Radar-fused ultra-low-altitude flight joint detection method and system

Abstract

The invention relates to the technical field of unmanned aerial vehicles, in particular to a radar-fused ultra-low-altitude flight joint detection method and system. The method comprises the following steps: acquiring an initial radar echo signal of an ultra-low-altitude area, and identifying a target radar echo signal corresponding to a flying target in the ultra-low-altitude area; obtaining a preliminary classification result of the flying target according to the Doppler frequency shift center sequence and the range walk sequence; if the preliminary classification result is a suspected small target, empirical mode decomposition is carried out on the target radar echo signal, and a micro-Doppler component set of the flying target is extracted based on an intrinsic mode function obtained through decomposition; spectral analysis is carried out on the target radar echo signal to obtain harmonic structure characteristics of the flight target; and performing trajectory association on the multiple frames of target plots through a Hungary algorithm, and determining that the flight target is an interferent or a rotor unmanned aerial vehicle according to the associated trajectory features. According to the invention, the overall accuracy and stability of ultra-low-altitude flight target detection can be improved.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) technology, and in particular to a method and system for joint detection of ultra-low-altitude flight using fused radar. Background Technology

[0002] In modern airspace safety management, the detection and identification of ultra-low-altitude flying targets is directly related to air defense security and flight control.

[0003] Current technologies primarily rely on receiving radar echo signals and using echo amplitude and time delay information to detect flying targets. However, this approach has significant shortcomings in data processing: when faced with multi-source, heterogeneous detection information, it lacks a scientifically effective data fusion and collaborative analysis mechanism, failing to organically integrate information from different sources. This results in insufficient information utilization and deviations in the determination of the flying target's status. Especially in application scenarios with complex target characteristics and severe environmental interference, existing technologies are highly prone to target misjudgment and missed detection.

[0004] In summary, existing ultra-low-altitude target detection schemes suffer from low accuracy and poor stability. Summary of the Invention

[0005] This application provides a method and system for joint detection of ultra-low-altitude flying targets using fusion radar, in order to improve the accuracy and stability of ultra-low-altitude target detection.

[0006] According to one aspect of this application, a method for joint detection of ultra-low-altitude flight using fused radar is provided, comprising:

[0007] The initial radar echo signal in the ultra-low altitude region is acquired, and the initial radar echo signal is processed by the least mean square adaptive filtering algorithm to generate a purified radar echo signal.

[0008] Target detection is performed on the purified radar echo signal to identify the target radar echo signal corresponding to the flying target in the ultra-low altitude area;

[0009] The Doppler frequency shift center sequence and range travel sequence of the target radar echo signal are acquired, and a preliminary classification result of the flight target is obtained based on the Doppler frequency shift center sequence and the range travel sequence; wherein, the preliminary classification result includes suspected small targets, large aircraft, and invalid targets;

[0010] If the preliminary classification result is the suspected small target, then empirical mode decomposition is performed on the target radar echo signal, and the micro-Doppler component set of the flight target is extracted based on the intrinsic mode function obtained from the decomposition; and spectrum analysis is performed on the target radar echo signal to obtain the harmonic structure characteristics of the flight target; wherein, the harmonic structure characteristics include the harmonic amplitude ratio and the proportion of harmonic energy;

[0011] The target radar echo signal is used to obtain multiple frames of target points of the flying target.

[0012] Using the Hungarian algorithm, combined with the micro-Doppler component set and the harmonic structure features, the trajectory of the multi-frame target points is correlated, and the flight target is determined to be either a jamming object or a rotary-wing UAV based on the correlated trajectory features.

[0013] Optionally, the step of acquiring the initial radar echo signal in the ultra-low altitude region and processing the initial radar echo signal using a least mean square adaptive filtering algorithm to generate a purified radar echo signal includes:

[0014] Initial radar echo signals in the ultra-low altitude region are acquired using a multi-channel sensor array.

[0015] Perform a Fourier transform on the initial radar echo signal to obtain the time spectrum of the target radar echo signal, and determine whether the initial radar echo signal is subject to strong electromagnetic interference based on the time spectrum.

[0016] If the initial radar echo signal is subject to strong electromagnetic interference, the initial radar echo signal is processed by the least mean square adaptive filtering algorithm to generate the purified radar echo signal; if the initial radar echo signal is not subject to strong electromagnetic interference, the initial radar echo signal is used as the purified radar echo signal.

[0017] Optionally, determining whether the initial radar echo signal is subject to strong electromagnetic interference based on the time-spectrum diagram includes:

[0018] Traverse each frequency band and time frame of the time-frequency spectrum to extract the amplitude and phase values ​​of the initial radar echo signal at each time-frequency point;

[0019] The amplitude jitter standard deviation of the initial radar echo signal is calculated based on the amplitude value; the phase jump angle of the initial radar echo signal is calculated based on the phase value.

[0020] The initial radar echo signal is determined to be subject to strong electromagnetic interference based on the amplitude jitter standard deviation and the phase jump angle.

[0021] Optionally, the step of performing target detection on the purified radar echo signal to identify the target radar echo signal corresponding to the flying target in the ultra-low altitude region includes:

[0022] The constant false alarm rate (CFAR) detection algorithm is used to process the purified radar echo signal to identify the target radar echo signal corresponding to the flying target in the ultra-low altitude region.

[0023] Optionally, acquiring the Doppler frequency shift center sequence and range travel sequence of the target radar echo signal, and obtaining the preliminary classification result of the flight target based on the Doppler frequency shift center sequence and the range travel sequence, includes:

[0024] Frequency domain analysis is performed on the target radar echo signal to obtain the Doppler frequency domain peak value. The Doppler frequency domain peak value is sorted by frame to obtain the Doppler frequency shift center sequence of the target radar echo signal.

[0025] Based on the principle of radar ranging, the radial distance from the flying target to the multi-channel sensor array is calculated according to the target radar echo signal. The radial distance is sorted by frame to obtain the range movement sequence of the target radar echo signal.

[0026] The radial velocity variation trend of the flight target is obtained based on the Doppler frequency shift center sequence, and the distance trajectory curvature of the flight target is obtained based on the distance movement sequence; wherein, the radial velocity variation trend includes the velocity change rate, velocity fluctuation standard deviation, and velocity mean deviation of the flight target, and the distance trajectory curvature includes the average trajectory curvature, curvature change amplitude, and trajectory fitting residual ratio of the flight target;

[0027] Using the radial velocity change trend and the distance trajectory curvature as inputs, a clustering algorithm is used to obtain the preliminary classification results of the flight target.

[0028] Optionally, if the preliminary classification result is the suspected small target, then empirical mode decomposition is performed on the target radar echo signal, and the micro-Doppler component set of the flying target is extracted based on the intrinsic mode functions obtained from the decomposition, including:

[0029] An adaptive filtering algorithm is used to preprocess the target radar echo signal;

[0030] Empirical mode decomposition is performed on the preprocessed target radar echo signal to obtain several eigenmode functions;

[0031] The intra-frame frequency features corresponding to the intrinsic mode functions are extracted by Hilbert transform, and the effective frequency components that conform to the small target category are selected. The effective frequency components are sorted from low to high frequency to obtain the micro-Doppler component set.

[0032] Optionally, the step of performing spectral analysis on the target radar echo signal to obtain the harmonic structure characteristics of the flying target includes:

[0033] A fast Fourier transform is performed on the preprocessed target radar echo signal to obtain the frequency domain spectrum of the target radar echo signal.

[0034] The ratio of the fundamental frequency to the harmonic amplitude of each harmonic of the target radar echo signal, and the proportion of the harmonic energy of each harmonic, are calculated based on the frequency domain spectrum.

[0035] Optionally, the step of using the Hungarian algorithm, combining the micro-Doppler component set and the harmonic structure features, to perform trajectory association on the multi-frame target points, and determining whether the flight target is an interference object or a rotary-wing UAV based on the associated trajectory features, includes:

[0036] Extract the frequency features from the set of microDoppler components;

[0037] Using a multi-feature weighted fusion algorithm, the frequency features, the harmonic amplitude ratio, and the harmonic energy ratio are used as association constraints to construct the association cost matrix between the target points in the multiple frames.

[0038] The optimal matching path for the flight target is solved using the Hungarian algorithm based on the correlation cost matrix.

[0039] The trajectory characteristics of the flight target are obtained based on the optimal matching path, and the flight target is determined to be a jamming object or a rotary-wing UAV based on the trajectory characteristics.

[0040] Optionally, obtaining the trajectory features of the flight target based on the optimal matching path, and determining whether the flight target is a jamming object or a rotary-wing UAV based on the trajectory features, includes:

[0041] The stable trajectory of the flight target is obtained according to the optimal matching path, and trajectory features are extracted based on the stable trajectory; wherein, the trajectory features include trajectory curvature, motion speed and trajectory continuity;

[0042] If the trajectory curvature is within a preset range of rotary-wing UAV curvature thresholds, the speed is within a preset range of rotary-wing UAV speed thresholds, and the trajectory continuity is within a preset range of rotary-wing UAV trajectory continuity thresholds, then the flying target is determined to be a rotary-wing UAV; otherwise, the flying target is determined to be an interference object.

[0043] According to another aspect of this application, a combined detection system for ultra-low-altitude flight using fused radar is provided, comprising:

[0044] The signal acquisition module is used to acquire the initial radar echo signal in the ultra-low altitude region, and process the initial radar echo signal through the least mean square adaptive filtering algorithm to generate a purified radar echo signal.

[0045] The target detection module is used to detect targets in the purified radar echo signal and identify the target radar echo signal corresponding to the flying target in the ultra-low altitude area.

[0046] The initial screening module is used to acquire the Doppler frequency shift center sequence and range travel sequence of the target radar echo signal, and to obtain the preliminary classification result of the flight target based on the Doppler frequency shift center sequence and the range travel sequence; wherein, the preliminary classification result includes suspected small targets, large aircraft, and invalid targets;

[0047] The feature acquisition module is used to perform empirical mode decomposition on the radar echo signal of the target when the preliminary classification result is the suspected small target, and extract the micro-Doppler component set of the flight target based on the intrinsic mode function obtained by the decomposition; and to perform spectrum analysis on the radar echo signal of the target to obtain the harmonic structure characteristics of the flight target; wherein, the harmonic structure characteristics include the harmonic amplitude ratio and the proportion of harmonic energy;

[0048] The target acquisition module is used to acquire multiple frames of target traces of the flying target based on the target radar echo signal.

[0049] The subdivision module is used to perform trajectory association on the multi-frame target points using the Hungarian algorithm, combined with the micro-Doppler component set and the harmonic structure features, and to determine whether the flight target is a jamming object or a rotary-wing UAV based on the associated trajectory features.

[0050] The technical solution of this application first processes the initial radar echo signal using a least mean square adaptive filtering algorithm to suppress environmental interference and clutter at the source. Then, based on the Doppler frequency shift center sequence and range-travel sequence, it performs preliminary classification of flying targets, acquiring multiple frames of target traces only for suspected small targets. Finally, combining the micro-Doppler component set and harmonic structure features, the Hungarian algorithm is used to complete trajectory association and further refine the target type determination. Specifically, this application employs a hierarchical classification and refined identification processing logic. First, preliminary classification eliminates ineffective targets and large aircraft, reducing invalid data interference and computational redundancy. Then, combining the micro-Doppler component set and harmonic structure features, the Hungarian algorithm is used to perform multi-constraint trajectory association. Based on the stable trajectory features obtained from the association, accurate identification of interfering objects and rotary-wing UAVs is achieved. Even in scenarios with complex target characteristics and severe environmental interference, this application can still ensure rigorous discrimination logic and reliable results, effectively reducing the probability of target misjudgment and missed detection, and comprehensively improving the overall accuracy and stability of ultra-low-altitude flying target detection.

[0051] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent from the following description. Attached Figure Description

[0052] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0053] Figure 1 A flowchart illustrating a method for joint detection of ultra-low-altitude flight using fusion radar, provided in an embodiment of this application;

[0054] Figure 2 A flowchart of another method for joint detection of ultra-low-altitude flight using fusion radar provided in this application embodiment;

[0055] Figure 3 A flowchart illustrating another method for joint detection of ultra-low-altitude flight using fusion radar provided in this application embodiment;

[0056] Figure 4 This is a schematic diagram of a fusion radar-based ultra-low-altitude flight joint detection system provided in an embodiment of this application. Detailed Implementation

[0057] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0058] It should be noted that the terms "target," "initial," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "including," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0059] Figure 1 This is a flowchart illustrating a method for joint detection of ultra-low-altitude flight using fusion radar, provided as an embodiment of this application. This embodiment is applicable to detecting various types of flying targets, and the method can be executed by a joint detection system for ultra-low-altitude flight using fusion radar. Figure 1 As shown, the method includes:

[0060] S101. Obtain the initial radar echo signal in the ultra-low altitude region, and process the initial radar echo signal using the least mean square adaptive filtering algorithm to generate a purified radar echo signal.

[0061] Specifically, the ultra-low altitude region refers to the airspace relatively close to the ground, typically ranging from a few meters to several hundred meters in altitude. The ultra-low altitude region is suitable for the activities of aircraft such as drones and small aircraft. The initial radar echo signal refers to the echo signal received by the radar after it transmits a detection signal, reflected by objects within the airspace. The initial radar echo signal contains information such as the target's distance and motion status. The purified radar echo signal refers to the high-quality, effective signal obtained after filtering to remove noise, interference, and other invalid components from the initial radar echo signal.

[0062] In this embodiment, noise and interference in the initial radar echo signal are removed using a least mean square adaptive filtering algorithm, thus preserving valid information reflecting the flying target. Especially in complex electromagnetic environments at ultra-low altitudes and under strong clutter interference, acquiring a purified radar echo signal with higher signal-to-noise ratio and fidelity provides a reliable data foundation for subsequent target detection, feature extraction, and target classification, avoiding the adverse effects of invalid interference on subsequent processing and ensuring the stability and accuracy of the overall detection method.

[0063] S102. Detect targets in the purified radar echo signal and identify the target radar echo signal corresponding to the flying target in the ultra-low altitude area.

[0064] Specifically, target detection refers to the process of identifying existing flying targets from processed radar echo signals, locating the targets, and then extracting valid radar echo signals that correspond only to the flying targets.

[0065] For example, a constant false alarm rate (CFAR) detection algorithm is used to process the purified radar echo signal to identify the target radar echo signal corresponding to the flying target in the ultra-low altitude region.

[0066] Specifically, the constant false alarm rate (CFAR) detection algorithm can adaptively estimate background clutter for radar target detection and dynamically adjust the detection threshold, effectively distinguishing target echoes from background interference while maintaining a constant false alarm probability. The specific process is as follows: First, the purified radar echo signal is divided into detection units and reference units. By statistically analyzing the signal information of the reference units, the power level of background clutter is adaptively estimated. Then, combined with a preset false alarm probability, a dynamic adaptive detection threshold is generated. The signal amplitude of the detection unit is compared with this threshold to determine whether the target exists. Finally, the detection results are aggregated and filtered to obtain the target radar echo signal of the corresponding flying target.

[0067] In this embodiment, the target radar echo signal can avoid bringing non-target clutter and environmental interference into the subsequent feature calculation and classification process, thereby reducing invalid calculations.

[0068] S103. Obtain the Doppler frequency shift center sequence and range travel sequence of the target radar echo signal, and obtain the preliminary classification results of the flight target based on the Doppler frequency shift center sequence and range travel sequence.

[0069] The preliminary classification results include suspected small targets, large aircraft, and invalid targets.

[0070] Specifically, the Doppler frequency shift center sequence refers to a series of frequency shift center values ​​calculated over a continuous signal observation period based on the Doppler effect of radar signals and the motion state of the flying target. The Doppler effect states that when a radar signal encounters a moving target, the echo signal frequency changes accordingly; the echo signal frequency increases when the target moves towards the radar, and decreases when the target moves away from the radar. The range movement sequence, during radar signal processing, is an information sequence characterizing the dynamic changes in the distance between the flying target and the radar over a continuous observation period, used to reflect the target's range-dimensional motion characteristics.

[0071] In this embodiment, preliminary classification enables the triage of different types of flying targets, ensuring that subsequent processes only involve high-precision, high-complexity feature extraction and identification of suspected small targets, thereby reducing computational redundancy in the detection system. Simultaneously, early elimination of invalid targets and clearly distinguishable large aircraft reduces interference factors in subsequent stages, ensuring the targeted and accurate nature of subsequent detailed analysis, and laying a classification foundation for the extraction of micro-motion and harmonic features and precise target identification.

[0072] S104. If the preliminary classification result is a suspected small target, then perform empirical mode decomposition on the target radar echo signal, extract the micro-Doppler component set of the flight target based on the intrinsic mode function obtained from the decomposition, and perform spectrum analysis on the target radar echo signal to obtain the harmonic structure characteristics of the flight target.

[0073] Among them, the harmonic structure characteristics include the harmonic amplitude ratio and the proportion of harmonic energy.

[0074] Specifically, Empirical Mode Decomposition (EMD) is an adaptive analysis method applicable to nonlinear and non-stationary signals, used to decompose complex signals into components at different frequency scales. After performing EMD on the target radar echo signal, a series of intrinsic mode functions (EMFs) can be obtained. The micro-Doppler component set refers to the set of micro-Doppler frequency shift components generated by the micro-motions of the flight target itself (such as rotor rotation, airframe vibration, etc.), used to characterize the subtle motion characteristics of the target. Spectrum analysis refers to the process of performing frequency domain transformation and analysis on the target radar echo signal to obtain the frequency distribution and amplitude intensity information of the signal in the frequency domain. Harmonic structure characteristics refer to the distribution and proportion characteristics of each harmonic component in the signal frequency domain, specifically including the harmonic amplitude ratio and the proportion of harmonic energy. The harmonic amplitude ratio refers to the ratio of the amplitude of each harmonic component to the fundamental component in the target radar echo signal. For example, the harmonic amplitude ratio of the nth harmonic can be obtained by dividing the amplitude of the nth harmonic by the amplitude of the fundamental wave. The harmonic energy proportion refers to the percentage of each harmonic component in the total effective harmonic energy of the target radar echo signal. For example, the harmonic energy proportion of the m-th harmonic can be obtained by dividing the energy of the m-th harmonic by the total energy of the fundamental wave and all harmonics in the target radar echo signal.

[0075] In this embodiment of the application, for flight targets that are initially classified as suspected small targets, the set of micro-Doppler components characterizing their micro-motion characteristics and the harmonic structure features characterizing their frequency domain structure are extracted in parallel. By combining the two types of complementary features, the inherent attributes of suspected small targets can be fully characterized, making up for the deficiency of the single feature characterization capability, and providing reliable feature support for the subsequent construction of trajectory association constraints and the accurate distinction between rotary-wing UAVs and interference objects.

[0076] S105. Obtain multiple frames of target points from the flying target based on the target radar echo signal.

[0077] Specifically, multi-frame target traces refer to multiple single-frame spatial position records corresponding to the same flying target, calculated based on the target radar echo signal within a continuous multi-frame radar scanning cycle, which are used to subsequently construct and analyze the target's motion trajectory.

[0078] For example, based on the time delay, amplitude, and azimuth information of the target radar echo signal, the target position can be analyzed and the coordinates can be calibrated for each frame of signal within a continuous multi-frame scanning cycle of the radar, generating multiple sets of independent spatial position data corresponding to the same flying target, i.e., the multi-frame target trace of the flying target.

[0079] In this embodiment of the application, the target radar echo signal in the continuous time domain is converted into discretized and framed spatial position data, which can provide a basic geometric position basis for subsequent trajectory association, thereby restoring and analyzing the target's true motion law.

[0080] S106. Using the Hungarian algorithm, combined with the micro-Doppler component set and harmonic structure characteristics, the trajectory of target points in multiple frames is correlated, and the flight target is determined to be a jamming object or a rotary-wing UAV based on the correlated trajectory characteristics.

[0081] Specifically, the Hungarian algorithm is a classic combinatorial optimization algorithm used to solve allocation problems. Interference refers to suspicious flying objects or false targets that are highly similar to rotary-wing UAVs in radar echo characteristics, motion states, and micro-motion performance, easily causing identification confusion. For example, interference can be birds, lightweight suspended debris such as plastic bags and paper scraps carried by ground winds, dense swarms of insects flying at low altitudes, false targets formed by radar sidelobes and multipath effects, and other non-target interference sources that can produce UAV-like micro-motion characteristics and harmonic spectrum features.

[0082] In this embodiment, the association cost matrix of the target point trace can be constructed first based on the micro-Doppler component set and harmonic structure features. Then, the Hungarian algorithm is used for iterative calculation to obtain the globally optimal matching path between target point traces in adjacent frames, thereby realizing accurate trajectory association of target point traces in multiple frames and improving the continuity and reliability of the target trajectory.

[0083] The technical solution of this application first processes the initial radar echo signal using a least mean square adaptive filtering algorithm to suppress environmental interference and clutter at the source. Then, based on the Doppler frequency shift center sequence and range-travel sequence, it performs preliminary classification of flying targets, acquiring multiple frames of target traces only for suspected small targets. Finally, combining the micro-Doppler component set and harmonic structure features, it uses the Hungarian algorithm to complete trajectory association and further refine the target type determination. Specifically, this solution employs a hierarchical classification and refined identification processing logic. First, it eliminates ineffective targets and large aircraft through preliminary classification, reducing invalid data interference and computational redundancy. Then, combining the micro-Doppler component set and harmonic structure features, it performs multi-constraint trajectory association using the Hungarian algorithm. Based on the stable trajectory features obtained from the association, it achieves accurate identification of interfering objects and rotary-wing UAVs. Even in scenarios with complex target features and severe environmental interference, this solution still ensures rigorous discrimination logic and reliable results, effectively reducing the probability of target misjudgment and missed detection, and comprehensively improving the overall accuracy and stability of ultra-low-altitude flying target detection.

[0084] Figure 2 A flowchart illustrating another method for joint detection of ultra-low-altitude flight using fusion radar, provided as an embodiment of this application. Based on the above embodiments, as... Figure 2 As shown, optionally, the method includes:

[0085] S201. Acquire initial radar echo signals in the ultra-low altitude region through a multi-channel sensor array.

[0086] Specifically, a multi-channel sensor array is a type of intelligent sensor, consisting of multiple radar sensing units arranged in a preset spatial configuration. Each sensing unit can synchronously transmit and receive radar signals, possessing the ability of spatial diversity reception and multi-dimensional information perception.

[0087] In this embodiment, the multi-channel sensor array can simultaneously acquire echo data of flying targets at different azimuths and angles, effectively improving the detection coverage and signal-to-noise ratio of complex ultra-low-altitude environments, and providing comprehensive and reliable raw data support for subsequent fusion processing of multi-source detection information and accurate target detection.

[0088] S202. Perform a Fourier transform on the initial radar echo signal to obtain the time spectrum of the target radar echo signal, and determine whether the initial radar echo signal is subject to strong electromagnetic interference based on the time spectrum.

[0089] Specifically, the Fourier transform is a classic signal processing method that converts time-domain signals to the frequency domain for analysis. It can decompose complex time-domain echo signals into a superposition of different frequency components, enabling the mutual conversion between the signal's time-domain and frequency-domain characteristics. The time-spectrum diagram is a two-dimensional graph that simultaneously characterizes the amplitude distribution of a radar signal in both the time and frequency dimensions. The time-spectrum diagram can intuitively reflect the variation of the frequency components of the initial radar echo signal over time, clearly presenting its frequency domain energy distribution characteristics. Strong electromagnetic interference refers to electromagnetic signals generated in ultra-low-altitude detection scenarios by external electromagnetic equipment, civilian communication signals, industrial electromagnetic radiation, etc., with intensity far exceeding that of the target echo and normal background noise. This type of interference can mask effective target echoes, leading to signal distortion and target detection failure.

[0090] Optionally, determining whether the initial radar echo signal is subject to strong electromagnetic interference based on the time-spectrum graph includes: traversing each frequency band and time frame of the time-spectrum graph, extracting the amplitude and phase values ​​of the initial radar echo signal at each time-frequency point. Calculating the amplitude jitter standard deviation of the initial radar echo signal based on the amplitude values; calculating the phase jump angle of the initial radar echo signal based on the phase values. Determining whether the initial radar echo signal is subject to strong electromagnetic interference based on the amplitude jitter standard deviation and the phase jump angle.

[0091] In this embodiment, firstly, the generated time-frequency spectrum is traversed globally, and the amplitude and phase values ​​of each time-frequency component corresponding to each time frame are extracted sequentially to form an amplitude matrix and a phase matrix. Secondly, a preset background noise reference frequency band is selected in the time-frequency spectrum, and the amplitude sequence of multiple consecutive frames within this frequency band is statistically analyzed. The amplitude jitter standard deviation of this sequence is calculated to quantify the degree of random fluctuation in signal amplitude. Then, based on the extracted phase values, the phase difference between adjacent time frames and the same frequency point is calculated. The phase difference is unwrapped to obtain a normalized phase jump angle, which is used to characterize the degree of abrupt change in signal phase. Finally, the calculated amplitude jitter standard deviation and phase jump angle are compared with preset interference judgment thresholds. If the amplitude jitter standard deviation is greater than the amplitude judgment threshold and the phase jump angle is greater than the phase judgment threshold, the initial radar echo signal is determined to be subject to strong electromagnetic interference; otherwise, it is determined not to be subject to strong electromagnetic interference.

[0092] S203. If the initial radar echo signal is subject to strong electromagnetic interference, the initial radar echo signal is processed by the least mean square adaptive filtering algorithm to generate a cleaned radar echo signal; if the initial radar echo signal is not subject to strong electromagnetic interference, the initial radar echo signal is used as the cleaned radar echo signal.

[0093] In the embodiments of this application, a differentiated signal processing strategy is formed based on the discrimination result of strong electromagnetic interference. This strategy can not only achieve adaptive interference suppression for signals subjected to strong interference and ensure signal quality, but also directly retain signals without interference, avoiding excessive filtering that leads to attenuation of the target's effective features, thus maximizing the balance between signal fidelity and interference suppression effect.

[0094] S204. Detect targets in the purified radar echo signal and identify the target radar echo signal corresponding to the flying target in the ultra-low altitude area.

[0095] S205. Perform frequency domain analysis on the target radar echo signal to obtain the Doppler frequency domain peak value. Sort the Doppler frequency domain peak values ​​by frame to obtain the Doppler frequency shift center sequence of the target radar echo signal.

[0096] Specifically, frequency domain analysis refers to a signal processing method that transforms target radar echo signals from the time domain to the frequency dimension for analysis. Frequency domain analysis can extract frequency domain characteristics such as the frequency distribution and amplitude intensity of the echo signal, providing a foundation for calculating Doppler frequency shift related parameters. The Doppler frequency domain peak value refers to the frequency point in the frequency domain curve where the amplitude reaches a local maximum after the target radar echo signal undergoes Doppler frequency domain transformation.

[0097] For example, firstly, for the target radar echo time-domain signal corresponding to each radar detection frame, a Fast Fourier Transform (FFT) is used for frequency domain conversion to obtain the Doppler spectrum data corresponding to that frame. The amplitude of each frame's Doppler spectrum data is traversed, and the amplitude maxima are selected. The frequency value corresponding to these maxima is determined, which is the Doppler frequency domain peak value of the current frame. False frequency domain peak values ​​caused by clutter residue and noise interference are removed, retaining the effective Doppler frequency domain peak values ​​related to the actual flight target motion. According to the chronological order of the radar detection time frames, the effective Doppler frequency domain peak values ​​obtained from each frame are arranged sequentially, ultimately forming the Doppler frequency shift center sequence of the target radar echo signal.

[0098] In this embodiment, the Doppler frequency domain features of a single frame are transformed into a time-series feature sequence. On the one hand, Doppler frequency domain peaks can be extracted through frequency domain analysis to accurately capture the core Doppler motion features of the target echo in each frame, eliminating the influence of secondary components and noise interference in the frequency domain. On the other hand, by sorting the Doppler frequency domain peaks of multiple frames in chronological order, a Doppler frequency shift center sequence is obtained, which is equivalent to transforming discrete single-frame frequency domain features into continuous time-series features. This can intuitively reflect the dynamic change trend of the flight target's motion speed, providing a reliable time-series motion feature basis for subsequent preliminary target classification based on this sequence.

[0099] S206. Based on the principle of radar ranging, the radial distance from the flying target to the multi-channel sensor array is calculated according to the target radar echo signal. The radial distance is sorted by frame to obtain the range movement sequence of the target radar echo signal.

[0100] Specifically, radar ranging principle refers to the process by which a radar system calculates the target distance. For example, using the pulse delay ranging method, the straight-line distance between the flying target and the radar is calculated by recording the time difference between the radar's transmitted signal and the received target echo signal, combined with the propagation speed of electromagnetic waves in space. Radial distance refers to the straight-line projection distance of the flying target relative to the radar observation point where the multi-channel sensor array is located, along the line connecting the two.

[0101] In this embodiment, the range-walk sequence can quantitatively characterize the continuous change of the radial distance of a flying target relative to the radar sensor over time, intuitively reflecting the radial motion trend and motion state differences of the flying target. The range-walk sequence and the Doppler frequency shift center sequence complement each other, providing core motion evidence in the range dimension for the preliminary classification of flying targets, thereby ensuring the effectiveness of the coarse classification results.

[0102] S207. Obtain the radial velocity change trend of the flight target based on the Doppler frequency shift center sequence, and obtain the curvature of the flight target's range trajectory based on the range movement sequence.

[0103] Among them, the radial velocity change trend includes the velocity change rate of the flight target, the standard deviation of velocity fluctuation, and the deviation of the mean velocity; the distance trajectory curvature includes the mean curvature of the flight target's trajectory, the magnitude of curvature change, and the trajectory fitting residual ratio.

[0104] Specifically, the velocity change rate is a quantitative indicator calculated based on the radial velocity time series, using the ratio of the velocity difference between adjacent frames to the time interval. It characterizes the increase or decrease in the radial velocity of a flight target per unit time. The velocity fluctuation standard deviation is the standard deviation value obtained by statistically calculating the radial velocity time series, used to measure the dispersion of the radial velocity in each frame relative to the velocity mean. The velocity mean deviation is the average value calculated by subtracting the radial velocity of each frame from the global velocity mean and taking the absolute value. It visually represents the overall deviation of the velocity sequence from the average level. The trajectory mean curvature is the value obtained by statistically averaging the curvature of each point on the flight target's distance time series trajectory. It describes the degree to which the trajectory curve deviates from a straight line. The curvature change amplitude refers to the difference between the maximum and minimum curvature values ​​at each point on the trajectory, used to characterize the dynamic range of change in the curvature of the distance trajectory. The trajectory fitting residual ratio is a quantitative indicator obtained by first performing least-squares linear fitting on the distance movement sequence, calculating the sum of squared residuals from all trajectory points to the fitted straight line, and then comparing this sum with the sum of squared residuals of the total trajectory energy. It reflects the degree to which the trajectory deviates from the ideal straight line.

[0105] For example, the Doppler frequency shift center sequence is converted into a time-series radial velocity, and the radial velocity change trend is obtained through smoothing and slope analysis; the distance movement sequence is constructed into a distance time-series trajectory, and after polynomial fitting, the degree of curvature of the flight target's distance trajectory is quantitatively characterized by the trajectory point fitting residual or curve curvature.

[0106] S208. Using the radial velocity change trend and the curvature of the distance trajectory as input, a clustering algorithm is used to obtain the preliminary classification results of the flight target.

[0107] Specifically, firstly, the velocity change rate, velocity fluctuation standard deviation, velocity mean deviation, trajectory mean curvature, curvature change amplitude, and trajectory fitting residual ratio are normalized to construct a multi-dimensional feature vector of the target. Secondly, the K-Means clustering algorithm is selected, with the number of cluster centers set to 3, and the parameters for initial cluster centers, maximum number of iterations, and convergence threshold are initialized. Then, the feature vector is input into the model, and cluster division is completed by calculating Euclidean distance, and the cluster centers are iteratively updated until the convergence condition is met. Finally, based on prior knowledge of motion features, cluster type labeling is completed. Clusters with stable motion and characteristics of small low-altitude aircraft are labeled as suspected small targets, clusters with high velocity amplitude and straight and stable trajectories are labeled as large aircraft, and clusters with irregular motion and corresponding noise clutter are labeled as invalid targets.

[0108] S209. If the preliminary classification result is a suspected small target, then perform empirical mode decomposition on the target radar echo signal, extract the micro-Doppler component set of the flight target based on the intrinsic mode function obtained from the decomposition, and perform spectrum analysis on the target radar echo signal to obtain the harmonic structure characteristics of the flight target.

[0109] S210: Obtain multiple frames of target points from the flying target based on the target radar echo signal.

[0110] S211. Using the Hungarian algorithm, combined with the micro-Doppler component set and harmonic structure characteristics, the trajectory of target points in multiple frames is correlated, and the flight target is determined to be a jamming object or a rotary-wing UAV based on the correlated trajectory characteristics.

[0111] The technical solution of this application first determines whether there is strong electromagnetic interference based on the amplitude jitter standard deviation and phase jump angle of the initial radar echo signal's spectrum. Then, it generates a purified radar echo signal by performing least mean square adaptive filtering as needed. Next, a constant false alarm rate (CFAR) detection algorithm is used to identify the target radar echo signal from the purified radar signal. Finally, frequency domain analysis is used to obtain the Doppler frequency shift center sequence and range travel sequence, extracting the radial velocity change trend and range trajectory curvature features. A clustering algorithm is then used to obtain the preliminary classification result of the flying target. This application can effectively improve the purification effect of the initial radar echo signal and the accuracy of target detection, laying a reliable foundation for subsequent accurate discrimination, while also enhancing the anti-interference capability and stability of ultra-low-altitude detection in complex electromagnetic environments.

[0112] Figure 3 A flowchart illustrating another method for joint detection of ultra-low-altitude flight using fusion radar, provided as an embodiment of this application. Based on the above embodiments, as... Figure 3 As shown, optionally, the method includes:

[0113] S301. Obtain the initial radar echo signal in the ultra-low altitude region, process the initial radar echo signal using the least mean square adaptive filtering algorithm, and generate a purified radar echo signal.

[0114] S302. Detect targets in the purified radar echo signal and identify the target radar echo signal corresponding to the flying target in the ultra-low altitude area.

[0115] S303. Acquire the Doppler frequency shift center sequence and range travel sequence of the target radar echo signal, and obtain the preliminary classification results of the flight target based on the Doppler frequency shift center sequence and range travel sequence.

[0116] S304. An adaptive filtering algorithm is used to preprocess the target radar echo signal.

[0117] Specifically, the adaptive filtering algorithm can further suppress residual clutter, noise, and non-micro-motion related interference in the target radar echo signal, highlight the effective echo component caused by the micro-motion of the flying target, improve the signal-to-noise ratio, provide high-quality input data for subsequent empirical mode decomposition, and avoid noise interference causing distortion of decomposition results.

[0118] S305. Perform empirical mode decomposition on the preprocessed target radar echo signal to obtain several eigenmode functions.

[0119] Specifically, each intrinsic mode function independently corresponds to a vibration or micro-motion component at a specific frequency in the target radar echo signal. This can separate the frequency components related to the micro-motion of the flying target in the target radar echo signal (such as signals generated by possible rotor rotation, fuselage shaking, etc.) from the low-frequency trend components and high-frequency noise components in the signal, avoiding mutual interference between different frequency characteristics, thereby accurately preserving and distinguishing the suspected micro-motion related signal components that can be identified later.

[0120] S306. Extract the intra-frame frequency features corresponding to the intrinsic mode functions through Hilbert transform, select the effective frequency components whose intra-frame frequency features conform to the small target category, and sort the effective frequency components from low to high frequency to obtain the micro-Doppler component set.

[0121] Specifically, the Hilbert transform can construct an analytical signal from the intrinsic mode functions, directly calculating the instantaneous frequencies within each frame while preserving the signal's temporal characteristics. Effective frequency components refer to the frequency components extracted from each intrinsic mode function that are related to the target's minute motion. Effective frequency components are mainly used to reflect the frequency characteristics of minute motions such as rotor rotation, airframe vibration, and irregular oscillations.

[0122] In the embodiments of this application, Hilbert transform is performed on each intrinsic mode function to obtain its time-frequency distribution information, and the effective frequency components in each frame are accurately extracted. The micro-motion mode function in the time domain is converted into quantifiable frequency features, which can realize the intuitive characterization and analysis of the target micro-motion frequency components.

[0123] S307. Perform a fast Fourier transform on the preprocessed target radar echo signal to obtain the frequency domain spectrum of the target radar echo signal.

[0124] Specifically, the Fast Fourier Transform (FFT) can transform discrete time-domain target radar echo signals into the frequency domain while significantly reducing computational complexity, achieving rapid mapping of the signal from the time dimension to the frequency dimension. The frequency domain spectrum, generated from the FFT result, clearly presents the dominant frequency, harmonic frequencies, noise floor, and the location and intensity of interference components of the target radar echo signal. Unlike the time-domain spectrum, which includes time information, the frequency domain spectrum focuses solely on the correspondence between frequency and amplitude.

[0125] S308. Calculate the ratio of the fundamental frequency to the harmonic amplitude of each harmonic of the target radar echo signal, as well as the proportion of the harmonic energy of each harmonic, based on the frequency domain spectrum.

[0126] For example, the fundamental frequency and its corresponding amplitude are extracted from the frequency domain spectrum. The harmonic frequencies that are integer multiples of the fundamental frequency are identified in sequence and their corresponding harmonic amplitudes are extracted. The ratio of each harmonic amplitude to the fundamental amplitude is calculated. The proportion of single harmonic energy to total harmonic and fundamental energy is statistically analyzed. The fundamental frequency, each harmonic frequency, the harmonic amplitude ratio and the harmonic energy proportion are integrated to form the harmonic structure characteristics of the flight target.

[0127] S309. Obtain multiple frames of target points from the flying target based on the target radar echo signal.

[0128] S310. Using the Hungarian algorithm, combined with the micro-Doppler component set and harmonic structure features, the trajectory of target points in multiple frames is correlated, and the flight target is determined to be a jamming object or a rotary-wing UAV based on the correlated trajectory features.

[0129] Optionally, frequency features are extracted from the micro-Doppler component set. A multi-feature weighted fusion algorithm is used, with frequency features, harmonic amplitude ratio, and harmonic energy proportion as correlation constraints, to construct a correlation cost matrix between target traces across multiple frames. The optimal matching path for the flight target is then solved using the Hungarian algorithm based on the correlation cost matrix. The trajectory features of the flight target are obtained from the optimal matching path, and the flight target is determined to be either a jamming object or a rotary-wing UAV based on these trajectory features.

[0130] Specifically, the association cost matrix is ​​used to characterize the matching similarity of target traces in adjacent frames. The association cost matrix is ​​a two-dimensional matrix constructed based on target features such as micro-Doppler component sets and harmonic structure characteristics, calculating the feature differences and positional deviations between the target to be matched in the current frame and targets in historical frames. Each element in the association cost matrix represents the association cost between the corresponding two targets; the smaller the cost value, the higher the matching similarity between the targets. The Hungarian algorithm is used to achieve trajectory association matching between radar-detected targets across multiple frames, achieving globally optimal matching results within polynomial time complexity. The optimal matching path refers to the target matching sequence obtained by the Hungarian algorithm that minimizes the total global association cost.

[0131] The process of obtaining the trajectory features of the flight target based on the optimal matching path and determining whether the flight target is a jamming object or a rotary-wing UAV based on the trajectory features includes: obtaining the stable trajectory of the flight target based on the optimal matching path and extracting trajectory features based on the stable trajectory; wherein, the trajectory features include trajectory curvature, motion speed and trajectory continuity; if the trajectory curvature is within a preset rotary-wing UAV curvature threshold range, the motion speed is within a preset rotary-wing UAV speed threshold range, and the trajectory continuity is within a preset rotary-wing UAV trajectory continuity threshold range, then the flight target is determined to be a rotary-wing UAV; otherwise, the flight target is determined to be a jamming object.

[0132] Specifically, based on the optimal matching path, the detection points that match the same target in multiple consecutive frames are sequentially associated in chronological order to form an initial motion trajectory. After outlier removal and smoothing, a stable trajectory can be obtained. Based on the temporal position data of this stable trajectory, the trajectory curvature, motion speed, and trajectory continuity of the flying target can be further calculated.

[0133] Furthermore, the threshold ranges for the curvature, trajectory continuity, and speed of rotary-wing UAVs were primarily determined through prior experiments and sample statistics. First, measured trajectory data of rotary-wing UAVs and various interfering objects were collected in advance. Statistical distribution analysis was performed on the trajectory curvature, speed, and trajectory continuity characteristics of the rotary-wing UAVs. The characteristic concentration intervals were extracted and optimized through verification experiments to obtain the corresponding threshold ranges that can effectively distinguish between UAVs and interfering objects.

[0134] The technical solution of this application, targeting suspected small targets, preprocesses their radar echo signals, extracts micro-Doppler component sets and harmonic features, and then performs multi-feature fusion, trajectory correlation, and threshold discrimination to ultimately achieve accurate differentiation between rotary-wing UAVs and interference objects. This application can significantly improve the anti-interference capability, target recognition reliability and accuracy of ultra-low altitude detection systems, effectively reduce their false positive rate, and adapt to complex ultra-low altitude detection scenarios.

[0135] Figure 4 This is a schematic diagram of a fusion radar-based ultra-low-altitude flight joint detection system provided in an embodiment of this application. Figure 4 As shown, the system includes:

[0136] The signal acquisition module 410 is used to acquire the initial radar echo signal in the ultra-low altitude region, and process the initial radar echo signal through the least mean square adaptive filtering algorithm to generate a purified radar echo signal.

[0137] The target detection module 420 is used to detect targets in the purified radar echo signal and identify the target radar echo signal corresponding to the flying target in the ultra-low altitude area.

[0138] The initial screening module 430 is used to acquire the Doppler frequency shift center sequence and range movement sequence of the target radar echo signal, and to obtain the preliminary classification results of the flight target based on the Doppler frequency shift center sequence and range movement sequence; among which, the preliminary classification results include suspected small targets, large aircraft and invalid targets.

[0139] The feature acquisition module 440 is used to perform empirical mode decomposition on the target radar echo signal when the preliminary classification result is suspected to be a small target, extract the set of micro-Doppler components of the flight target based on the intrinsic mode function obtained by decomposition, and perform spectrum analysis on the target radar echo signal to obtain the harmonic structure characteristics of the flight target; wherein, the harmonic structure characteristics include the harmonic amplitude ratio and the proportion of harmonic energy.

[0140] The target acquisition module 450 is used to acquire multiple frames of target targets based on the target radar echo signal.

[0141] The subdivision module 460 is used to perform trajectory association on target points in multiple frames using the Hungarian algorithm, combined with the micro-Doppler component set and harmonic structure features, and to determine whether the flight target is a jamming object or a rotary-wing UAV based on the associated trajectory features.

[0142] The ultra-low-altitude flight joint detection system of fusion radar provided in this application embodiment can execute the ultra-low-altitude flight joint detection method of fusion radar provided in any embodiment of this application, and has the corresponding functional modules and beneficial effects of the execution method.

Claims

1. A method for joint detection of ultra-low-altitude flight using fused radar, characterized in that, include: The initial radar echo signal in the ultra-low altitude region is acquired, and the initial radar echo signal is processed by the least mean square adaptive filtering algorithm to generate a purified radar echo signal. Target detection is performed on the purified radar echo signal to identify the target radar echo signal corresponding to the flying target in the ultra-low altitude area; The Doppler frequency shift center sequence and range travel sequence of the target radar echo signal are acquired, and a preliminary classification result of the flight target is obtained based on the Doppler frequency shift center sequence and the range travel sequence; wherein, the preliminary classification result includes suspected small targets, large aircraft, and invalid targets; If the preliminary classification result is the suspected small target, then empirical mode decomposition is performed on the target radar echo signal, and the micro-Doppler component set of the flight target is extracted based on the intrinsic mode function obtained from the decomposition; and spectrum analysis is performed on the target radar echo signal to obtain the harmonic structure characteristics of the flight target; wherein, the harmonic structure characteristics include the harmonic amplitude ratio and the proportion of harmonic energy; The target radar echo signal is used to obtain multiple frames of target points of the flying target. Using the Hungarian algorithm, combined with the micro-Doppler component set and the harmonic structure features, the trajectory of the multi-frame target points is correlated, and the flight target is determined to be either a jamming object or a rotary-wing UAV based on the correlated trajectory features.

2. The method for joint detection of ultra-low-altitude flight using fusion radar according to claim 1, characterized in that, The process of acquiring the initial radar echo signal in the ultra-low altitude region and processing it using a least mean square adaptive filtering algorithm to generate a purified radar echo signal includes: Initial radar echo signals in the ultra-low altitude region are acquired using a multi-channel sensor array. Perform a Fourier transform on the initial radar echo signal to obtain the time spectrum of the target radar echo signal, and determine whether the initial radar echo signal is subject to strong electromagnetic interference based on the time spectrum. If the initial radar echo signal is subject to strong electromagnetic interference, the initial radar echo signal is processed by the least mean square adaptive filtering algorithm to generate the purified radar echo signal; if the initial radar echo signal is not subject to strong electromagnetic interference, the initial radar echo signal is used as the purified radar echo signal.

3. The method for joint detection of ultra-low-altitude flight using fusion radar according to claim 2, characterized in that, The step of determining whether the initial radar echo signal is subject to strong electromagnetic interference based on the time-spectrum diagram includes: Traverse each frequency band and time frame of the time-frequency spectrum to extract the amplitude and phase values ​​of the initial radar echo signal at each time-frequency point; The amplitude jitter standard deviation of the initial radar echo signal is calculated based on the amplitude value; the phase jump angle of the initial radar echo signal is calculated based on the phase value. The initial radar echo signal is determined to be subject to strong electromagnetic interference based on the amplitude jitter standard deviation and the phase jump angle.

4. The method for joint detection of ultra-low-altitude flight using fusion radar according to claim 1, characterized in that, The step of performing target detection on the purified radar echo signal to identify the target radar echo signal corresponding to the flying target in the ultra-low altitude region includes: The constant false alarm rate (CFAR) detection algorithm is used to process the purified radar echo signal to identify the target radar echo signal corresponding to the flying target in the ultra-low altitude region.

5. The method for joint detection of ultra-low-altitude flight using fusion radar according to claim 2, characterized in that, The process of acquiring the Doppler frequency shift center sequence and range travel sequence of the target radar echo signal, and obtaining the preliminary classification result of the flight target based on the Doppler frequency shift center sequence and the range travel sequence, includes: Frequency domain analysis is performed on the target radar echo signal to obtain the Doppler frequency domain peak value. The Doppler frequency domain peak value is sorted by frame to obtain the Doppler frequency shift center sequence of the target radar echo signal. Based on the principle of radar ranging, the radial distance from the flying target to the multi-channel sensor array is calculated according to the target radar echo signal. The radial distance is sorted by frame to obtain the range movement sequence of the target radar echo signal. The radial velocity variation trend of the flight target is obtained based on the Doppler frequency shift center sequence, and the distance trajectory curvature of the flight target is obtained based on the distance movement sequence; wherein, the radial velocity variation trend includes the velocity change rate, velocity fluctuation standard deviation, and velocity mean deviation of the flight target, and the distance trajectory curvature includes the average trajectory curvature, curvature change amplitude, and trajectory fitting residual ratio of the flight target; Using the radial velocity change trend and the distance trajectory curvature as inputs, a clustering algorithm is used to obtain the preliminary classification results of the flight target.

6. The method for joint detection of ultra-low-altitude flight using fusion radar according to claim 1, characterized in that, If the preliminary classification result is the suspected small target, then empirical mode decomposition is performed on the target radar echo signal, and the micro-Doppler component set of the flying target is extracted based on the intrinsic mode function obtained from the decomposition, including: An adaptive filtering algorithm is used to preprocess the target radar echo signal; Empirical mode decomposition is performed on the preprocessed target radar echo signal to obtain several eigenmode functions; The intra-frame frequency features corresponding to the intrinsic mode functions are extracted by Hilbert transform, and the effective frequency components that conform to the small target category are selected. The effective frequency components are sorted from low to high frequency to obtain the micro-Doppler component set.

7. The method for joint detection of ultra-low-altitude flight using fusion radar according to claim 6, characterized in that, The step of performing spectral analysis on the target radar echo signal to obtain the harmonic structure characteristics of the flying target includes: A fast Fourier transform is performed on the preprocessed target radar echo signal to obtain the frequency domain spectrum of the target radar echo signal. The fundamental frequency of the target radar echo signal is calculated based on the frequency domain spectrum, along with the harmonic amplitude ratio of each harmonic and the harmonic energy percentage of each harmonic.

8. The method for joint detection of ultra-low-altitude flight using fusion radar according to claim 1, characterized in that, The process involves using the Hungarian algorithm, combined with the micro-Doppler component set and the harmonic structure features, to perform trajectory association on the multi-frame target points. Based on the associated trajectory features, the flight target is determined to be either a jamming object or a rotary-wing UAV, including: Extract the frequency features from the set of microDoppler components; Using a multi-feature weighted fusion algorithm, the frequency features, the harmonic amplitude ratio, and the harmonic energy ratio are used as association constraints to construct the association cost matrix between the target points in the multiple frames. The optimal matching path for the flight target is solved using the Hungarian algorithm based on the correlation cost matrix. The trajectory characteristics of the flight target are obtained based on the optimal matching path, and the flight target is determined to be a jamming object or a rotary-wing UAV based on the trajectory characteristics.

9. The method for joint detection of ultra-low-altitude flight using fusion radar according to claim 8, characterized in that, The step of obtaining the trajectory characteristics of the flight target based on the optimal matching path, and determining whether the flight target is a jamming object or a rotary-wing UAV based on the trajectory characteristics, includes: The stable trajectory of the flight target is obtained according to the optimal matching path, and trajectory features are extracted based on the stable trajectory; wherein, the trajectory features include trajectory curvature, motion speed and trajectory continuity; If the trajectory curvature is within a preset range of rotary-wing UAV curvature thresholds, the speed is within a preset range of rotary-wing UAV speed thresholds, and the trajectory continuity is within a preset range of rotary-wing UAV trajectory continuity thresholds, then the flying target is determined to be a rotary-wing UAV; otherwise, the flying target is determined to be an interference object.

10. A combined radar and ultra-low-altitude flight detection system, characterized in that, include: The signal acquisition module is used to acquire the initial radar echo signal in the ultra-low altitude region, and process the initial radar echo signal through the least mean square adaptive filtering algorithm to generate a purified radar echo signal. The target detection module is used to detect targets in the purified radar echo signal and identify the target radar echo signal corresponding to the flying target in the ultra-low altitude area. The initial screening module is used to acquire the Doppler frequency shift center sequence and range movement sequence of the target radar echo signal, and to obtain the preliminary classification result of the flight target based on the Doppler frequency shift center sequence and the range movement sequence; wherein, the preliminary classification result includes suspected small targets, large aircraft, and invalid targets; The feature acquisition module is used to perform empirical mode decomposition on the radar echo signal of the target when the preliminary classification result is the suspected small target, and extract the micro-Doppler component set of the flight target based on the intrinsic mode function obtained by decomposition; and to perform spectrum analysis on the radar echo signal of the target to obtain the harmonic structure characteristics of the flight target; wherein, the harmonic structure characteristics include the harmonic amplitude ratio and the proportion of harmonic energy; The target acquisition module is used to acquire multiple frames of target traces of the flying target based on the target radar echo signal. The subdivision module is used to perform trajectory association on the multi-frame target points using the Hungarian algorithm, combined with the micro-Doppler component set and the harmonic structure features, and to determine whether the flight target is a jamming object or a rotary-wing UAV based on the associated trajectory features.

Images