A shipborne ground wave radar target detection method based on adaptive beam RDT

Abstract

The application discloses a shipborne ground wave radar detection and tracking integrated beam RDT data adjustment and repair method, relates to the field of shipborne ground wave radar target detection, and has the following basic steps: fixed beam RDT first frame construction and preprocessing; suspected target classification and initial beam determination; comprehensive attitude information is used to judge whether a target is in a detection area; RDT beam dynamic determination; RDT repair; DP-TBD target integrated detection; and multi-beam target result fusion. The application fully utilizes shipborne ground wave radar multi-beam data, adaptively constructs beam RDT three-dimensional data based on target azimuth information and platform attitude information, dynamically adjusts beam RDT data in the target integrated detection process in combination with target self-motion and change of the heading of the shipborne platform, improves the target signal-to-noise ratio, and proposes an RDT data repair method for the target signal loss problem caused by the change of the heading, so that the target integrated detection process can be continued, the shipborne ground wave radar can realize long-time stable tracking of the target, and the target detection performance of the shipborne ground wave radar is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of shipborne ground wave radar target detection, and particularly relates to a detection and tracking integrated method for shipborne ground wave radar beam RDT adaptive construction. BACKGROUND

[0002] High frequency surface wave radar (HFSWR) uses vertically polarized electromagnetic waves (3-30MHz) to propagate around the sea surface with small attenuation, which can achieve large-scale, over-the-horizon continuous monitoring of sea targets, and provide real-time information such as the position, speed and direction of moving targets. Compared with traditional ground-based ground wave radar, shipborne ground wave radar is more mobile and flexible, further expanding the range of ocean monitoring, and has wide application value in the fields of maritime rights protection and maritime traffic management. In the ground wave radar target detection method, the detection and tracking integrated method extends the detection to the time dimension, does not distinguish whether there is a target in a single frame of echo data, but improves the signal-to-noise ratio of the target through accumulation of multiple frames of data, and finally completes target detection and track tracking at the same time, further improving the detection performance of weak target signals in compact ground wave radar.

[0003] The construction of Range-Doppler-Time (RDT) three-dimensional data is a key step for ground wave radar target detection and tracking integration. At present, the existing shipborne ground wave radar target detection and tracking integrated method usually uses single-channel echo information or simple average information of multiple channels when constructing RDT data, and the signal-to-noise ratio of the target in these information is usually low, which is not conducive to the detection and tracking of weak targets. The beam data obtained by beam synthesis of multi-channel data has a higher signal-to-noise ratio and can be used to better detect and track targets. However, for shipborne ground wave radar, the heading of the platform will change dramatically under the influence of ocean dynamic environmental factors such as waves and currents, causing changes in the radar detection area and exacerbating the change of the target's azimuth in the radar coordinates. In a short period of time, the target may cross multiple beams, or even some targets may be moved out of the radar detection area. At this time, in the RDT constructed according to the traditional method, the target moved out of the radar detection area lacks effective echo signals for a period of time, which causes the detection and tracking integration for the target to be unable to continue, and further causes the target track to be broken or lost.

[0004] The application provides a shipborne ground wave radar target detection and tracking integrated method, which fully utilizes target azimuth information and shipborne platform attitude information and simultaneously carries out integrated detection in multi-beam RDT data. In the integrated detection, the beam RDT data is adaptively and dynamically adjusted, the signal-to-noise ratio and detection performance of the target in the integrated detection are improved, and simultaneously, aiming at the problem that the target signal is temporarily lost due to the dramatic change of the platform heading, an RDT data repairing method is provided, the integrated detection process on the related target is maintained, and the continuous and stable tracking on the target is realized. SUMMARY

[0005] (1) Technical problems to be solved

[0006] The application aims to provide a shipborne ground wave radar target detection and tracking integrated method based on adaptive beam RDT construction, which carries out integrated detection in multi-beam RDT data. The radar multi-channel data, target azimuth information and platform attitude information are fully utilized, the problems of low target signal energy, track breakage and loss caused by signal loss under the condition that the shipborne platform heading changes dramatically are solved, the continuous and stable tracking on the target is realized, and the target detection performance of the shipborne ground wave radar is improved.

[0007] (2) Technical solutions

[0008] The application comprises the following steps:

[0009] (1) Fixed beam RDT construction and first frame preprocessing

[0010] Multi-channel time domain data of the shipborne radar and attitude data of the shipborne platform are acquired, the attitude data comprises platform speed v p (k), platform heading Radar main shaft angle rs p (k) and the like. Wherein k represents the frame number of data, k=1, 2, 3,..., K, K represents the total frame number of data.

[0011] The number of beams required to cover the entire detection area is set as N (set according to the actual situation), and N beam RDT three-dimensional data structures are constructed. The angle of each beam pointing angle is θ(j), j is the number of each beam in the counterclockwise direction, j=1, 2,..., N. The first frame data of the constructed fixed beam RDT is preprocessed (CFAR detection), and a suspected target point track data set Point is obtained, which comprises target amplitude value f (unit dB), azimuth angle α (relative to the radar main shaft), radial distance r, Doppler velocity v, distance cell r g , Doppler cell v g , longitude lon, latitude lat, target signal-to-noise ratio highest beam flag bit flag (default value-1) and the like.

[0012] (2) Suspected target classification and initial beam determination

[0013] The suspected target track data set Point obtained in (1) is subjected to beam classification processing. If the absolute value of the difference between the target azimuth and the beam pointing angle is minimum, it indicates that the target is closest to the beam, and the signal-to-noise ratio of the target in the beam is highest. Taking the ith target as an example:

[0014]

[0015] Wherein, a(i, k) represents the azimuth of the ith target in the kth frame, θ(j) represents the pointing angle of the jth beam, and flag(i, k) = j represents that the signal-to-noise ratio of the ith target in the kth frame is highest in the jth beam.

[0016] (3) Determine whether the target is in the detection area by integrating attitude information

[0017] From the second frame (k = 2), before the detection and tracking integration processing is performed, the target azimuth and the platform heading change amount need to be considered to determine whether each target is in the effective detection area of the radar. The formula is as follows:

[0018]

[0019] Wherein, b(i, k) = 0 represents that the target is out of the effective detection area of the radar in the kth frame; b(i, k) = 1 represents that the target is still in the effective detection area of the radar in the kth frame. Wherein, represents the platform heading change amount in the kth frame, and Ω is the boundary of the effective detection area of the radar, which is usually set to 60° according to the actual situation.

[0020] If b(i, k) = 1, step (4) is entered.

[0021] If b(i, k) = 0, it is necessary to determine whether the heading of the ship-borne platform is in a periodic change. Only when the platform is in a periodic change state, the target can return to the effective detection area of the radar again. First, the platform attitude data of the previous 20 frames is obtained, and Fourier transform is used to analyze whether the platform heading is in a periodic change; then, the The size of the platform heading change amount is determined.

[0022]

[0023] Wherein, d is the size of the heading change amount, which is generally set to 5°.

[0024] s = 0, indicating that the platform heading change is small, and after the target moves out of the detection area, it is difficult to return to the detection area again by changing the platform heading, and the detection and tracking integration process of the target is terminated. s = 1, indicating that the heading is in a periodic change state and the heading change is large, and after the target moves out of the detection area, it can return to the detection area again, then step (5) is entered.

[0025] (4) RDT beam dynamic determination

[0026] RDT beam dynamic determination needs to consider the target azimuth and heading change at the same time. Taking the k-1 frame target i as an example:

[0027]

[0028] flag(i, k) = j, that is, the beam where the signal-to-noise ratio of the kth frame target i is the highest beam is the jth beam, and then the jth beam RDT data is used for detection and tracking integration of target i in the kth frame.

[0029] (5) RDT repair

[0030] The target moves out of the radar detection area for a period of time, and the effective echo signal is missing. In order to continue the detection and tracking integration process of the target, the position and amplitude information of the previous target is used to predict the position and amplitude of the next frame target, and RDT repair is completed.

[0031] The distance and speed of target i in the k-1 frame are r(i, k-1) and v(i, k-1), respectively. According to the cosine theorem, the radial distance r(i, k) of the kth frame target can be approximately expressed as:

[0032]

[0033] In combination with the attached Figure 4 It is explained that β(i, k-1) is the angle between the target motion direction and the echo direction, and φ(i, k-1) is the angle between the echo direction and the ship-borne platform motion direction; t represents the frame interval time, which is usually 60s. The angle β(i, k) between the target motion direction and the echo direction in the kth frame will also change:

[0034] β(i, k) = β(i, k-1) + Δβ

[0035]

[0036] The kth frame platform speed v p (k), the heading change is Then the kth frame target speed is as follows:

[0037]

[0038] The cell (r g , v g ) where the target is located in the kth frame can be calculated from the distance and velocity resolution. g , v g ). After obtaining the position of the target, the amplitude values of the five cells centered on the cell where the target is located need to be estimated.

[0039] Taking the three cells where the target is extended in the distance dimension as an example, the highest point m and the two next highest points n and p of the target in the (k-1)th frame are taken, the amplitudes of m and p are a(i, k-1) and c(i, k-1) respectively, and a Gaussian curve fitting is performed using the three points (see FIG. 2). Figure 3 The coordinates of the extreme point of the Gaussian curve are H(m', h), and h is the amplitude (also the highest amplitude point) of the cell (r g , v g ) where the target is located in the kth frame. Then, the difference Δμ of the corresponding distance of the highest amplitude of the target in the kth frame and the (k-1)th frame is calculated, and according to Δμ, the other two points on the curve are moved by Δμ, and the amplitude values b(i, k) and c(i, k) after the movement are calculated.

[0040] The estimation of the amplitude value in the velocity dimension is the same as above.

[0041] (6) Integrated detection of DP-TBD targets

[0042] Let X k be the coordinate set of the target in the kth frame of the RD spectrum, X k = {[r g , v g , α] T}, r g ∈ [1, r max ], r max is the number of distance dimension cells, v g ∈ [1, v max ], v max is the number of velocity dimension cells. Let Z k = {z k (r g , v g )} be the set of echo intensity measurement values of the target in the kth frame of the RD spectrum, where zk(r g , v g ) is the echo intensity measurement value of the target in the kth frame of the RD spectrum. The target track is defined as:

[0043]

[0044] where V T is the amplitude threshold. Let I(x k ) be the value function of the kth frame, and Ψ k (xk ) is I(x k-1 ) takes maximum value corresponding to x k-1 .

[0045] k=1, for all x1={ [r g , v g , α] T}∈X1, I(x1)=z1(r g , v g ), Ψ1(x1)=[0,0] T ; for 2≤k≤K, have:

[0046]

[0047]

[0048] When , the confirmation of the real target is completed, corresponding to the unique number trace(i), i=1, 2, 3…, L, L is the number of tracks. The track backtracking of the real target is carried out to obtain a target track coordinate set.

[0049] (7) Multi-beam target result fusion

[0050] Since the track of the target is formed by searching in multiple beams, the fusion of the multi-beam track results needs to be carried out after the search of all target tracks is completed. For example, the track of the target in the jth beam RDT is maintained for m frames, and the track in the j+1th beam RDT is maintained for n frames, and the two tracks need to be fused into a complete track of m+n frames. In the integration detection process, each target track has a unique track number trace(i), and the track number remains unchanged when the target dynamically adjusts the beam RDT, so only the target tracks with the same track number need to be associated to obtain the complete track of the target.

[0051] (Three) beneficial effects

[0052] The advantages of the present application are embodied in:

[0053] The application provides a shipborne ground wave radar target detection and tracking integrated method based on adaptive beam RDT construction, which utilizes multi-beam data, target azimuth information and platform attitude information obtained by the radar, first constructs an RDT three-dimensional data structure based on different beam information, and carries out first frame data preprocessing; then uses beam RDT data where the target highest signal-to-noise ratio is located to carry out integrated detection, and adjusts or repairs the beam RDT data in real time; finally, the complete track of the target is obtained by fusing multi-beam results. After the adjustment and repair of the RDT data, the three-dimensional RDT data where the target is located is dynamically changed, so that each target can obtain the best echo signal, the target signal-to-noise ratio is obviously improved, and the target detection and tracking integrated processing is beneficial. BRIEF DESCRIPTION OF DRAWINGS

[0054] Figure 1 It is a basic flowchart of the application.

[0055] Figure 2 It is a RDT beam dynamic adjustment schematic diagram.

[0056] Figure 3 It is a RDT repair schematic diagram.

[0057] Figure 4 It is a shipborne ground wave radar target detection schematic diagram. DETAILED DESCRIPTION

[0058] In order to make the purpose, content and advantages of the application more clear, the specific embodiment of the application is further described in detail below with reference to the drawings:

[0059] Reference Figure 1 , the specific implementation steps of the application are:

[0060] (1) Fixed beam RDT construction and first frame preprocessing

[0061] Obtain multi-channel time domain data of the shipborne radar and attitude data of the shipborne platform, the attitude data of the shipborne platform includes platform speed v p (k), platform heading Radar main shaft angle rs p (k) and the like, the radar main shaft is relative to the north direction, clockwise is positive, wherein k represents the frame number of data, k=1, 2, 3,..., K, K represents the total frame number of data.

[0062] In order to make the coverage width of adjacent beams not less than 3dB, generally, multiple beams with different azimuths are set, each beam points to θ(j) (angle relative to the radar main axis), j is the number of each beam in clockwise direction, j = 1, 2, …, N. The relationship between beam width and beam pointing angle is as follows, M antenna elements and linear array with interval d0, beam main lobe width θ Bw The formula is as follows:

[0063]

[0064] Wherein, c is the speed of light, f c is the radar frequency. Set the number of beams N required to cover the entire detection area (set by actual situation), construct N-beam RDT three-dimensional data structure. The first frame data of the constructed N-beam RDT is preprocessed, that is, CFAR detection and target azimuth estimation, to obtain a suspected target track data set Point, the track information includes target amplitude value f (unit dB), azimuth angle α (relative to the radar main axis), radial distance r, Doppler velocity v, distance cell r g , Doppler cell v g , longitude lon, latitude lat, target signal-to-noise ratio highest beam flag bit flag (default value is -1) and other information.

[0065] (2) Suspected target classification and initial beam determination

[0066] The suspected target track data Point obtained in (1) is subjected to beam classification processing, and the difference between the target azimuth and the pointing angle of each beam is calculated to determine which beam the target is closest to. If the absolute value of the difference between the target azimuth and the pointing angle of the beam is the smallest, it means that the target is closest to the beam, that is, the target has the highest signal-to-noise ratio in the beam, such as the i-th target:

[0067]

[0068] Wherein, α(i, k) represents the azimuth of the i-th target in the k-th frame, θ(j) represents the pointing angle of the j-th beam, and flag(i, k) = j represents that the i-th target in the k-th frame has the highest signal-to-noise ratio in the j-th beam.

[0069] (3) Determine whether the target is in the detection area by comprehensively considering the attitude information

[0070] From the second frame (k = 2), before the detection and tracking integration processing is performed, the target azimuth angle and the change amount of the platform heading are comprehensively considered to determine whether each target is in the effective detection area of the radar, and the formula is as follows:

[0071]

[0072] Wherein, b(i, k) = 0, indicates that the target is out of the effective detection area of the radar in the kth frame; b(i, k) = 1, indicates that the target is still in the effective detection area of the radar in the kth frame. Indicates the change amount of the heading of the platform in the kth frame, and Ω is the boundary of the effective detection area of the radar, which is usually set to 60° according to the actual situation.

[0073] If b(i, k) = 1, step (4) is entered.

[0074] If b(i, k) = 0, it is necessary to judge whether the heading of the ship-borne platform is in periodic change. Only when the platform is in periodic change, the target can return to the effective detection area of the radar again.

[0075] To judge whether the heading of the platform is in effective periodic change, first, the platform attitude data of the previous 20 frames is acquired, and Fourier transform is used to analyze whether the heading of the platform is in periodic change or only changes in one direction; then, the size of the change amount of the heading of the platform is judged.

[0076]

[0077] Wherein, d is the size of the change amount of the heading, which is usually set to 5°.

[0078] s = 0, indicates that the change amount of the heading of the platform is small and does not meet the requirement of periodic change of the heading. After the target moves out of the detection area, it is difficult for the target to return to the detection area again through the change of the heading of the platform, so the detection and tracking integration process of the target is terminated. s = 1, indicates that the heading is in periodic change and the change amount of the heading is large. After the target moves out of the detection area, the target will return to the detection area again due to the periodic change of the heading, so step (5) is entered.

[0079] (4) RDT beam dynamic determination

[0080] To perform RDT beam dynamic determination, the change amount of the heading of the target in the kth frame and the change amount of the heading of the target in the k-1th frame are considered, the sum of the change amount of the heading of the target in the kth frame and the change amount of the heading of the target in the k-1th frame is calculated, and then the difference between the sum and the pointing angle of each beam is calculated. When the absolute value of the difference is the smallest, it indicates that the target in the kth frame is closest to the beam, and the target in the next frame is replaced to the beam RDT data to perform the detection and tracking integration process. Taking the target i in the k-1th frame as an example:

[0081]

[0082] flag(i, k) = j, that is, the beam with the highest signal-to-noise ratio of the target i in the kth frame is the jth beam, so the jth beam RDT data is used to perform the detection and tracking integration of the target i in the kth frame.

[0083] (5) RDT repair ​

[0084] The effective echo signal of the target is missing for a period of time when the target moves out of the radar detection area. In order to continue the detection and tracking of the target, the position and amplitude of the next frame of the target are predicted by using the position and amplitude of the previous frame of the target, and the RDT is repaired.

[0085] The distance and velocity of the known target i in the k-1th frame are respectively r(i, k-1) and v(i, k-1), wherein:

[0086] v(i, k-1) = v r (i)cos[β(i, k-1)] + v p (k-1)cos[φ(i, k-1)]

[0087] The accompanying drawings are incorporated into and constitute a part of this specification. Figure 4 It is explained that v r (i) is the real speed of the target i, and the sea target generally moves at a constant speed, that is, v r (i) is constant; v p (k-1) is the moving speed of the ship-borne platform, β(i, k-1) is the included angle between the target moving direction and the echo direction, and φ(i, k-1) is the included angle between the echo direction and the moving direction of the ship-borne platform.

[0088] According to the cosine theorem, the radial distance r(i, k) of the target in the kth frame can be approximately expressed as:

[0089]

[0090] Wherein, t represents the frame interval time, and is usually 60s. The included angle β(i, k) between the target moving direction and the echo direction in the kth frame also changes:

[0091] β(i, k) = β(i, k-1) + Δβ

[0092]

[0093] The attitude information of the ship-borne platform in the kth frame is known, the speed is v p (k), and the change of the heading is Δφ The target speed v(i, k) in the kth frame is:

[0094]

[0095] The cell (r g , v g ) in which the target in the kth frame is located can be calculated according to the distance and speed resolution. After the position of the target is obtained, the amplitude values of the five cells centered on the cell in which the target is located need to be estimated.

[0096] Taking three cells extended by the distance dimension target as an example, the target k-1 frame distance dimension amplitude highest point m and two amplitude second highest points n and p are taken, the m, n and p three points amplitudes are a(i, k-1), b(i, k-1) and c(i, k-1), the three points are used to make Gaussian curve fitting (reference attached Figure 3 ), the Gaussian curve extreme point coordinates H(m', h) are obtained, h is the amplitude (also the amplitude highest point) of the target cell (r g , v g ) in the k frame. Then, the difference Δμ of the target amplitude highest corresponding distance of the k frame and the k-1 frame is calculated, according to Δμ, the other two points on the curve are moved Δμ, and the amplitude values b(i, k) and c(i, k) after moving are obtained.

[0097] The estimation of the velocity dimension amplitude value is the same as above.

[0098] (6) DP-TBD target integration detection

[0099] Let X k be the target coordinate set in the k frame RD spectrum, X k = {[r g , v g , α] T}, r g ∈ [1, r mnax ], r mmax is the distance dimension cell number, v g ∈ [1, v mmax ], v mmax is the velocity dimension cell number. Let the target echo intensity measurement value set in the k frame RD spectrum be Z k = {z k (r g , v g )}, wherein z k (r g , v g ) is the target echo intensity measurement value in the k frame RD spectrum, and is expressed as:

[0100]

[0101] , wherein A k is the target echo amplitude value, and ω k (r g , v g ) is the noise amplitude value. The target track is defined as:

[0102]

[0103] , wherein V T is the amplitude threshold value. Set I(xk ) is the value function of the kth frame, Ψ k (x k ) is I(x k-1 ) taking the maximum value corresponding to x k-1 .

[0104] When k=1, for all x1={ [r g , v g , α] T}∈X1, I(x1)=z1(r g , v g ), Ψ1(x1)=[0,0] T ; for 2≤k≤K, there are:

[0105]

[0106]

[0107] When , the confirmation of the real target is completed, and the corresponding unique number trace(i) is i=1, 2, 3..., L, L is the number of tracks. Track backtracking is performed on the real target:

[0108]

[0109] In the above formula, k=K-1, K-2,...1. After track backtracking, the target track coordinate set

[0110] (7) Multi-beam target result fusion

[0111] Since the track of the target is formed by searching in multiple beams, multi-beam track result fusion needs to be performed after all target track searches are completed. For example, the track of the target in the jth beam RDT is maintained for m frames, and the track in the j+1th beam RDT is maintained for n frames, and the two tracks need to be fused into a complete track of m+n frames. In the integration detection process, each target track has a unique track number trace(i), and the track number remains unchanged when the target dynamically adjusts the beam RDT, so only the target tracks with the same track number need to be associated to obtain the complete track of the target.

[0112] For each target, steps (3), (4), (5) and (6) are repeated until the detection and tracking integration process of all targets is completed. Thus, the detection and tracking integration of the shipborne ground wave radar based on the adaptive beam RDT is completed, and the target track and the detection result are obtained.

[0113] The innovation of the present application lies in the following aspects:

[0114] The application makes full use of shipborne ground wave radar multi-beam data, and adaptively constructs beam RDT three-dimensional data based on target azimuth information and platform attitude information. In combination with target self-motion and change of the heading of the shipborne platform, the beam RDT data is dynamically adjusted in the target integrated detection process, and the target signal-to-noise ratio is improved. For the problem of target signal loss caused by the change of the heading, the application proposes an RDT data repairing method, which can continue the target integrated detection process, realize long-time stable tracking of the target by the shipborne ground wave radar, and improve the target detection performance of the shipborne ground wave radar.

Claims

1.A shipborne ground wave radar target detection method based on adaptive beam RDT, comprising the following steps: (1) fixed beam RDT construction and first frame preprocessing Obtaining multi-channel time-domain data of a shipborne radar and attitude data of a shipborne platform, the attitude data including platform speed v p (k), platform heading and radar principal axis angle rs p (k), wherein k represents the frame number, k = 1, 2, 3, …, K, and K represents the total number of frames of data; Set the number of beams required to cover the entire detection area as N, N is set by the actual situation, construct N beam RDT three-dimensional data structure, each beam pointing angle is θ(j), j is each beam numbered in counterclockwise direction, j = 1, 2, …, N, the first frame data of the constructed beam RDT is preprocessed, the preprocessing is CFAR detection, and the suspected target point track data set Point is obtained, including target amplitude value f, f unit is dB, azimuth angle α, α is relative to radar main shaft angle, radial distance r, Doppler velocity v, distance cell r g , Doppler cell v g , longitude lon, latitude lat and target signal-to-noise ratio highest beam flag bit flag, the default value of flag is -1; (2) suspected target classification and initial beam determination The suspected target point data set Point obtained in (1) is subjected to beam classification processing, and if the absolute value of the difference between the target azimuth and the beam pointing angle is the smallest, it indicates that the target is closest to the beam, and the signal-to-noise ratio of the target in the beam is the highest. For the ith target: Wherein, α(i, k) represents the azimuth of the ith target in the kth frame, θ(j) represents the pointing angle of the jth beam, and flag(i, k) = j represents that the signal-to-noise ratio of the ith target in the kth frame is the highest in the jth beam; (3) judging whether the target is in the detection region by comprehensively considering the attitude information From the second frame, before the detection and tracking integration process is performed, the target azimuth and the change amount of the platform heading are comprehensively considered to judge whether each target is in the effective detection region of the radar, and the formula is as follows: Wherein, b(i, k) = 0 indicates that the target exceeds the effective detection area of the radar at the kth frame; b(i, k) = 1 indicates that the target is still in the effective detection area of the radar at the kth frame, wherein, represents the change amount of the bow direction of the platform at the kth frame, and Ω is the boundary of the effective detection area of the radar, which is set to 60° according to the actual situation. If b(i, k) = 1, step (4) is entered; If b(i, k) = 0, it is needed to judge whether the bow direction of the ship-borne platform is in periodic change. Only when the platform is in periodic change, the target can return to the effective detection area of the radar again. First, the platform posture data of the previous 20 frames is acquired, and whether the bow direction of the platform has periodic change is analyzed by using Fourier transform. Then, the size of the change of the bow direction of the platform is judged in combination with the target detection range of the radar and the target detection range of the radar in the previous 20 frames. the size of the change of the bow direction of the platform is judged in combination with the target detection range of the radar and the target detection range of the radar in the previous 20 frames. Wherein, d is the size of the heading change amount, which is set to 5°; s = 0, indicating that the change amount of the platform heading is small, and after the target moves out of the detection region, it is difficult for the target to return to the detection region again through the change of the platform heading, and the detection and tracking integration process of the target is terminated; s = 1, indicating that the heading is in a periodic change state and the change amount of the heading is large, and after the target moves out of the detection region, the target can return to the detection region again, and step (5) is entered; (4) RDT beam dynamic determination The RDT beam dynamic determination needs to consider the target azimuth and the change amount of the heading at the same time, and for the target i in the k-1th frame: That is, the beam with the highest signal-to-noise ratio of the target i in the kth frame is the jth beam, and then the jth beam RDT data is used for the detection and tracking integration of the target i in the kth frame; (5) RDT repair The effective echo signal is missing for a period of time after the target moves out of the radar detection region. In order to continue the detection and tracking integration process of the target, the position and amplitude of the next frame of target are predicted through the position and amplitude information of the previous frame of target, and the RDT repair is completed; The distance and speed of the target i in the k-1th frame are r(i, k-1) and v(i, k-1) respectively, and the radial distance r(i, k) of the target in the kth frame is approximately represented as: β(i, k-1) is the included angle between the target motion direction and the echo direction, φ(i, k-1) is the included angle between the echo direction and the shipborne platform motion direction; t represents the frame interval time, which is 60s; the included angle β(i, k) between the target motion direction and the echo direction in the kth frame also changes: β(i, k) = β(i, k-1) + Δβ The platform velocity v of the kth frame p (k), the amount of change in the heading is The target velocity of the kth frame is as follows: The cell in which the target in the kth frame is located is calculated from the distance resolution and the speed resolution, the amplitude values of the 5 cells centered on the cell in which the target is located are estimated, and then the RDT repair is completed. In the three cells of the target distance dimension extension, take the target k-1 frame distance dimension amplitude highest point m and two amplitude second high points n, p, the amplitudes of m, n, p are a(i, k-1), b(i, k-1), c(i, k-1) respectively, use the three points to make Gaussian curve fitting, the Gaussian curve extreme point coordinates H(m', h), h is the amplitude of the k frame target cell, then calculate the difference of the k frame and the k-1 frame target amplitude highest corresponding distance, recorded as Δμ, according to Δμ, move the other two points on the curve by Δμ, and calculate the corresponding amplitude values b(i, k), c(i, k) after moving; The estimation of the velocity dimension amplitude value is the same as above; (6) DP-TBD target integration detection Let X k be the set of target coordinates in the kth frame of RD spectrum, X k = {[r g , v g , α} T}, r g ∈ [1, r max ], r max is the number of distance dimension cells, v g ∈ [1, v max ], v max is the number of velocity dimension cells; let Z k = {z k (r g , v g )} be the set of target echo intensity measurements in the kth frame of RD spectrum, where z k (r g , v g ) is the target echo intensity measurement in the kth frame of RD spectrum; the target track is defined as: Wherein, V T is the amplitude threshold; set I(x k ) as the kth frame value function, Ψ k (x k ) as the x k-1 corresponding to the maximum value of I(x k-1 ); When k=1, for all x1={r g ,v g ,α] T For any x ∈ X1, we have I(x1) = z1(r) g ,v g ), Ψ1(x1)=[0,0] T For 2≤k≤K, we have: When the real target is confirmed, the corresponding unique number trace(i), i = 1, 2, 3…, L, L is the number of tracks, and track backtracking is performed on the real target to obtain a target track coordinate set. (7) Multi-beam target result fusion Since the target track is formed by searching among multiple beams, it is necessary to fuse the multi-beam track results after all target track searches are completed. When the target track in the jth beam RDT is maintained for m frames and the track in the j+1th beam RDT is maintained for n frames, the two tracks are fused into a complete track of m+n frames. In the integration detection process, each target track has a unique track number trace(i), and the track number remains unchanged when the target dynamically adjusts the beam RDT. Therefore, only the target tracks with the same track number need to be associated to obtain the complete track of the target.

Images