A method and system for coherent accumulation of high-speed cross-beam targets of a general exploration radar

Abstract

The application discloses a kind of pan-probing radar high-speed cross-beam target coherent accumulation method and system, belong to radar signal processing field;Firstly, phase compensation factor is constructed using beam pointing information, and the phase jump caused by beam switching is eliminated;Then end offset mode and normalized beam domain distance are introduced, candidate cross-beam trajectory is generated, each beam echo segment is extracted and spliced into virtual long observation signal;Then, by center slow time flip and second Keystone transformation, first-order distance migration and second-order distance curvature are sequentially eliminated, and distance migration decoupling is realized;Finally, phase compensation and coherent accumulation are carried out in acceleration domain, target detection and parameter estimation are completed;The application does not need multi-dimensional joint search, effectively realizes cross-beam phase continuous splicing and multi-dimensional multi-scale migration collaborative compensation under low computing power condition, significantly improves the long-time coherent accumulation gain of high-speed weak target, and is suitable for space-based, airborne and other pan-probing radar platforms.

Description

Technical Field

[0001] This invention relates to the field of radar signal processing and target detection technology, and in particular to a method and system for cross-beam coherent accumulation and target detection in a generalized radar system, targeting cross-beam, cross-range, and cross-Doppler phenomena generated during long-term observation of high-speed maneuvering targets. Background Technology

[0002] In radar detection applications targeting high-speed, weak targets, the physical bottleneck of radar power-aperture product necessitates extending the accumulation time to improve the echo signal-to-noise ratio, which has become an important technical approach to enhance radar system detection performance. Traditional phased array radars typically employ a narrow-beam scanning system, resulting in a limited dwell time for targets within a single beam, making it difficult to meet the requirements for long-term observation and long-term coherent accumulation of high-tangential-velocity targets. Digital array-based general-purpose radars typically employ a low-gain, wide-beam transmission and multi-beam reception system, enabling continuous reception of target echoes over a larger airspace coverage area, thus providing conditions for long-term observation of high-speed, weak targets. Simultaneously, during long-term coherent accumulation, the motion of high-speed targets often induces a "three-span" effect across range, Doppler, and beam dimensions, causing the target echo to spread and migrate in the range, Doppler, and beam dimensions, making it difficult to effectively superimpose the echo energy under a unified phase reference. For this type of problem, traditional long-time accumulation methods such as Radon Fourier transform, Keystone transform, and fractional Fourier transform are mainly aimed at range travel and Doppler spread compensation under single-beam or fixed observation channel conditions. In the scenario of simultaneous multi-beam reception by a panoptic radar, the consideration of phase changes caused by beam migration and beam switching is relatively insufficient.

[0003] In existing technologies, one approach is to suppress cross-range and cross-Doppler cell effects by compensating for target motion parameters within the traditional long-time accumulation framework. For example, Li et al. proposed a coherent detection and parameter estimation method based on windowed Radon fractional Fourier transform (WRFRFT). This method primarily addresses situations where the target's entry and exit times are unknown. By setting a time window and searching along the target's possible trajectory, it extracts effective echoes from the original observation data and then accumulates energy in the fractional Fourier transform domain to complete target detection and motion parameter estimation. This type of method can achieve good long-time accumulation under unknown entry and exit times and can also address cross-range cell and Doppler spread issues. However, it mainly targets general radar observation scenarios and is not specifically designed for cross-beam migration in multi-beam receiving systems of general radar systems. Therefore, it is difficult to directly solve the phase unification and cross-beam energy convergence problems between multiple beams.

[0004] Another approach focuses directly on multi-beam scenarios. Rao et al., addressing the problem of weak target cross-beam detection, established a three-dimensional temporal model including fast time, slow time, and beam time, along with a corresponding three-dimensional signal model. They proposed multi-beam correlation coherent accumulation algorithms for both time-division shared multi-beam and space-division shared multi-beam modes. The basic idea of ​​this approach is to establish correlations between multiple beams, uniformly process the echoes of the target in different beams, and then use frequency domain accumulation to achieve weak target detection. This research shows that by establishing correlations in the beam dimension, the energy dispersion problem caused by beam migration can be alleviated to some extent, thereby achieving cross-beam coherent accumulation.

[0005] To address the situation where a target traverses multiple beams within the coherent processing time and the entry and exit times of the target in each beam are unknown, Hu et al. proposed a multi-beam target accumulation and detection method combining multi-scale sliding window phase difference and spatial projection. This method first constructs a multi-beam phase compensation function to compensate for phase changes caused by the target's cross-beam motion; then, it uses a second-order Keystone transform to correct range travel; next, it estimates the target's entry and exit times and positions in the beams using multi-scale sliding window phase difference; finally, it employs spatial projection to achieve joint accumulation and detection of multi-beam echoes. This scheme systematically considers the target detection problem under cross-beam conditions and has certain applicability when the entry and exit times of the beams are unknown. However, its processing flow involves multiple steps such as sliding window analysis, phase difference estimation, and spatial projection, making the overall implementation process relatively complex.

[0006] However, existing technologies still have the following shortcomings. First, traditional long-term accumulation methods mainly focus on compensation for range and Doppler units under single-beam systems, which are not well-suited to the cross-beam coherence problem in multi-beam receiving scenarios of generalized radar. Second, although some multi-beam accumulation methods consider beam shift, they often rely on prior information such as the target's entry or exit time from the beam, initial beam position, target trajectory parameters, or beam domain rotation parameters, or require processing through sliding windows, spatial projection, or multi-parameter joint search, resulting in high algorithm complexity. Third, for high-speed targets, the effects of range, Doppler, and beam units often coexist and are coupled. If they cannot be compensated collaboratively within a unified framework, echo energy defocusing is still likely to occur, making it difficult to obtain sufficient coherent accumulation gain. Especially in generalized radar scenarios, due to the unknown tangential motion of the target, the exact time of the target entering and exiting each receiving beam is difficult to determine in advance, and beam pointing changes also introduce additional phase changes, making it difficult for existing technologies to balance detection performance and computational complexity in engineering applications.

[0007] Therefore, for high-speed cross-beam targets in general-purpose radar, it is still necessary to propose a coherent accumulation method that can take into account cross-beam phase uniformity, range-Doppler motion compensation, and effective cross-beam energy convergence under the condition of limited prior information, so as to improve the long-term coherent accumulation gain and detection capability of high-speed weak targets.

[0008] In summary, existing coherent accumulation methods have the following drawbacks when dealing with high-speed cross-beam targets in generalized radar:

[0009] 1. Unacceptable beam switching phase modulation: Traditional multi-beam processing methods ignore the phase jump introduced by the sudden change in beam pointing when the target moves through different beams, which makes it impossible for the signal to achieve coherent alignment between pulses and causes cross-beam coherent accumulation to fail.

[0010] 2. Multidimensional coupling leads to an explosion in computational load: If the generalized Radon Fourier transform (GRFT) is used for the joint search of range, velocity, acceleration and beam trajectory, the computational load of searching the multidimensional parameter space increases exponentially, making it difficult to implement in real time on radar platforms with limited computing power. Summary of the Invention

[0011] In view of this, the purpose of this invention is to provide a method and system for high-speed cross-beam target coherent accumulation in a probe radar, aiming to eliminate the coupling of phase jump and range migration caused by beam switching in a probe radar with low computing power, and to achieve efficient accumulation and robust detection of high-speed cross-beam target energy.

[0012] A method for high-speed cross-beam coherent accumulation of targets using a general-purpose radar includes the following steps: The first step is to preprocess the received echo signal and construct a beam phase compensation factor using the beam pointing parameter to perform beam phase compensation on the echo signal and eliminate phase jumps. The second step, for scenarios where the target spans multiple receiving beams, defines the end offset mode by introducing the start-end offset factor and the end-end offset factor, constructs a two-dimensional candidate beam domain trajectory, calculates the normalized beam domain distance between the two-dimensional candidate beam domain trajectory and each receiving beam under each end offset mode, and then extracts the receiving beam corresponding to the minimum distance pulse by pulse from the preprocessed multi-beam data to construct a cross-beam splicing signal, and realigns the energy of the target scattered in different beams in the time domain to form a single virtual long-term observation signal; The third step involves constructing a symmetrical slow-time variable based on the observation center time, using the cross-beam splicing signal to construct a time-varying reversal signal, and then performing decoupling multiplication on the cross-beam splicing signal to cancel the first distance travel, and then eliminating the second distance curvature coupling term to obtain the focused signal. The fourth step involves constructing an acceleration search grid for the focused signal, constructing a phase compensation factor for each trial acceleration and performing a fast Fourier transform to achieve coherent accumulation between pulses, obtaining a range-Doppler two-dimensional plane output, and extracting the effective target and its motion parameters through constant false alarm rate detection.

[0013] Preferably, the first step specifically includes: The linear frequency modulated signal transmitted by the radar is expressed as a complex signal, as follows: (1) in, It is a rectangular envelope function. The duration of the pulse. For frequency modulation slope, For signal bandwidth, For carrier frequency, For fast time variables, For slow time variables, The pulse sequence number. The number of pulses. The pulse repetition period; If the system generates The receiving digital beam, ignoring amplitude and noise terms, after down-conversion and fast-time matched filtering, the... The beam at the 1st The baseband pulse compression echo signal under one pulse is represented as: (2) in, Represents the speed of light. The radial distance between the target and the radar at different slow time points. These are the initial distance, initial velocity, and initial acceleration at the start of the observation; , The first The beam at the 1st The amplitude and phase responses formed under each pulse; Constructing compensation factors using beam pointing parameters: (3) in, For wavelength, The first The pointing center of each beam is oriented along the azimuth cosines in the x and y directions. These represent the element spacing in the x and y directions, respectively. The aforementioned compensation factor is used to compensate the baseband pulse-compression echo signal, eliminating phase jumps caused by beam switching, and transforming the signal to the fast time-frequency domain. The preprocessed signal is obtained: (4) in, This represents the Fourier transform along the fast time.

[0014] Preferably, in the second step, the starting bias factor and termination bias factor The end bias mode constitutes the above ,Pick , , and These correspond to four cross-beam crossing modes: all-in / all-out, all-in / half-out, half-in / all-out, and half-in / half-out.

[0015] In the second step, the method for calculating the normalized beam domain distance includes: For any starting beam Termination Beam and end bias mode Construct normalized trajectory parameters: (7) This yields the two-dimensional candidate beam domain trajectories of the start and end beam pairs in each end offset mode: (8) in, , These represent the two-dimensional cosine vectors pointing towards the center for the starting and ending beams, respectively. Define the normalized beam domain distance as: (9) in, The azimuth cosine vector of the beam. and Corresponding to and The azimuth cosine of the direction. and These respectively represent the received beam at direction and Half-power beamwidth in the direction; Substituting formula (8) into formula (9), the left side of the equation in formula (9) becomes equal to the left side of the equation. Replace with The right side of the equals sign Replace with , Replace with The normalized beam domain distance between the two-dimensional candidate beam domain trajectory and the received beam is obtained under each end offset mode.

[0016] Preferably, the third step specifically includes: Introducing observation center time Symmetric slow-time variables ; Constructing a time-varying flip signal: (12) Multiplying the spliced ​​signal by the inverted signal yields the decoupled product signal: (13) Subsequently, a second-order Keystone transform is used to eliminate secondary coupling, and the scaling relationship is defined as follows: (14) in, For the new slow-time variable; The decoupled product signal is scaled and interpolated along a slow time path, and then subjected to a fast-time inverse Fourier transform to obtain the focused signal. (15); in, This represents the inverse Fourier transform.

[0017] Preferably, the fourth step specifically includes: Construct an acceleration search grid set, and test each acceleration in the set. Construct phase compensation factor And perform a fast Fourier transform to achieve coherent accumulation between pulses, and obtain the output of the range-Doppler two-dimensional plane: (16) in, For Doppler frequency variables.

[0018] A high-speed cross-beam target coherent accumulation system for a general-purpose radar, used to implement the above method, includes: The signal preprocessing and phase compensation module is used to perform down-conversion, matched filtering, and beam phase jump compensation on the echo signal received by the array. The cross-beam trajectory generation and signal stitching module is used to extract the corresponding beam signals based on the assumed beam trajectory and stitch them into a long observation signal. The distance migration decoupling and focusing module is used to eliminate first-order and second-order distance migration by using slow-time flipping and second-order Keystone transformation; The coherent accumulation and target detection module is used to perform acceleration compensation, Doppler analysis, and constant false alarm rate detection on the focused signal.

[0019] The present invention has the following beneficial effects: 1. Achieve phase continuity stitching of cross-beam echoes and improve long-time coherent accumulation capability.

[0020] This invention addresses the problem of targets spanning multiple receiving beams in multi-beam radar scenarios. It utilizes beam pointing information to construct a beam phase compensation factor and, combined with an assumed beam trajectory, extracts echo segments of the target in different beams, recombining the echoes originally dispersed across multiple beams into a single, phase-continuous virtual long observation signal. This effectively suppresses phase inconsistencies caused by beam switching or beam migration, improving the coherent accumulation gain for targets spanning multiple beams.

[0021] 2. Achieve coordinated compensation for cross-beam, cross-distance, and cross-Doppler effects.

[0022] This invention does not address range migration or Doppler spread in isolation. Instead, based on cross-beam trajectory matching, it further utilizes center slow-time flip and second-order Keystone transform to compensate for first-order range migration and second-order range curvature caused by the radial motion of the target. This allows it to address cross-beam, cross-range cell, and cross-Doppler cell issues within the same processing framework, reducing energy defocusing caused by multidimensional coupling.

[0023] 3. Reduces the computational complexity of multidimensional joint search, making it suitable for platforms with limited computing power.

[0024] If a joint search for range, velocity, acceleration, and two-dimensional beam trajectory is directly performed using methods such as the generalized Radon Fourier transform, the computational complexity increases rapidly with the parameter dimension. This invention first stitches the beam trajectory, then uses center-time flipping to eliminate one-stage range travel, and employs a second-order Keystone transform to compensate for secondary range curvature. This eliminates the need for large-scale joint searches of some motion parameters, thereby reducing algorithm complexity and making it more suitable for space-based, airborne, and other radar platforms with limited power consumption and computing power.

[0025] 4. Applicable to two-dimensional area array radar and high-speed maneuvering target scenarios.

[0026] Existing multi-beam accumulation methods mainly focus on modeling one-dimensional beam sequences or uniformly moving targets. This invention is designed for two-dimensional digital array general-purpose radar, considering the target's migration in the azimuth and elevation two-dimensional beam domains as well as its higher-order radial motion states. This is more in line with the long-term observation requirements of high-speed, weak targets in practical engineering scenarios such as space-based radar and airborne radar. Attached Figure Description

[0027] Figure 1 This is a schematic diagram illustrating the multi-dimensional cross-section of the objectives of this invention.

[0028] Figure 2 This is a flowchart of the algorithm of the present invention.

[0029] Figure 3 This represents the trajectory of the target between the receiving beams during the implementation of the method of the present invention.

[0030] Figure 4 This shows the changes in the target's fast and slow time echoes during the implementation of the method of this invention.

[0031] Figure 5 This demonstrates the signal change process during the implementation of the method of the present invention. Detailed Implementation

[0032] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0033] The purpose of this invention is to provide a method for coherent accumulation and detection of high-speed cross-beam targets in a panoptic radar, which solves the problems of echo energy defocusing and detection failure caused by cross-beam, cross-range, and cross-Doppler elements when panoptic radar observes high-speed maneuvering targets for a long time.

[0034] like Figure 1 As shown, when a general-purpose radar observes a high-speed target, the target echo may simultaneously undergo range dimension crossing and two-dimensional beam domain crossing within the coherent accumulation time. Figure 1 The spatial coordinate system in the radar is used to represent the three-dimensional observation geometry of the target relative to the radar. The positional changes of the target at the start and end of the observation cause it to span multiple range cells in the range dimension and multiple receiving beamwidths in the two-dimensional beam domain. Figure 1 The range traversal diagram on the right illustrates the migration of the target echo peak between range cells over slow time, while the beam traversal diagram illustrates the migration of the target angular position from one receiving beam region to an adjacent receiving beam region over slow time. Therefore, if long-term coherent accumulation is performed directly within a fixed receiving beam, the target energy will diffuse simultaneously in both the range and beam dimensions. This invention achieves energy focusing for this type of multidimensional traversing target through beam phase compensation, cross-beam trajectory generation, signal stitching, range migration decoupling, and coherent accumulation.

[0035] This invention provides a method and system for high-speed cross-beam target coherent accumulation and detection using a general-purpose radar. The system includes: a signal preprocessing and phase compensation module, a cross-beam trajectory generation and signal stitching module, a range migration decoupling and focusing module, and a coherent accumulation and target detection module. Specifically, the signal preprocessing and phase compensation module performs down-conversion, matched filtering, and intra-beam phase jump compensation on the echo received by the array; the cross-beam trajectory generation and signal stitching module extracts the corresponding beam signal based on the spatially assumed trajectory and stitches it into a long observation signal; the range migration decoupling and focusing module eliminates first-order and second-order range migration using slow-time flipping and second-order Keystone transform; and the coherent accumulation and target detection module performs acceleration compensation, Doppler analysis, and constant false alarm rate (CFAR) detection on the focused signal.

[0036] This invention provides a method for high-speed cross-beam target coherent accumulation and detection using a general-purpose radar, specifically comprising: The first step, the signal preprocessing and phase compensation module, performs preprocessing and beam phase compensation on the echo signal received by the array: The linear frequency modulated signal transmitted by the radar is expressed as a complex signal, as follows: (1) in, It is a rectangular envelope function. The duration of the pulse. For frequency modulation slope, For signal bandwidth, For carrier frequency, For fast time variables, For slow time variables, The pulse sequence number. The number of pulses. This is the pulse repetition period.

[0037] If the system generates The receiving digital beam, ignoring amplitude and noise terms, after down-conversion and fast-time matched filtering, the... The beam at the 1st The baseband pulse compression echo signal under one pulse is represented as: (2) in, Represents the speed of light. The radial distance between the target and the radar at different slow time points. These represent the initial distance, initial velocity, and initial acceleration at the start of the observation. , The first The beam at the 1st The amplitude and phase responses formed under each pulse.

[0038] Since dynamic switching of beam pointing introduces spatial phase modulation into the echo, the signal preprocessing and phase compensation module uses beam pointing parameters to construct a compensation factor: (3) in, For wavelength, The first The pointing center of each beam is oriented along the azimuth cosines in the x and y directions. These represent the element spacing in the x and y directions, respectively.

[0039] This factor is used to compensate for the original echo, eliminate phase jumps caused by beam switching, and transform the signal to a fast time-frequency domain. The preprocessed signal is obtained: (4) in, This represents the Fourier transform along the fast time.

[0040] The second step involves the cross-beam trajectory generation and signal stitching module recombining the energy fragments of the target spanning multiple beams. Let the receiving beam set be Then the first The two-dimensional cosine vector pointing towards the center of each receiving beam is: (5) Considering that the times when high-speed targets enter the starting beam and leave the ending beam are usually unknown, this invention introduces a starting-end offset factor. and termination bias factor This is used to describe the different dwelling truncation modes of the target at both ends of the observation aperture; the set of end offset modes is defined as: (6) Among them, the end bias mode Pick , , and These correspond to four cross-beam traversal modes: all-in / all-out, all-in / half-out, half-in / all-out, and half-in / half-out. An offset factor of 0 indicates that the corresponding end has a complete beam dwell section; a value of... This indicates that the corresponding end has only half of the equivalent beam dwell segment.

[0041] For any starting beam Termination Beam and end bias mode Construct normalized trajectory parameters: (7) This yields the two-dimensional candidate beam domain trajectories of the start and end beam pairs in each end offset mode: (8) in, , These represent the two-dimensional cosine vectors pointing towards the center for the starting and ending beams, respectively. To accommodate situations where the azimuth and elevation beamwidths may differ in a two-dimensional array, the normalized beam domain distance is defined as: (9) in, The azimuth cosine vector of the beam. and Corresponding to and The azimuth cosine of the direction. and These respectively represent the received beam at direction and Half-power beamwidth in the direction.

[0042] Based on the nearest neighbor beam association criterion and the definition of formula (9), the two-dimensional candidate beam domain trajectory and the received beam set under each end offset mode shown in formula (8) are calculated. The normalized beam domain distance for each beam n is calculated; the beam number corresponding to the minimum distance is obtained, and this number is used as the beam number of the nth beam. The candidate trajectory in the 1st The receiving beam number corresponding to each pulse : (10) When substituting formula (8) into formula (9), the left side of the equation in formula (9) will change. Replace with The right side of the equals sign Replace with , Replace with .

[0043] Based on candidate trajectory number A given receive beam number allows you to determine its corresponding quadruple. ; For the The candidate beam trajectory generation and signal stitching module extracts the fast time-frequency domain echo of the corresponding beam pulse by pulse from the preprocessed multi-beam data to construct the cross-beam stitched signal: (11) in, Indicates the received beam The preprocessed signal after phase jump compensation and conversion to the fast time domain is calculated according to formula (4).

[0044] This operation realigns the energy of the target dispersed in different beams in the time domain, forming a single virtual long-term observation signal.

[0045] Step 3: Distance migration decoupling and focusing module decouples the fast-time frequency and slow-time of the spliced ​​signal. To decouple the fast time-frequency and velocity coupling, a method based on the observation center time is introduced. Symmetric slow-time variables In a symmetric slow-time coordinate system, the target motion model is equivalent to... ,in, , These represent the distance at the center time and the radial velocity at the center time, respectively. Constructing a time-varying flip signal: (12) Multiplying the spliced ​​signal by the inverted signal yields the decoupled product signal: (13) By taking advantage of the property that the phase caused by velocity is an odd function of time, the multiplication operation can completely cancel out the first distance travel, leaving only the second distance bending coupling term caused by acceleration in the signal.

[0046] Subsequently, a second-order Keystone transform (SOKT) is used to eliminate secondary coupling, and the scaling relationship is defined as follows: (14) in, This is a new slow-time variable.

[0047] The product signal is resampled and interpolated along a slow time scale, and then subjected to a fast-time inverse Fourier transform to obtain the focused signal: (15) At this point, the signal energy eliminates migration across distance cells and is fully focused within the same distance cell.

[0048] Step 4: The coherent accumulation and target detection module realizes target energy focusing and parameter extraction. At this time, the focus signal A time-varying quadratic phase term caused by radial acceleration still remains. An acceleration search mesh set is constructed. For each trial acceleration The coherent accumulation and target detection module constructs a phase compensation factor. And perform a fast Fourier transform to achieve coherent accumulation between pulses, and obtain the output of the range-Doppler two-dimensional plane: (16) in, For Doppler frequency variables.

[0049] Ultimately, in the four-dimensional parameter space composed of beam trajectory, acceleration, range, and Doppler... Extract the detection statistics. Then, square the modulus. Compare with the preset constant false alarm threshold: (17) in, This serves as the detection threshold. Each peak point that meets this condition and exhibits local maxima characteristics is determined to be a valid target, and its spatial coordinates correspond to the high-precision motion parameter estimation results for that target, including the optimal beam crossing trajectory, radial acceleration, initial distance, and radial velocity.

[0050] Example: like Figure 2 As shown in the flowchart, this invention proposes a high-speed cross-beam target coherent accumulation method for panoptic radar. This method improves the signal accumulation performance by combining multiple beam processing methods.

[0051] In this embodiment, a uniform planar digital array radar is used, employing wide-beam transmission and multiple narrow-beam signal reception to stare at a certain airspace range. The radar parameters are shown in Table 1. Table 1 Radar Parameter Table

[0052] Furthermore, considering the practical research scenario of low-orbit space-based radar targeting weak targets in space, this invention sets up three targets. Two targets have trajectories that are compared, while the trajectory of the third target does not intersect with the first two, representing targets under different scenarios. The motion parameters of these three targets in three-dimensional space are summarized in Table 2.

[0053] Table 2 Target Parameters

[0054] In addition, to visually demonstrate the target's cross-beam motion process within the two-dimensional beam domain, such as Figure 3 As shown. In this invention, the two-dimensional beam domain refers to the domain formed by... Direction and azimuth cosine and Direction and azimuth cosine The diagram depicts a receiving beam pointing towards a plane. Circular regions represent the half-power beamwidth coverage area of ​​each receiving beam within the two-dimensional azimuth cosine plane, and numbers within the circles indicate the index numbers of the corresponding receiving beams. The three trajectories of different line types in the diagram represent the beam domain motion trajectories of target 1, target 2, and target 3 within one coherent accumulation time. The starting point of the trajectory is indicated by a dot, and the ending point by a triangle. The initial positions of the three targets are located in the receiving beam regions near indices 34, 32, and 30, respectively, and their ending positions are located in the receiving beam regions near indices 15, 16, and 23, respectively. Target 1 passes through approximately four receiving beams sequentially during the observation period, target 2 passes through approximately four receiving beams sequentially, and target 3 passes through approximately two receiving beams sequentially during the observation period. This diagram illustrates that high-speed targets undergo significant cross-beam migration within the coherent accumulation time, and the target energy is dispersed across multiple receiving beams. Therefore, it is necessary to select the corresponding receiving beam pulse-by-pulse based on the candidate beam domain trajectory and construct a cross-beam splicing signal.

[0055] like Figure 4 As shown, Figure 4 The fast-time and slow-time echo distributions of the target trajectory across different receiving beams are presented. The horizontal axis represents range, and the vertical axis represents slow time; fast time corresponds to the range dimension, and slow time corresponds to the pulse sequence dimension. Figure 4 The sub-graphs in the diagram correspond to the echo distribution of targets 1, 2, and 3 in different receiving beams. For target 1, its echo is mainly distributed in receiving beams with indices 34, 28, 21, and 15; for target 2, its echo is mainly distributed in receiving beams with indices 32, 27, 21, and 16; and for target 3, its echo is mainly distributed in receiving beams with indices 30 and 23. Figure 4 It is evident that the echo energy of a single target during observation is not fixed within a single receiving beam, but rather appears sequentially in multiple receiving beams along with the target beam domain trajectory. Therefore, in the second step, this invention extracts the preprocessed signal of the corresponding beam pulse by pulse based on the candidate beam domain trajectory, and stitches the target echoes dispersed in multiple beams into a single virtual long-term observation signal. Figure 5 The diagram illustrates the signal changes during the implementation of the method described herein. It can be seen that, after processing, the target energy is effectively focused between beams.

[0056] like Figure 5 As shown, Figure 5 The process of signal processing results changing during the implementation of the method of the present invention is presented. Figure 5Subgraph (a) shows the target's preset beam trajectory in the two-dimensional receiving beam domain, used to determine the receiving beam corresponding to the target at each pulse moment; subgraph (b) shows the fast-time-slow-time processing result after slow-time flipping and second-order Keystone transform of the cross-beam spliced ​​signal, used to illustrate that the range migration of the target echo is corrected and focused in the range dimension; subgraph (c) shows the range-Doppler two-dimensional output result formed after further acceleration compensation and Doppler analysis, used to illustrate that the target energy forms a concentrated peak in the range-Doppler plane. Figure 5 As can be seen, the method of the present invention can convert the target echoes that were originally dispersed in multiple receiving beams and multiple range cells into a concentrated response in the range-Doppler plane through cross-beam signal splicing, range migration decoupling focusing, acceleration compensation and coherent accumulation processing, thereby realizing the effective detection of high-speed cross-beam targets.

[0057] In summary, the above are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for high-speed cross-beam coherent accumulation of targets in a general-purpose radar, characterized in that, Includes the following steps: The first step is to preprocess the received echo signal and construct a beam phase compensation factor using the beam pointing parameter to perform beam phase compensation on the echo signal and eliminate phase jumps. The second step, for scenarios where the target spans multiple receiving beams, defines the end offset mode by introducing the start-end offset factor and the end-end offset factor, constructs a two-dimensional candidate beam domain trajectory, calculates the normalized beam domain distance between the two-dimensional candidate beam domain trajectory and each receiving beam under each end offset mode, and then extracts the receiving beam corresponding to the minimum distance pulse by pulse from the preprocessed multi-beam data to construct a cross-beam splicing signal, and realigns the energy of the target scattered in different beams in the time domain to form a single virtual long-term observation signal; The third step involves constructing a symmetrical slow-time variable based on the observation center time, using the cross-beam splicing signal to construct a time-varying reversal signal, and then performing decoupling multiplication on the cross-beam splicing signal to cancel the first distance travel, and then eliminating the second distance curvature coupling term to obtain the focused signal. The fourth step involves constructing an acceleration search grid for the focused signal, constructing a phase compensation factor for each trial acceleration and performing a fast Fourier transform to achieve coherent accumulation between pulses, obtaining a range-Doppler two-dimensional plane output, and extracting the effective target and its motion parameters through constant false alarm rate detection.

2. The method according to claim 1, characterized in that, The first step specifically includes: The linear frequency modulated signal transmitted by the radar is expressed as a complex signal, as follows: （1） in, It is a rectangular envelope function. The duration of the pulse. For frequency modulation slope, For signal bandwidth, For carrier frequency, For fast time variables, For slow time variables, The pulse sequence number. The number of pulses. The pulse repetition period; If the system generates The receiving digital beam, ignoring amplitude and noise terms, after down-conversion and fast-time matched filtering, the... The beam at the 1st The baseband pulse compression echo signal under one pulse is represented as: （2） in, Represents the speed of light. The radial distance between the target and the radar at different slow time points. These are the initial distance, initial velocity, and initial acceleration at the start of the observation; , The first The beam at the 1st The amplitude and phase responses formed under each pulse; Constructing compensation factors using beam pointing parameters: （3） in, For wavelength, The first The pointing center of each beam is oriented along the azimuth cosines in the x and y directions. These represent the element spacing in the x and y directions, respectively. The aforementioned compensation factor is used to compensate the baseband pulse-compression echo signal, eliminating phase jumps caused by beam switching, and transforming the signal to the fast time-frequency domain. The preprocessed signal is obtained: （4） in, This represents the Fourier transform along the fast time.

3. The method according to claim 2, characterized in that, In the second step, the initial bias factor and termination bias factor The end bias mode constitutes the above ,Pick , , and These correspond to four cross-beam crossing modes: all-in / all-out, all-in / half-out, half-in / all-out, and half-in / half-out.

4. The method according to claim 3, characterized in that, In the second step, the method for calculating the normalized beam domain distance includes: For any starting beam Termination Beam and end bias mode Construct normalized trajectory parameters: （7） This yields the two-dimensional candidate beam domain trajectories of the start and end beam pairs in each end offset mode: （8） in, , These represent the two-dimensional cosine vectors pointing towards the center for the starting and ending beams, respectively. Define the normalized beam domain distance as: （9） in, The azimuth cosine vector of the beam. and Corresponding to and The azimuth cosine of the direction. and These respectively represent the received beam at direction and Half-power beamwidth in the direction; Substituting formula (8) into formula (9), the left side of the equation in formula (9) becomes equal to the left side of the equation. Replace with The right side of the equals sign Replace with , Replace with The normalized beam domain distance between the two-dimensional candidate beam domain trajectory and the received beam is obtained under each end offset mode.

5. The method according to claim 4, characterized in that, The third step specifically includes: Introducing observation center time Symmetric slow-time variables ; Constructing a time-varying flip signal: （12） Multiplying the spliced ​​signal by the inverted signal yields the decoupled product signal: （13） Subsequently, a second-order Keystone transform is used to eliminate secondary coupling, and the scaling relationship is defined as follows: （14） in, For the new slow-time variable; The decoupled product signal is scaled and interpolated along a slow time path, and then subjected to a fast-time inverse Fourier transform to obtain the focused signal. （15）； in, This represents the inverse Fourier transform.

6. The method according to claim 5, characterized in that, The fourth step specifically includes: Construct an acceleration search grid set, and test each acceleration in the set. Construct phase compensation factor And perform a fast Fourier transform to achieve coherent accumulation between pulses, and obtain the output of the range-Doppler two-dimensional plane: （16） in, For Doppler frequency variables.

7. A high-speed cross-beam target coherent accumulation system for a general-purpose radar, used to implement the method according to any one of claims 1 to 6, characterized in that, include: The signal preprocessing and phase compensation module is used to perform down-conversion, matched filtering, and beam phase jump compensation on the echo signal received by the array. The cross-beam trajectory generation and signal stitching module is used to extract the corresponding beam signals based on the assumed beam trajectory and stitch them into a long observation signal. The distance migration decoupling and focusing module is used to eliminate first-order and second-order distance migration by using slow-time flipping and second-order Keystone transformation; The coherent accumulation and target detection module is used to perform acceleration compensation, Doppler analysis, and constant false alarm rate detection on the focused signal.

Images