue. Existing methods for monitoring objects dropped from high places rely on visual detection technology. However, vision-based monitoring solutions are highly susceptible to environmental factors, with performance severely deteriorating in dim light (at night) or extreme weather conditions (fog, dust, etc.).

[0003] The fundamental principles of radar give it a natural advantage in detecting moving targets. Millimeter-wave radar, operating in the millimeter-wave (30-300 GHz) frequency band, has rapidly developed in recent years due to its compact size, all-weather operation, and extremely high spatial resolution. It is widely used in the detection of moving targets and has great potential in detecting a wide variety of high-altitude objects. However, due to the complex electromagnetic environment in cities, millimeter-wave radar can detect interference targets unrelated to high-altitude objects, posing a greater challenge to the identification of high-altitude object dropping.

[0004] For millimeter-wave radar, the motion characteristics of high-altitude projectile targets and the extraction of their radar range-Doppler characteristic curves are of great significance to improving the ability to determine the behavior of high-altitude projectiles. Summary of the Invention

[0005] In response to the shortcomings of current millimeter-wave radar research on high-altitude parabolic target detection and the lack of effective means to distinguish parabolic targets from interference targets, this paper proposes to use the range-Doppler characteristic curve of the target in radar detection to distinguish the behavior of high-altitude parabolic objects. By comparing it with the range-Doppler simulation characteristics of parabolic motion in classical mechanics, the reliability of high-altitude parabolic object discrimination is improved.

[0006] To achieve the above technical objectives, the technical solution of the present invention is: a method for detecting and distinguishing high-altitude projectiles based on millimeter-wave radar data fitting, comprising the following steps:

[0007] Real-time monitoring of suspected parasitic areas in high-rise buildings through millimeter-wave radar systems;

[0008] Radar echo signal sampling and processing to obtain the range-Doppler detection trajectory of the moving target;

[0009] Adjust the parabolic model parameters and fit it to the range-Doppler detection trajectory of the moving target to obtain the fitting error;

[0010] The parabolic discriminant equation is set according to the fitting error, and the parabolic discriminant equation is used to determine whether the moving target is a parabolic target.

[0011] Furthermore, the millimeter-wave radar system is installed on the ground side of a building, and the detection beam direction of the radar antenna is vertically upward and parallel to the plane of the building wall.

[0012] Furthermore, the millimeter-wave radar system uses a single antenna to transmit signals and four antennas with the same spacing and arranged linearly to receive signals; the transmission signal of the millimeter-wave radar system is a linear frequency modulated continuous wave signal.

[0013] Furthermore, the linear frequency modulation continuous wave signal adopts a frequency band of 76-81 GHz.

[0014] Furthermore, the radar echo signal is processed by two fast Fourier transforms, the background interference points and noise are processed by a high-pass filter and a constant false alarm algorithm, and then the range-Doppler detection trajectory of the moving target is obtained by clustering.

[0015] Furthermore, the high-pass filter adopts an FIR high-pass filter; and the constant false alarm algorithm adopts a unit maximum constant false alarm method.

[0016] Furthermore, the parabolic model is a classical mechanics parabolic model that regards the moving target as a point mass.

[0017] Furthermore, the least squares method is used for fitting to obtain a simulated range-Doppler map; the mean square error between the coordinates of the points actually detected in the range-Doppler map and the coordinates of the sampling points in the simulated range-Doppler map is calculated, and the minimum value of the mean square error is taken as the fitting error; the determination coefficient is calculated using the residual square sum of the actual detection points and the fitting points and the regression square sum of all detection points.

[0018] Furthermore, the parabolic discriminant equation is:

[0019] p=β·(R 2 +τ·ε+ψ(n))

[0020] Among them, β represents the monitoring coefficient, τ represents the error normalization coefficient, ε represents the fitting error, R 2 represents the coefficient of determination, n is the number of valid points that satisfy the parabolic trajectory, and ψ(n) is the normalized limit function.

[0021] Furthermore, when the radar detects a target, it starts to calculate the value p of the parabolic discriminant equation in real time. When p>=1, it is judged as a high-altitude parabolic behavior, an alarm signal is issued and the data of the corresponding multi-frequency modulation cycle is recorded; when p<1, it is judged as an interference target, and continues to monitor and update the p value.

[0022] The present invention provides a method for detecting and distinguishing high-altitude projectiles based on millimeter-wave radar data fitting. This method, combined with radar clutter suppression methods, effectively extracts the range-Doppler trajectory of detected target objects by analyzing the range-Doppler response of radar echoes. By combining the classical mechanics parabolic model with theoretical simulation and actual range-Doppler curve fitting, the present invention provides a method for determining the trajectory of high-altitude projectiles using millimeter-wave radar, effectively improving the reliability of millimeter-wave radar in determining high-altitude projectile behavior. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A flowchart of a specific embodiment of a method for detecting and distinguishing high-altitude projectiles based on millimeter-wave radar data fitting;

[0024] Figure 2 A parabolic range-Doppler response diagram obtained by detecting and removing a large number of interference points in a radar multi-frequency modulation period in a specific embodiment of the method of the present invention;

[0025] Figure 3 This is a distance-Doppler detection curve obtained by direct clustering processing in a specific embodiment of the method of the present invention;

[0026] Figure 4 This is a comparison diagram of a simulated range-Doppler curve with the minimum fitting error and an actual detection curve in a specific embodiment of the method of the present invention. DETAILED DESCRIPTION

[0027] Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. In the following description, when referring to the drawings, like numbers in different figures represent like or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present invention, as detailed in the appended claims.

[0028] The terms used in this invention are for the purpose of describing specific embodiments only and are not intended to limit the invention. The singular forms "a," "the," and "the" used in this invention and the appended claims are also intended to include plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0029] It should be understood that although the terms "first," "second," "third," etc. may be used in the present invention to describe various information, such information should not be limited to these terms. These terms are merely used to distinguish information of the same type from one another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the present invention. Depending on the context, the term "if" as used herein may be interpreted as "when," "when," or "in response to determining."

[0030] The technical method of the present invention will be further described below in conjunction with the accompanying drawings in the embodiments of the present invention. Based on the embodiments of the present invention, other embodiments obtained by ordinary technicians in this field without creative work all fall within the scope of protection of the present invention.

[0031] One embodiment of the present invention, as Figure 1 As shown, the present invention proposes a high-altitude projectile detection and discrimination method based on millimeter-wave radar processing, which includes the following specific steps:

[0032] (1) Deploy the millimeter-wave radar system: Set up the millimeter-wave radar system in the high-rise building monitoring area;

[0033] Exemplarily and preferably, the millimeter wave radar system is installed on the ground near the north and south sides of the building, and the radar antenna detection beam direction is vertically upward and parallel to the plane of the building wall.

[0034] Furthermore, the millimeter-wave radar system uses a single antenna to transmit signals and four equally spaced, linearly arranged antennas to receive them. The echoes received by the four antennas can be non-coherently accumulated to improve the signal-to-noise ratio and subsequently used for parabolic target position analysis.

[0035] Furthermore, the transmission signal of the millimeter wave radar system is a linear frequency modulated continuous wave signal with adjustable bandwidth in the 76-81 GHz frequency band.

[0036] (2) Sampling of radar echo signals with a single frequency modulation period: The millimeter-wave radar system transmits signals to monitor the suspected parabolic area of a high-rise building, and performs down-conversion demodulation and real-time digital sampling on the radar echo to obtain the original sampling data of the radar echo.

[0037] (3) Processing of radar echo signals with multiple frequency modulation cycles: The original complex signal data obtained by sampling the echo signals with multiple frequency modulation cycles is subjected to two fast Fourier transforms to obtain a range-Doppler response diagram containing the parabolic target and background interference targets. The static background clutter interference is filtered through a high-pass filter, and the constant false alarm (CFAR) algorithm is used for detection to obtain a range-Doppler response diagram with only a few interference points remaining, thereby obtaining the range-Doppler detection trajectory of the moving target under multiple pulses. Specifically, since parabolic objects cannot be regarded as ideal particles, when there are high-altitude parabolic targets, multiple detection points on the range-Doppler response diagram usually correspond to the target. It is necessary to perform clustering processing and analysis on the range-Doppler response diagram with only a few interference points remaining to obtain a reasonable detection target trajectory on the range-Doppler diagram.

[0038] Specifically, the background interference target may be a disturbance in the environment within the detection range, such as the fluttering of clothes drying on the balcony, the swaying of leaves, etc.; or a non-parabolic target appearing within the detection range, such as a flying bird.

[0039] Specifically, the constant false alarm algorithm adjusts the detection threshold of the radar system to keep the false detection probability constant, and determines whether the detected signal is a target or noise by comparing the signal strength with the detection threshold.

[0040] The constant false alarm rate (CFAR) algorithm includes a unit average method (CA-CFAR), a unit maximum method (GO-CFAR), an ordered statistics method (OS-CFAR) algorithm, and the like.

[0041] More preferably, see Figure 2 In this embodiment of the present invention, radar echo signals within multiple frequency modulation periods are processed using the unit maximum constant false alarm (GO-CFAR) method and FIR high-pass filtering (Finite Impulse Response filter) to obtain a range-Doppler response diagram of a parabolic target with a large number of interference points removed. By superimposing the amplitudes of all the processed range-Doppler response diagrams, a relatively clear range-Doppler response change process can be obtained. Figure 2 .

[0042] Further preferably, in the embodiment of the present invention, a direct clustering method is used for all existing range-Doppler response maps to obtain the target center position in each range-Doppler map, thereby obtaining a reasonable target range-Doppler trajectory, see Figure 3 .

[0043] (4) Parabolic range-Doppler characteristic simulation and fitting: When there is a detected range-Doppler trajectory of a moving target, the parabolic motion of a particle target in classical mechanics is simulated and the range-Doppler trajectory obtained by simulation is used for fitting. The model parameters are adjusted to obtain a simulation trajectory with the minimum fitting error with the detected trajectory.

[0044] Specifically, a parabolic falling model is established based on classical mechanics, and the ideal range-Doppler trajectory simulation results are obtained according to the radial distance and velocity from the detected target to the radar position.

[0045] Further preferably, the parabolic model has the following differential equation form, and the air resistance experienced by the object is proportional to the quadratic of the velocity, specifically:

[0046]

[0047] is the object's speed, is the object motion coordinate, is the acceleration due to gravity, and α is the drag coefficient.

[0048] Specifically, the range-Doppler simulation characteristic trajectory and the detection trajectory are fitted using the least squares method, and the fitting model parameters are adjusted to obtain the simulated range-Doppler trajectory.

[0049] Furthermore, the fitting error is estimated using the mean square error to ensure that the difference between the range-Doppler curve simulated by adjusting the parameters and the actual detected range-Doppler curve is minimized, that is, (r det ,v det ) is the coordinate of the point actually detected in the range-Doppler map, r det is the set of distances actually detected, v det is the set of actually detected speeds, (r sim ,v sim ) is the coordinate of the sampling point of the range-Doppler diagram obtained in the parabolic simulation, r sim is the set of distances calculated by simulation, v sim is the set of simulated velocities, n is the number of detection points in the range-Doppler diagram. Record the fitting model obtained by the least squares method and the coefficient of determination R of the fitting result. 2 and the fitting error ε.

[0050] Specifically, the fitting error ε and the determination coefficient R 2 The general expression is:

[0051]

[0052] The numerator of the second term in formula (3) is the residual sum of squares between all actual detection points and all fitting points, and the denominator is the regression sum of squares between all actual detection points and the average value of the detection points.

[0053] In the embodiment of the present invention, see Figure 4 The figure shows the actual detection results of the range-Doppler trajectory and the optimal fitting results. The optimal drag coefficient α is 1.1718, the optimal initial target height is 17.75m, the optimal initial parabolic speed is 3.12m / s, and the determination coefficient R 2 is 0.87, and the fitting error ε is 0.42.

[0054] Furthermore, through subsequent simulation of the range-Doppler trajectory of various parabolic objects, the radar feature set of each object can be obtained to ensure the accuracy of high-altitude parabolic behavior judgment.

[0055] (5) Use the fitting error results to perform parabolic discrimination: According to the needs of high-altitude parabolic monitoring in actual buildings and the fitting error, a parabolic discrimination equation is set. If the equation value of the parabolic discrimination equation is greater than the set reasonable threshold, it is judged as a high-altitude parabolic behavior, a warning signal is issued and the data of the corresponding multi-frequency modulation cycle is recorded to provide analytical data for subsequent parabolic floor estimation; if the equation value of the parabolic discrimination equation is less than the interference threshold, it is judged as an interference target.

[0056] Specifically, combined with the obtained goodness of fit parameters, the high-altitude parabolic judgment equation is set as:

[0057] p=β·(R 2 +τ·ε+ψ(n)) (4)

[0058] Where β represents the monitoring coefficient and can be set between 0.3 and 0.4 depending on the building monitoring requirements. For buildings with high risk and high frequency of high-altitude parabolic projections, it is set to 0.4, while for buildings with low risk and low frequency, it is set to 0.3. τ represents the error normalization coefficient, which is related to the radar installation environment and the type of parabolic projection. In practice, it can be determined based on known parabolic projection experiments and is generally initially set to 0.1. n is the number of valid points in the parabolic trajectory (also the number of times the target appears in the radar RD map), and ψ(n) is the normalized limit function Sigmoid, which is formulated as follows:

[0059] Furthermore, when the radar detects a target, it begins to calculate the value p of the parabolic discrimination equation in real time. When p>=1, it is judged as a high-altitude parabolic behavior, an alarm signal is issued and the corresponding multi-frequency modulation cycle echo data is recorded to provide analysis for subsequent parasitic resident judgment; when p<1, it is judged as an interference target, and continues to monitor and update the p value until the judgment conditions are met.

[0060] See also Figure 2、 3 4 Experimental results: In the embodiment of the present invention, the monitoring coefficient β is set to 0.36, the normalization coefficient τ is set to 2.4, the number of valid points n in the parabolic trajectory is 82, and the value p of the discriminant equation obtained by calculation is 1.04, which can be determined to be a high-altitude parabolic target.

[0061] The above embodiments are intended only to illustrate the design concepts and features of the present invention. Their purpose is to enable those skilled in the art to understand the contents of the present invention and implement them accordingly. The scope of protection of the present invention is not limited to the above embodiments. Therefore, any equivalent changes or modifications made based on the principles and design concepts disclosed in the present invention are within the scope of protection of the present invention.

Claims

1. A method for detecting and distinguishing high-altitude projectiles based on millimeter-wave radar data fitting, characterized in that: The following steps are involved: Real-time monitoring of suspected parasitic areas in high-rise buildings through millimeter-wave radar systems; Radar echo signal sampling and processing to obtain the range-Doppler detection trajectory of the moving target; Adjusting the parabolic model parameters and fitting it with the range-Doppler detection trajectory of the moving target to obtain a fitting error; including: using the least squares method to fit and obtain a simulated range-Doppler map; calculating the mean square error between the coordinates of the points actually detected in the range-Doppler map and the coordinates of the sampling points in the simulated range-Doppler map, and taking the minimum value of the mean square error as the fitting error; and calculating the determination coefficient using the residual square sum of the actual detection point and the fitting point and the regression square sum of all detection points; The parabolic discriminant equation is set according to the fitting error, and the parabolic discriminant equation is used to determine whether the moving target is a parabolic target.

2. The method for detecting and distinguishing high-altitude projectiles based on millimeter-wave radar data fitting according to claim 1 is characterized in that: The millimeter wave radar system is installed on the ground side of the building, and the radar antenna detection beam direction is vertically upward and parallel to the plane of the building wall.

3. The method for detecting and distinguishing high-altitude projectiles based on millimeter-wave radar data fitting according to claim 1, characterized in that: The millimeter-wave radar system uses a single antenna to transmit signals and four antennas with the same spacing and arranged linearly to receive signals; the transmission signal of the millimeter-wave radar system is a linear frequency modulated continuous wave signal.

4. The method for detecting and distinguishing high-altitude projectiles based on millimeter-wave radar data fitting according to claim 3 is characterized in that: The linear frequency modulation continuous wave signal adopts a frequency band of 76-81 GHz.

5. The method for detecting and distinguishing high-altitude projectiles based on millimeter-wave radar data fitting according to claim 1, characterized in that: The radar echo signal is processed by two fast Fourier transforms, the background interference points and noise are processed by high-pass filter and constant false alarm algorithm, and the range-Doppler detection trajectory of the moving target is obtained by clustering.

6. The method for detecting and distinguishing high-altitude projectiles based on millimeter-wave radar data fitting according to claim 5 is characterized in that: The high-pass filter adopts an FIR high-pass filter; the constant false alarm algorithm adopts a unit average method, a unit maximum method or an ordered statistics method.

7. The method for detecting and distinguishing high-altitude projectiles based on millimeter-wave radar data fitting according to claim 1, characterized in that: The parabolic model is a classical mechanics parabolic model that regards the moving target as a mass point.

8. The method for detecting and distinguishing high-altitude projectiles based on millimeter-wave radar data fitting according to claim 1, characterized in that: The parabolic discriminant equation is: p=β·(R 2 +τ·ε+ψ(n)) Among them, β represents the monitoring coefficient, τ represents the error normalization coefficient, ε represents the fitting error, R 2 represents the coefficient of determination, n is the number of valid points that satisfy the parabolic trajectory, and ψ(n) is the normalized limit function.

9. The method for detecting and distinguishing high-altitude projectiles based on millimeter-wave radar data fitting according to claim 8, characterized in that: When the radar detects a target, it starts to calculate the value p of the parabolic discriminant equation in real time. When p>=1, it is judged as a high-altitude parabolic behavior, an alarm signal is issued and the data of the corresponding multi-frequency modulation cycle is recorded; when p<1, it is judged as an interference target, and continues to monitor and update the p value.

Images