Estimation Method for Occurrence Time of Typical Separation Events Based on Piecewise Fitting

Abstract

The present invention discloses a method for estimating the occurrence time of typical separation events based on piecewise fitting, comprising the following steps: S1: Obtain a radar echo signal, and perform pulse compression on the radar echo signal to obtain a range profile signal; S2: Perform envelope alignment processing on the range profile sequence in the range profile signal; S3: Perform threshold detection on each range profile in the range profile signal, extract the peak information of the range profile, and obtain a range profile peak sequence signal; S4: Extract line segments by frames based on the range profile peak sequence signal; S5: Perform fusion processing on the line segments extracted in each frame, retain the line segments with a length greater than threshold one, and obtain a target sequence; S6: Estimate the separation time based on the fused line segment information and obtain the respective echo signals of the separated targets. The present invention can accurately reflect the occurrence time of the separation event and obtain the independent echo signals of each target after separation without the need for empirical estimation of the position and velocity of the separation point time.

Description

Technical Field

[0001] The present invention relates to the technical field of signal and information processing, and particularly relates to a method for estimating the occurrence time of typical separation events based on piecewise fitting. Background Art

[0002] During the flight of a rocket, various separation events will occur. After these events occur, the overall radar echo of the target shows that one target gradually separates into two or more targets, forming a group of targets. Estimating the separation time as early as possible after the separation event occurs and effectively separating the echo signals of each target can be used as the information source for subsequent feature extraction and is an important preprocessing step in the target recognition process.

[0003] Existing studies at home and abroad all assume that the judgment of the separation event has been achieved, and the research focus is on how to separate the echo signals of multiple targets after separation. However, there is no report on how to achieve high-precision estimation of the separation time, and the research on separating group target signals mostly focuses on using time-frequency diagrams, which have high requirements for the parameter performance of the radar. After a typical separation event occurs during rocket flight, the number of scattering centers will increase and the distribution range will expand in the range dimension of the echo, and the existing target observation radars can obtain the range dimension information of the target. Therefore, estimating the separation time and separating multiple target signals based on range information has good applicability and a wide range of application scenarios, but the existing technology has not given an effective solution. Summary of the Invention

[0004] In view of the above problems, the present invention aims to provide a method for estimating the occurrence time of typical separation events based on piecewise fitting, and uses the distribution information of the target in the range dimension in the radar continuous observation echo sequence to estimate the occurrence time of the typical separation event during rocket flight and separate the corresponding signals.

[0005] The technical solution of the present invention is as follows:

[0006] A method for estimating the occurrence time of typical separation events based on piecewise fitting, comprising the following steps:

[0007] S1: Obtain the radar echo signal, and perform pulse compression on the radar echo signal to obtain a range image signal;

[0008] S2: Perform envelope alignment processing on the range image sequence in the range image signal;

[0009] S3: Perform threshold detection on each range image in the range image signal, extract the peak information of the range image, and obtain a range image peak sequence signal;

[0010] S4: Extract line segments frame by frame based on the range image peak sequence signal;

[0011] S5: Fuse the line segments extracted from each frame, retain the line segments with lengths greater than threshold one, and obtain the target sequence;

[0012] S6: Estimate the separation moment based on the fused line segment information and obtain the respective echo signals of the separated targets.

[0013] Preferably, in step S1, the pulse compression of the radar echo signal specifically includes the following sub-steps:

[0014] Let the radar radiation signal be h(t), and the received echo signal of the nth pulse be s r (n,t), and the signal corresponding to the matched filtering process is s o (n,t), and the corresponding processing process is as follows:

[0015]

[0016] In the formula: T r is the pulse duration; h(t - x) is the pulse response function; t is the fast time of the echo;

[0017] Performing range compression on multiple consecutive pulses can obtain the range image sequence;

[0018] For s o (n,t), after sampling with t s , the corresponding digital signal is represented as s o (n,m), where m represents the number of range sampling units, 1 ≤ m ≤ M, M represents the total number of discrete values of the one-dimensional range image, and the range corresponding to one sampling unit is Δr = ct s / 2, and c represents the speed of light.

[0019] Preferably, when performing pulse compression, the pulse compression process is implemented in the time domain or through Fourier transform in the frequency domain.

[0020] Preferably, in step S2, when performing envelope alignment processing:

[0021] Use s o (n ref ,m) to represent the reference range image, s o (n c ,m) to represent the range image to be aligned, and Δm is the range cell offset, then its correlation coefficient form is as follows:

[0022]

[0023] In the formula: R(Δm) represents the correlation coefficient under the condition of Δm;

[0024] When R(Δm) reaches the maximum value, it means that the correlation coefficient between the echoes is the largest, that is, the alignment of the two echoes is completed;

[0025] When the envelopes are aligned, the first one-dimensional range image is selected as the reference range image. The reference range image and the adjacent echoes are envelope-aligned based on the maximum correlation coefficient. After the alignment is completed, the adjacent range images are calibrated and aligned with it as the reference range image, and so on until all range images are envelope-aligned. The range image signal of the nth pulse after alignment is expressed as

[0026] Preferably, in step S3, when performing threshold detection, the following formula is used for threshold detection:

[0027]

[0028] In the formula: represents the result of the threshold detection; represents the range image signal of the nth pulse after alignment; k represents the detection coefficient.

[0029] Preferably, in step S4, when extracting line segments by frame, the Hough transform is used for extraction. When extracting line segments from a certain frame of signal using the Hough transform, the following specific sub-steps are included:

[0030] S41: Generate a quantization parameter space (ρ, θ) according to the image size and initialize each element to 0;

[0031] S42: For each non-zero point in the peak image of the range image, according to ρ i = mcosθ i + nsinθ i calculate its corresponding point in the Hough transform parameter space and increment the value of the corresponding point by 1, where (m, n) are the coordinates corresponding to the non-zero point in the peak image, m is the range sampling unit, and n is the corresponding time sampling unit;

[0032] S43: After the statistics, the peak points in the (ρ, θ) parameter space that are greater than the second threshold correspond to a line segment;

[0033] Through the above processing, assuming that the lk line segments are extracted from the kth frame echo sequence, the corresponding parameters are expressed as where L k represents the set of estimated line segment parameters; represents the estimated slant range corresponding to the lkth line segment; represents the estimated angle corresponding to the lkth line segment.

[0034] Preferably, in step S43, the value of the second threshold is M r / 2, where M r represents the segmentation length.

[0035] Preferably, in step S5, the fusion process specifically includes the following sub-steps:

[0036] Let be a certain line segment extracted from the k-th frame, and

[0037]

[0038] be a certain line segment extracted from the (k + 1)-th frame. When the following relationship is satisfied, the two line segments are considered to be successfully matched and belong to the same target: r where: Δθ represents the angular error threshold; M

[0039] represents the segment length; Δρ represents the distance error threshold; represents the line segment parameter corresponding to the k-th frame, represents the line segment parameter corresponding to the (k + 1)-th frame;

[0040] When there is still a line segment in the (k + 2)-th frame that is successfully matched with the line segment in the (k + 1)-th frame, the line segment parameter corresponding to the (k + 2)-th frame is added to Comp_Line;

[0041] According to the above judgment conditions, the line segments of {L1, L2}, {L2, L3}....{L K-1 , L K} are sequentially matched to obtain the fused line segment; each fused line segment represents a target, and its corresponding line segment parameter can be expressed as:

[0042]

[0043] where: p k represents the starting frame number; p j represents the continuous frame number.

[0044] In step S5, retaining the line segment with a length greater than threshold one means retaining the valid line segment with a continuous frame number greater than 1.

[0045] Preferably, in step S6, when estimating the separation moment, the following two conditions are used to determine whether a separation event has occurred:

[0046] (1) The number of valid targets increases;

[0047] (2) The distance difference between the positions of the added targets and the original targets is less than threshold three;

[0048] When both of the above conditions are satisfied, it is determined that a separation event has occurred; the corresponding frame number is the separation moment.

[0049] Preferably, step S6 specifically includes the following sub-steps:

[0050] S61: Find the segment with the earliest start time from the fused segments as the reference segment;

[0051] S62: Compare one by one the distance corresponding to the start time of other segments with the distance difference of the reference segment at the corresponding time. When the distance difference is less than threshold three, it is considered that the target corresponding to the segment is separated; these two fused segments are used as a matching segment for a separation event {Comp_Line(p ref ), Comp_Line(p p )};

[0052] S63: Use the segment with the second earliest start time as the reference segment, and determine whether the remaining subsequent segments belong to its separated target; and so on until all segments are processed;

[0053] S64: After obtaining a fused segment {Comp_Line(p ref ), Comp_Line(p p )} for a certain matching separation event, the corresponding separation time sampling unit is p p .M r , and the separation time is:

[0054] t p = p p .M r .t s (6)

[0055] According to the fused segment of the separation event, the sampling time unit corresponding to the target before separation is [p ref .M r , p p .M r - 1], the sampling time unit corresponding to target 1 after separation is [p p .M r , p ref,j .M r - 1], and the sampling unit corresponding to the target is [p p .M r , p p,j .M r - 1];

[0056] For each target, in each segment, the distance unit corresponding to the target is obtained according to the parameters of the corresponding segment, so as to obtain its RCS as where is the corresponding distance unit, (ρ′, θ′) are the parameters of the corresponding segment, and n is the corresponding time sampling unit.

[0057] The beneficial effects of the present invention are:

[0058] The present invention does not require empirical estimation of the position and velocity at the separation point moment. Meanwhile, it has a good inhibitory effect on the tracking error of the reflective radar, with stable output track, faster convergence speed and higher accuracy of the landing point prediction result. The present invention has low requirements for radar performance parameters and can be applied to existing target observation radars. Its results can accurately reflect the occurrence moment of the separation event and obtain independent echo signals of each target after separation. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0060] Figure 1 It is a schematic flow chart of the method for estimating the occurrence moment of a typical separation event based on piecewise fitting according to the present invention;

[0061] Figure 2 It is a schematic diagram of the original range profile sequence of the separated target in a specific embodiment;

[0062] Figure 3 It is a schematic diagram of the range profile sequence after envelope alignment in a specific embodiment;

[0063] Figure 4 It is a schematic diagram of the peak detection result in a specific embodiment;

[0064] Figure 5 It is a schematic diagram of the piecewise fitting result of frame-by-frame segmentation in a specific embodiment;

[0065] Figure 6 It is a schematic diagram of the result after frame-by-frame line segment fusion in a specific embodiment;

[0066] Figure 7 It is a schematic diagram of the RCS of the separated target in a specific embodiment;

[0067] Figure 8 It is a schematic diagram of the range cell obtained from the separated target in a specific embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0068] The present invention will be further described below in conjunction with the accompanying drawings and embodiments. It should be noted that, without conflict, the embodiments in the present application and the technical features in the embodiments may be combined with each other. It should be pointed out that, unless otherwise specified, all technical and scientific terms used in the present application have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present application belongs. The term "including" or "comprising" or similar words used in the disclosure of the present invention means that the element or object appearing before the word encompasses the element or object listed after the word and its equivalents, without excluding other elements or objects.

[0069] As Figure 1 shown, the present invention provides a method for estimating the occurrence time of a typical separation event based on piecewise fitting, including the following steps:

[0070] S1: Obtain the radar echo signal, and perform pulse compression on the radar echo signal to obtain a range profile signal.

[0071] In a specific embodiment, performing pulse compression on the radar echo signal specifically includes the following sub-steps:

[0072] Let the radar radiation signal be h(t), and the received echo signal of the nth pulse be s r (n,t), and the corresponding signal for matched filtering processing is s o (n,t), and the corresponding processing process is as follows:

[0073]

[0074] In the formula: T r is the pulse duration; h(t - x) is the pulse response function; t is the fast time of the echo;

[0075] Optionally, the above pulse compression process can be implemented in the time domain or in the frequency domain through Fourier transform. Performing range compression on multiple consecutive pulses can obtain the range profile sequence.

[0076] For s o (n,t), the corresponding digital signal after sampling with t s is represented as s o (n,m), where m represents the number of range sampling units, 1 ≤ m ≤ M, M represents the total number of discrete values of the one-dimensional range profile, and the range corresponding to one sampling unit is Δr = ct s / 2, and c represents the speed of light.

[0077] S2: Perform envelope alignment processing on the range profile sequence in the range profile signal.

[0078] For the signal after pulse compression, the actual range image sequence obtained has serious interleaving phenomenon, so the echoes must be aligned first to achieve compensation of the pulse envelope.

[0079] In a specific embodiment, when performing envelope alignment processing: using s o (n ref ,m) represents the reference distance image, s o (n c ,m) represents the range image to be aligned, Δm is the range unit offset, and the correlation coefficient is as follows:

[0080]

[0081] Where: R(Δm) represents the correlation coefficient under Δm conditions;

[0082] When R(Δm) reaches the maximum value, it means that the correlation coefficient between echoes is the largest, that is, the alignment of the two echoes is completed;

[0083] When aligning the envelope, the first one-dimensional range image is selected as the reference range image, and the reference range image is aligned with the adjacent echo based on the maximum correlation coefficient. After the alignment is completed, the reference range image is used as the reference range image to calibrate and align the adjacent range images, and so on until the envelope alignment of all range images is completed; the range image signal of the nth pulse after alignment is expressed as

[0084] S3: Perform threshold detection on each range image in the range image signal, extract peak information of the range image, and obtain a range image peak sequence signal.

[0085] The threshold detection method is used to extract the peak information of each range image to obtain the range image peak sequence signal, which can reduce the amount of calculation for subsequent processing. The detection principle of the threshold detection method is as follows:

[0086]

[0087] Where: Indicates the result of over-threshold detection; represents the range image signal of the nth pulse after alignment; k represents the detection coefficient.

[0088] The detection coefficient k is determined by the detection probability and the false alarm probability. For a single range image, there are more false alarm information in the detection results, but from the range image sequence, the real target echo is consistent and appears as a continuous line segment. The false alarm information detected by a single range image will not affect the subsequent detection results.

[0089] S4: extracting line segments based on the range profile peak sequence signal by frame division.

[0090] For a short continuous observation time, the motion of the target can be approximately considered as uniform linear motion. Therefore, the change in its position is linear and is represented as a line segment in the range image peak information sequence. The false alarm peak points are discrete. Therefore, the effective extraction of the target can be achieved through the Hough transform.

[0091] In order to approximate the motion of the target as uniform motion within the observation time, the signal is segmented. Let the segment length be M. r , then the echo sequence is divided into K frames, where the 1st to M r pulse signals are the first frame, M r +1 to 2M r is the 2nd frame, and so on. The last frame with less than M r pulse signals is taken as one frame.

[0092] In a specific embodiment, when using the Hough transform to extract the line segment from a certain frame of signal, it specifically includes the following sub-steps:

[0093] S41: Generate a quantization parameter space (ρ, θ) according to the image size and initialize each element to 0;

[0094] S42: For each non-zero point in the peak image of the range image, calculate its corresponding point in the Hough transform parameter space according to ρ i = mcosθ i + nsinθ i , and add 1 to the corresponding point value, where (m, n) is the coordinate corresponding to the non-zero point in the peak image, m is the range sampling unit, and n is the corresponding time sampling unit;

[0095] S43: After the statistics, the peak points in the (ρ, θ) parameter space greater than the second threshold correspond to a line segment;

[0096] Through the above processing, assuming that the lk line segments are extracted from the kth frame echo sequence, its corresponding parameters are expressed as where L k represents the set of estimated line segment parameters; represents the estimated slant range corresponding to the lkth line segment; represents the estimated angle corresponding to the lkth line segment.

[0097] In a specific embodiment, the value of the second threshold is M r / 2.

[0098] S5: Perform fusion processing on the line segments extracted from each frame, retain the line segments with a length greater than the first threshold, and obtain the target sequence.

[0099] For the line segments estimated frame by frame using the Hough transform, when two line segments extracted between two adjacent frames belong to the same target, their corresponding line segment parameters are consistent. Therefore, in a specific embodiment, the fusion process specifically includes the following sub-steps:

[0100] Let be a certain line segment extracted in the k-th frame, and

[0101]

[0102] be a certain line segment extracted in the (k + 1)-th frame. When the following relationship is satisfied, the two line segments are considered to be successfully matched and belong to the same target: r where: Δθ represents the angular error threshold; M

[0103] represents the segmentation length; Δρ represents the distance error threshold; If the matching is correct, the matched line segments are fused and represented as 1 line segment where represents the line segment parameters corresponding to the k-th frame,

[0104] and

[0105] represents the line segment parameters corresponding to the (k + 1)-th frame; When there are still line segments in the (k + 2)-th frame that are successfully matched with the line segments in the (k + 1)-th frame, the line segment parameters corresponding to the (k + 2)-th frame are added after Comp_Line; K-1 , L K} are processed for matching to obtain the fused line segments; Each fused line segment represents a target, and its corresponding line segment parameters can be expressed as:

[0106]

[0107] where: p k represents the starting frame number; p j represents the continuous frame number.

[0108] To retain the line segments with a length greater than threshold one, specifically, the temporary line segments are deleted, and the effective line segments that are paired in multiple frames are retained. For the temporary line segments, their continuous frame number is small, generally only one frame, that is, p j = 1, while for the effective target line segments, they can be continuously observed within multiple frames. Therefore, p j > 1. Therefore, according to whether p j is equal to 1, the temporary line segments are deleted, and only the effective line segments with a long duration are retained. The retained line segments correspond to the positions of the effectively observed targets. Therefore, in this embodiment, retaining the line segments with a length greater than threshold one means retaining the effective line segments with a continuous frame number greater than 1.

[0109] S6: Estimate the separation moment based on the fused line segment information and obtain the respective echo signals of the separated targets.

[0110] In a specific embodiment, when estimating the separation moment, it is determined whether a separation event has occurred through the following two conditions: (1) The number of valid targets increases; (2) The distance difference between the positions of the increased targets and the original target positions is less than threshold three; When both of the above conditions are satisfied, it is determined that a separation event has occurred; The corresponding frame number at this time is the separation moment.

[0111] In a specific embodiment, step S6 specifically includes the following sub-steps:

[0112] S61: Find the segment with the earliest starting moment from the fused line segments as the reference segment.

[0113] S62: Compare one by one the distance corresponding to the starting moment of other segments with the distance of the reference segment at the corresponding moment. When the distance difference is less than threshold three, it is considered that the target corresponding to this segment is the separated target; These two fused line segments are used as a matching line segment for a separation event {Comp_Line(p ref ), Comp_Line(p p )}.

[0114] S63: Use the segment with the second earliest starting moment as the reference segment and determine whether the remaining subsequent segments belong to its separated target; And so on until all line segments are processed;

[0115] S64: After obtaining a fused line segment {Comp_Line(p ref ), Comp_Line(p p )} for a certain matching separation event, the corresponding separation time sampling unit is p p .M r , and the separation time is:

[0116] t p = p p .M r .t s (6)

[0117] According to the fused line segment of the separation event, the sampling time unit corresponding to the target before separation is [p ref .M r , p p .M r - 1], the sampling time unit corresponding to target 1 after separation is [p p .M r , p ref,j .M r - 1], and the sampling unit corresponding to the target is [pp .M r ,p p,j .M r -1];

[0118] For each target, in each segment, the distance unit corresponding to the target is obtained according to the parameters of the corresponding line segment, so as to obtain its RCS as where is the corresponding distance unit, (ρ′, θ′) are the parameters of the corresponding line segment, and n is the corresponding time sampling unit.

[0119] In a specific embodiment, the method for estimating the occurrence time of the separation event based on piecewise fitting of typical separation events described in the present invention is used to estimate the occurrence time of the separation event. Specifically:

[0120] In this embodiment, the simulation parameters are set as follows: assuming that the radar center frequency is 10 GHz, the bandwidth is 30 MHz, the sampling time interval t s = 0.01 s, the total number of discrete values of the one-dimensional range profile M = 1000, that is, a total of 1000 pulses are observed (corresponding to a time of 10 s), the corresponding separation time is 1.25 s, the speed difference between the two targets after separation is 7.48 m / s, and the range profile sequence obtained after pulse compression of the echo is as Figure 2 shown. After envelope alignment of the echo, the obtained result is as Figure 3 shown. From Figure 2 and Figure 3 it can be seen that after envelope alignment processing, the range profiles are effectively aligned.

[0121] For Figure 3 each range profile in, threshold detection is performed to extract the peak information of the range profile, and a range profile peak sequence signal is obtained. The result is as Figure 4 shown. Let the segment length M r = 200, that is, 200 sampling times (corresponding to a time of 2 s) are taken as 1 frame, and Hough transform is used to extract line segments from each frame of image. The obtained extraction result is as Figure 5 shown. From Figure 5 it can be seen that Hough transform correctly estimates the corresponding line segments from each distance segment. The result after fusing each frame of data is as Figure 6 shown. From Figure 6 it can be seen that in this embodiment, the fusion of the corresponding line segments of the two targets after separation is realized. Take the threshold three k dist = 5. Finally, it is judged that a separation event occurs at 2 s, and the corresponding estimation error is 0.75 s. The RCS data of each target is as Figure 7 shown, and the distance unit corresponding to each target in each time sampling unit is as Figure 8 shown. From Figure 7 and Figure 8It can be seen that the present invention realizes the correct estimation of the separation moment and can obtain the RCS values of each target before and after separation.

[0122] In summary, the present invention can accurately reflect the occurrence moment of the separation event and obtain the independent echo signals of each target after separation without the need for empirical estimation of the position and velocity at the separation point moment. Compared with the prior art, the present invention has made remarkable progress.

[0123] The above are only the preferred embodiments of the present invention and do not impose any form of limitation on the present invention. Although the present invention has been disclosed above with the preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications to equivalent embodiments by using the disclosed technical content within the scope of the technical solution of the present invention. However, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A method for estimating the occurrence time of a typical separation event based on piecewise fitting, characterized in that, It includes the following steps: S1: Obtain the radar echo signal, and perform pulse compression on the radar echo signal to obtain a range image signal; S2: Perform envelope alignment processing on the range image sequence in the range image signal; S3: Perform threshold detection on each range image in the range image signal, extract the peak information of the range image, and obtain a range image peak sequence signal; S4: Based on the range image peak sequence signal, use the Hough transform to extract line segments frame by frame; S5: Perform fusion processing on the line segments extracted from each frame, retain the line segments with a length greater than threshold one, and obtain a target sequence; S6: Estimate the separation moment based on the fused line segment information and obtain the respective echo signals of the separated targets.

2. The method for estimating the occurrence time of a typical separation event based on piecewise fitting according to claim 1, wherein In step S1, the specific steps of performing pulse compression on the radar echo signal include the following sub-steps: Let the radar radiation signal be h(t), and the received echo signal of the nth pulse be s r (n,t), and the signal corresponding to the matched filtering process is s o (n,t), and the corresponding processing procedure is as follows: （1） where: T r is the pulse duration; h(t - x) is the impulse response function; t is the fast echo time; Performing range compression on multiple consecutive pulses can obtain the range image sequence; For s o (n, t) is sampled at a sampling time interval t s The corresponding digital signal after sampling is denoted as s o (n, m), where m represents the number of range sampling cells, 1 ≤ m ≤ M, M represents the total number of discrete values of the one-dimensional range profile, and the range corresponding to one sampling cell is Δr = ct s / 2, where c represents the speed of light.

3. The method for estimating the occurrence time of a typical separation event based on piecewise fitting according to claim 2, wherein When performing pulse compression, the pulse compression process is implemented in the time domain or through Fourier transform in the frequency domain.

4. The method for estimating the occurrence time of a typical separation event based on piecewise fitting according to claim 1, wherein In step S2, when performing envelope alignment processing: Use s o (n ref , m) represents the reference range image, s o (n c , m) represents the range image to be aligned, and Δm is the range cell offset. Then its correlation coefficient form is as follows: （2） Where: R(Δm) represents the correlation coefficient under the condition of Δm; m represents the number of range sampling units; M represents the total number of discrete values of the one-dimensional range image; When R(Δm) reaches the maximum value, it indicates that the correlation coefficient between echoes is the largest, that is, the alignment of the two echoes is completed; When performing envelope alignment, select the first one-dimensional range image as the reference range image, perform envelope alignment on the reference range image and the adjacent echo based on the maximum correlation coefficient, and after the alignment is completed, use it as the reference range image to calibrate and align its adjacent range images, and so on until all range images are completed with envelope alignment; The range image signal of the nth pulse after alignment is expressed as .

5. The method for estimating the occurrence time of a typical separation event based on piecewise fitting according to claim 1, wherein In step S3, when performing threshold detection, perform threshold detection through the following formula: （3） In the formula: represents the result of over-threshold detection; represents the range image signal of the nth pulse after alignment; m represents the number of range sampling units; k represents the detection coefficient; M represents the total number of discrete values of the one-dimensional range image.

6. The method for estimating the occurrence time of a typical separation event based on piecewise fitting according to claim 1, wherein In step S4, when using the Hough transform to extract line segments from a certain frame of signal, it specifically includes the following sub-steps: S41: Generate a quantization parameter space according to the image size , and initialize each element to 0; S42: For each non-zero point in the peak image of the range image, according to calculate its corresponding point in the Hough transform parameter space, and add 1 to the value of the corresponding point, where is the coordinate corresponding to the non-zero point in the peak image, m is the range sampling unit, and n is the corresponding time sampling unit; S43: After the statistics are completed, The peak points in the parameter space that are greater than threshold two correspond to a line segment; Through the above processing, assuming that the k-th frame echo sequence extracts lk segments of line segments, the corresponding parameters are expressed as , where L k represents the set of estimated line segment parameters; represents the slant range corresponding to the estimated lk-th line segment; represents the angle corresponding to the estimated lk-th line segment.

7. The method for estimating the occurrence time of a typical separation event based on piecewise fitting according to claim 6, wherein In step S43, the value of the second threshold is M r / 2, where M r represents the segmentation length.

8. The method for estimating the occurrence time of a typical separation event based on piecewise fitting according to claim 1, wherein In step S5, the specific steps of performing fusion processing include the following sub-steps: Let be a certain line segment extracted from the k-th frame, be a certain line segment extracted from the (k + 1)-th frame. When the following relationship is satisfied, it is considered that the two line segments are successfully matched and belong to the same target: （4） Where: Δθ represents the angle error threshold; M r represents the segment length; Δρ represents the distance error threshold; If the matching is correct, the fused line segments after matching are represented as one line segment , where represents the line segment parameters corresponding to the k-th frame, represents the line segment parameters corresponding to the (k + 1)-th frame; When there is still a line segment in the k + 2 frame that successfully matches the line segment in the k + 1 frame, then add the line segment parameters corresponding to the k + 2 frame after Comp_Line; According to the above judgment conditions, successively , .... of the line segments are matched to obtain the fused line segments; each fused line segment represents a target, and its corresponding line segment parameters are expressed as: （5） Where: p k represents the starting frame number; p j represents the number of consecutive frames; In step S5, retaining the line segments with a length greater than threshold one means retaining the valid line segments with a continuous number of frames greater than 1.

9. The method for estimating the occurrence time of a typical separation event based on piecewise fitting according to claim 1, characterized in that In step S6, when estimating the separation moment, judge whether a separation event has occurred through the following two conditions: (1) The number of valid targets increases; (2) The distance difference between the positions of the added targets and the original target positions is less than threshold three; When both of the above conditions are met at the same time, it is judged that a separation event has occurred; the corresponding frame number is the separation moment.

10. The method for estimating the occurrence time of a typical separation event based on piecewise fitting according to claim 9, characterized in that, Step S6 specifically includes the following sub-steps: S61: Find the earliest starting segment from the fused line segments as the reference segment; S62: Compare one by one the distance differences between the distances corresponding to other line segments at the starting moment and the distance of the reference line segment at the corresponding moment. When the distance difference is less than Threshold Three, it is considered that the target corresponding to the line segment is separated; use these two fused line segments as a matching line segment for a separation event ; S63: Use the line segment with the second earliest starting moment as the reference segment, and judge whether the remaining subsequent line segments belong to its separated target; and so on until all line segments are processed; S64: Obtain the merged segment of a certain matching separation event After that, the corresponding separation time sampling unit is , and the separation time is: （6） where: t s represents the sampling time interval; M r represents the segment length; Fusing the line segments according to the separation event, the sampling time unit corresponding to the target before separation is , and the sampling time unit corresponding to target 1 after separation is , and the sampling unit corresponding to the target is ; For each target, in each segment, obtain the distance unit corresponding to the target according to the parameters of the corresponding line segment, so as to obtain its RCS as , where is the corresponding distance unit, are the parameters of the corresponding line segment, and n is the corresponding time sampling unit.

Images