Target Detection Method Based on Agile Long-Time Coherent Integration with Micro-Motion Sliding Window

Abstract

The present invention belongs to the research field of radar weak target detection and is applicable to solving the problem of long-time accumulation detection of targets by hypersonic missile-borne radars. Aiming at the problem of difficult detection of low-observable targets by hypersonic missile-borne terminal guidance radars, a target detection method based on micro-motion sliding window and agile long-time coherent accumulation is provided. The computational complexity of long-time coherent accumulation is improved through a hybrid accumulation method of short-time coherent accumulation within a frame and long-time compensation coherent accumulation between frames. Data is updated through the mode of sliding window micro-motion, solving the problem of high data rate output in long-time coherent accumulation and ensuring the real-time performance of target detection. By iteratively using the stationary data within the sliding window, the computational amount is reduced. This method extends the long-time coherent accumulation of early warning radars to missile-borne radars, solving the contradiction between long-time accumulation and high data rate requirements for weak target detection by missile-borne radars. The algorithm has strong real-time performance and is convenient for engineering implementation.

Description

Technical Field

[0001] The present invention belongs to the field of radar weak target detection research and is applicable to solving the real-time detection problem of long-time accumulation of targets by hypersonic missile-borne radars. Background Art

[0002] Near-space hypersonic missiles are a new type of strategic weapon, which are fast, highly maneuverable, and have good stealth performance. However, due to the high speed of the missiles, there are problems in detecting targets. In order to achieve the real-time stability detection of high-speed low-observable targets, increasing the pulse accumulation time to improve the signal-to-noise ratio has been a research hotspot in recent years. Most research mainly focuses on coherent accumulation, mainly studying the problems of range cell migration and Doppler cell migration faced in the long-time accumulation of high-speed targets. Among them, methods such as Keytone and Radon-Fourier transform can achieve displacement compensation for range migration caused by high speed. Methods such as fractional Fourier transform and S-method can compensate for phase migration caused by the maneuver of the target in the radial direction. In response to the more complex problem of simultaneous range and Doppler cross-gate migration that may exist in high-speed targets, some scholars have proposed and studied Chirp-Fourier Transform, Radon-fractional Fourier transform, Radon-linear canonical transform, Radon-Lv’s distribution transform, generalized Radon-Fourier transform (GRFT), adaptive Radon-Fourier transform, and Radon-S transform.

[0003] For the situation of target range cell migration that occurs during the long-time accumulation of hypersonic missile-borne radars, if the traditional coherent accumulation method is continued to be used, range gate and Doppler gate migration phenomena will occur, affecting the accumulation effect, and motion compensation for the echo is required; however, long-time accumulation will result in not meeting the requirements of the guidance data rate. For a missile to hit the target, it is necessary to meet the real-time and accuracy of the target information of the terminal guidance radar, which makes the existing long-time coherent accumulation still have adaptability problems when applied to the terminal guidance radar.

[0004] The present invention provides a target detection method based on micro-motion sliding window for agile long-time coherent accumulation to detect targets. By combining coherent accumulation and non-coherent accumulation in a hybrid accumulation manner with the sliding window method, the real-time performance of long-time radar accumulation is solved, the data rate of terminal guidance radar guidance is ensured, and the stability of target detection is improved, enabling the missile to better hit the target. Summary of the Invention

[0005] Aiming at the problem that it is difficult for the terminal guidance radar to detect weak targets at high maneuvering speeds, a target detection method based on micro-motion sliding window and agile long-time coherent accumulation is provided. The computational complexity of long-time coherent accumulation is improved through a hybrid accumulation method of intra-frame short-time coherent accumulation and inter-frame long-time compensation coherent accumulation. The data rate requirement is solved by updating the data in the mode of a sliding window to ensure the real-time performance of target detection. To solve the above technical problems, the technical solutions of the present invention are as follows:

[0006] 1. A target detection method based on micro-motion sliding window and agile long-time coherent accumulation, characterized by including the following technical measures:

[0007] Step 1: Group and process the radar measurement echo signals, and the specific method is as follows:

[0008] The radar signals determine an accumulation time period according to the accumulation time. The radar signals in each time period are defined as a data block, and there are N radar echoes in each data block;

[0009] Step 2: Accumulate the data blocks, and the specific method is as follows:

[0010] For the total time when the terminal guidance radar receives echoes, M data blocks will be generated. First, perform short-time coherent accumulation on the N radar signals in each data block. After short-time coherent accumulation, M time-frequency data blocks will be obtained. Then, perform long-time coherent accumulation on these M time-frequency data blocks. Finally, obtain the time-frequency distribution signal through maximum detection.

[0011] Step 3: Update the data in the way of a sliding window, and the specific method is as follows:

[0012] After the initial accumulation, there are M data blocks. However, due to the real-time requirement of the terminal guidance radar for data, the accumulated echo data needs to be updated. Taking the M data blocks of the initial accumulation as a sliding window, the data blocks within the sliding window are from 1 to M. When the next data block M + 1 arrives, through the movement of the sliding window, the first data block is removed and the new data block is added. At this time, the data blocks within the sliding window are from 2 to M + 1. At the same time, when the sliding window is updated, the data of the 2nd to Mth data blocks in the sliding window are retained, and only the data of the newly arrived (M + 1)th data block are calculated, realizing the long-time accumulation of data and the requirement of data rate.

[0013] 2. A target detection method based on micro-motion sliding window and agile long-time coherent accumulation, characterized in that the data update by means of a sliding window specifically includes the following technical measures:

[0014] Step 1: Take N radar echo signals as a frame of signals, perform the first-level coherent accumulation of short-time Fourier transform, and obtain the data S1 m (k), where k represents the range cell and m represents the number of frames;

[0015] Step 2: Using velocity and acceleration as compensation parameters, with a certain step, create a same cell array of size I×J for each echo, where the number of velocity search values is I and the number of acceleration search values is J;

[0016] Step 3: Construct a cell array data(i,j,m) of size M×I×J, where i represents the i-th velocity search parameter, j represents the j-th acceleration search parameter, and m represents the frame number, i = 1, 2, …, I, j = 1, 2, …, J, m = 1, 2, …, M; The data S m (k) after displacement compensation with parameter velocity v i and phase compensation with acceleration a j The signal S ijm (k), is correspondingly stored in the cell array at the i-th row and j-th column data(i,j,m);

[0017] Step 4: Extract the signals compensated with the same compensation parameters from the cell arrays from 1 to M, and construct a two-dimensional matrix S ij (k,m), where k represents the range cell and m represents the frame number, m = 1, 2, …, M;

[0018] Step 5: For all data matrices S ij (k,m), perform secondary coherent integration using the Fourier transform to obtain the signal S2 ij (k,m), and obtain the maximum value by detecting the maximum value of a certain parameter v I and a J The corresponding signal S2 IJ (k,m) to obtain the maximum value, then take this maximum value as the output of the coherent integration.

[0019] Step 6: When the data of the M + 1-th frame enters, delete the cell array data(i,j,1) corresponding to the 1-st frame data, add the data of the M + 1-th frame data(I,J,M + 1), and for the corresponding data S ijm (k) of the 2-nd to M-th frames, no longer perform compensation operations, only need to compensate the data of the M + 1-th frame to obtain the compensated signal S ijM+1 (k), and store them into the cell array data(i,j,M + 1) respectively. Slide the cell array, regard the data of the 2-nd to M + 1-th frames as the new 1-st to M-th frame data, and return to Step 5.

[0020] The beneficial effects of the present invention are:

[0021] Compared with the prior art, the high PRF radar high-speed maneuvering target detection method based on multi-hypothesis fuzzy matching Radon transform described in this technical solution has the following beneficial effects:

[0022] (1) This method uses a hybrid accumulation method that combines short-time Fourier transform within a frame and long-time compensation for inter-frame coherent accumulation to reduce the algorithm complexity of long-time coherent accumulation for hypersonic missile-borne radars.

[0023] (2) The data for target detection is updated quickly by means of a micro-motion sliding window, which increases the real-time performance of the algorithm and ensures the data rate requirement for target detection by the terminal guidance radar of hypersonic missiles.

[0024] As described above, it is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person familiar with the technology can extend to other modifications, changes and applications within the technical scope disclosed by the present invention, and all should be covered within the scope of the present invention. Description of the Drawings

[0025] Appendix Figure 1 The flowchart of the method steps of the present invention;

[0026] Appendix Figure 2 is a schematic diagram of the sliding window; Specific Embodiment

[0027] Next, in combination with the drawings, the technical solution of the present invention will be described in detail. Referring to Appendix Figure 1 , the specific steps of the present invention include:

[0028] Step 1: Group the radar measurement echo signals, and the specific method is as follows:

[0029] The radar signals determine an accumulation time period according to the accumulation time. The radar signals of each time period are defined as a data block, and there are N radar echoes in each data block.

[0030] Step 2: Accumulate the data blocks, and the specific method is as follows:

[0031] For the total time when the terminal guidance radar receives echoes, M data blocks will be generated. First, perform short-time coherent accumulation on the N radar signals in each data block. After short-time coherent accumulation, M time-frequency data blocks will be obtained. Then, perform long-time coherent accumulation on these M time-frequency data blocks. Finally, obtain the time-frequency distribution signal through maximum detection.

[0032] Step 3: Update the data by means of a sliding window, and the specific method is as follows:

[0033] After the initial accumulation, there are M data blocks. However, due to the real-time requirement of the terminal guidance radar for data, the accumulated echo data needs to be updated. Taking the M data blocks of the initial accumulation as a sliding window, the data blocks within the sliding window are numbered from 1 to M. When the next data block M+1 arrives, through the movement of the sliding window, the first data block is removed and the new data block is added. At this time, the data blocks within the sliding window are numbered from 2 to M+1. At the same time, when updating the sliding window, the data from the 2nd to the Mth data blocks in the sliding window are retained, and only the data of the newly arrived (M+1)th data block is calculated to meet the requirements of long-term data accumulation and data rate. The specific technical measures for updating data through the sliding window are as follows:

[0034] Step 1: Taking N radar echo signals as a frame of signals, perform the first-level coherent accumulation of short-time Fourier transform to obtain the data S1 m (k), where k represents the range cell and m represents the number of frames;

[0035] Step 2: Using velocity and acceleration as compensation parameters, with a certain step, establish a cell array of the same size I×J for each echo, where the number of velocity search values is I and the number of acceleration search values is J;

[0036] Step 3: Construct a cell array data(i,j,m) of size M×I×J, where i represents the ith velocity search parameter, j represents the jth acceleration search parameter, m represents the number of frames, i = 1, 2, …, I, j = 1, 2, …, J, m = 1, 2, …, M; The data S m (k) of the mth frame, after displacement compensation with velocity v i and phase compensation with acceleration a j The signal S ijm (k), is correspondingly stored in the cell array at the ith row and jth column data(i,j,m);

[0037] Step 4: Extract the signals compensated with the same compensation parameters from the cell arrays numbered from 1 to M to construct a two-dimensional matrix S ij (k,m), where k represents the range cell and m represents the number of frames, m = 1, 2, …, M;

[0038] Step 5: Perform the second-level coherent accumulation on all data matrices S ij (k,m) using Fourier transform again to obtain the signal S2 ij (k,m). By detecting the maximum value, a certain parameter v I and a J The corresponding signal S2 IJ (k,m) obtains the maximum value, and then this maximum value is taken as the output of the coherent accumulation.

[0039] Step 6: When the (M + 1)-th frame of data enters, delete the cell array data(i,j,1) corresponding to the 1st frame of data, add the (M + 1)-th frame data data(I,J,M + 1), and correspond with the data S of the previous 2 to M frames ijm (k) No longer perform compensation operations, only need to compensate the (M + 1)-th frame of data to obtain the compensated signal S ijM+1 (k), and store them into the cell array data(i,j,M + 1) respectively. Perform a sliding window translation on the cell array, regard the 2 to (M + 1)-th frame of data as the new 1 to M frames of data, and return to Step 5.

Claims

1. A target detection method based on agile long-time coherent accumulation with a micro-motion sliding window, characterized in that it includes the following technical measures: Step 1: Group the radar measured echo signals, and the specific method is as follows: The radar signals determine an accumulation time period according to the accumulation time. The radar signals in each time period are defined as a data block, and there are N radar echoes in each data block; Step 2: Accumulate the data blocks, and the specific method is as follows: For the total time when the terminal guidance radar receives echoes, M data blocks will be generated. First, perform short-time coherent accumulation on the N radar signals in each data block. After short-time coherent accumulation, M time-frequency data blocks will be obtained. Then, perform long-time coherent accumulation on these M time-frequency data blocks. Finally, obtain the time-frequency distribution signal through maximum detection; Step 3: Update the data in the form of a sliding window, and the specific method is as follows: After the initial accumulation, there are M data blocks. However, due to the real-time requirement of the terminal guidance radar for data, the accumulated echo data needs to be updated. Take the M data blocks of the initial accumulation as a sliding window. The data blocks within the sliding window are from 1 to M. When the next data block M + 1 arrives, through the movement of the sliding window, the first data block is removed and the new data block is added. At this time, the data blocks within the sliding window are from 2 to M + 1. At the same time, when performing sliding window update, retain the data of the 2nd to Mth data blocks in the sliding window, and only calculate the data of the newly arrived (M + 1)th data block to meet the requirements of long-time data accumulation and data rate.

2. The target detection method based on agile long-time coherent accumulation with a micro-motion sliding window according to claim 1, characterized in that updating the data in the form of a sliding window specifically includes the following technical measures: Step 1: Take N radar echo signals as a frame of signals, perform the first-level coherent accumulation of short-time Fourier transform, and obtain data S1 m (k), where k represents the range cell and m represents the number of frames; Step two: Use speed and acceleration as compensation parameters, and with a certain step, establish a cell array of the same size of I×J for each echo, where the number of speed search is I and the number of acceleration search is J; Step 3: Construct a cell array data(i,j,m) of size M×I×J, where i represents the i-th velocity search parameter, j represents the j-th acceleration search parameter, and m represents the frame number, with i = 1, 2, …, I, j = 1, 2, …, J, and m = 1, 2, …, M; the data S m (k) after displacement compensation with parameter velocity v i and acceleration a j phase compensation is the signal S ijm (k), which is correspondingly stored in data(i,j,m) at the i-th row and j-th column of the cell array; Step 4: Extract the signals compensated by the same compensation parameters in 1 to M cell arrays, and construct a two-dimensional matrix S ij (k, m), where k represents the range cell and m represents the number of frames, m = 1, 2, …, M; Step Five: For all data matrices S ij (k,m), perform second-level coherent integration using Fourier transform to obtain signal S2 ij (k,m), and obtain a certain parameter v through the detection of the maximum value I and a J For the corresponding signal S2 IJ (k,m), if the maximum value is obtained, then take this maximum value as the output of the coherent integration; Step 6: When the (M + 1)-th frame of data enters, delete the cell array data(i, j, 1) corresponding to the 1st frame of data, add the (M + 1)-th frame data data(I, J, M + 1), and correspond with the data S of the previous 2nd to M-th frames ijm (k) No longer perform compensation operations, only need to compensate the (M + 1)-th frame of data to obtain the compensated signal S ijM+1 (k), and store them into the cell array data(i, j, M + 1) respectively. Perform a sliding window translation on the cell array, regard the 2nd to (M + 1)-th frames of data as the new 1st to M-th frames of data, and return to Step 5.

Images