Large grazing angle data unit alignment method based on mean square and maximum

Abstract

The invention relates to a large grazing angle data unit alignment method based on mean square and maximum, and belongs to the technical field of radar signal processing. The method comprises the following steps: 1) calculating average amplitude; 2) selecting reference frame data; 3) calculating the number of offset distance units; and step 4) data offset. According to the method, the problem that the target distance units are difficult to align is solved, and the data distance units of the mobile platform are aligned to adapt to various inter-frame accumulation methods.

Description

Technical Field

[0001] The present invention relates to a large ground-grazing angle data unit alignment method based on maximum mean square sum, and belongs to the technical field of radar signal processing. Background Art

[0002] Top-down attack is one of the important means of striking marine targets. Radar target detection at large ground-grabbing angles is the core technology supporting this method. However, due to the extremely low signal-to-noise ratio at large ground-grabbing angles, the detection performance is low if only the current frame is used for detection. Inter-frame accumulation can greatly improve the detection performance. When the radar is working at a large ground-grabbing angle, the platform is in motion, and the distance unit where the target is located is constantly changing, making inter-frame accumulation more difficult. There are many methods for inter-frame accumulation at present, including the use of inter-frame temporal features, the use of historical frame distribution prior information, etc. These methods can greatly improve the target signal-to-noise ratio, but these methods are all implemented in a static scene between the target and the radar. For the case of a moving platform, some scholars currently use Doppler information to track the target to achieve accumulation, but due to the limited Doppler resolution and the influence of wave speed, this method is not very accurate. Therefore, this patent proposes a large ground-grabbing angle data unit alignment method based on the maximum mean square sum to align the target distance units to adapt to various inter-frame accumulation methods. Summary of the Invention

[0003] The purpose of the present invention is to solve the shortcomings of the above-mentioned prior art and provide a method for aligning large ground-grabbing angle data units based on mean square and maximum, which solves the problem of difficulty in aligning target distance units. It can align the data distance units of the moving platform to adapt to various inter-frame accumulation methods.

[0004] The invention provides a method for aligning large ground-grabbing angle data units based on the mean square sum maximum, which is special in that it includes the following steps:

[0005] Step 1) Calculation of average amplitude: Using the echo data after pulse compression processing, the average amplitude is calculated based on the number of pulses used for radar speed compensation to obtain the reference data required for unit alignment;

[0006] Step 2) Selecting reference frame data: Select the first frame of reference data as the reference frame data. If the radar is not working stably when acquiring the first frame of data, select the first frame of reference data after it is working stably as the reference frame data.

[0007] Step 3) Calculate the number of offset distance units: Set the range of the number of search units, slide the current frame data within the range of the search unit number, and calculate the mean square sum with the reference frame data. The search unit number with the largest mean square sum is the number of offset distance units;

[0008] Step 4) Offset data: The original data corresponding to each frame of reference data is shifted by the number of offset distance units to obtain the aligned data.

[0009] Preferably, the specific steps for calculating the average amplitude in step 1) are:

[0010] The velocity compensation is performed once every 32 or 64 pulses for the large ground-grazing angle data. The compensated data are coherent. In order to maintain the coherence to meet the needs of subsequent data processing, each set of coherent data is used as the minimum number of units for distance alignment, and the average amplitude is used as a reference for calculating the number of offset distance units for each set of data. The calculation formula is shown in (1):

[0011]

[0012] x m (c) represents radar echo data, i represents the number of coherent data groups, c represents the number of range cells, and N represents the number of pulses in each coherent data group. When the average amplitude is used as a reference, a single abnormal data pulse will not have a significant impact on the overall data, and the robustness is relatively strong.

[0013] Preferably, the specific steps of step 3) are:

[0014] (1) Setting the search unit number range: Select an integer interval symmetrical about the zero point. The size of the interval is related to the radar tracking capability. When the radar tracking capability is strong, the target range unit fluctuates less, and the required search unit number range is smaller. Conversely, a larger search range is required. Since the current frame reference data needs to subtract the search unit number when performing the mean square calculation in the next step to enable it to slide in the reference frame data, the upper limit of the interval should be smaller than the number of distance units from the target range unit to the data boundary.

[0015] (2) Calculate the mean square sum by traversing the search unit: Crop the reference data of the current frame. The number of cropped units on the left and right ends are the absolute values of the lower and upper bounds of the search unit number range, respectively. The cropped data is traversed and slid in the reference frame data with the center distance unit of the reference frame as the starting point. The mean square sum is calculated as follows:

[0016]

[0017] Among them, AA S1 Represents the numerical vector of the reference frame data within the sliding window, Represents the transpose of the cropped current frame reference data numeric vector.

[0018] (3) Calculate the maximum number of offset units: Calculate the maximum value of the mean square sum calculated by traversing the search unit for each reference data frame. The number of offset units corresponding to the maximum mean square sum is the maximum number of offset units.

[0019] Compared with the existing technology, the large ground-grazing angle data unit alignment method based on the mean square sum maximum described in this solution has the following beneficial effects:

[0020] (1) The method proposed in the present invention utilizes the characteristic that the relative positions of the target and several strong scattering points in the radar detection scene remain unchanged, and aligns the distance units where the target is located based on the mean square maximum principle. The calculation is simple and the practicability is strong.

[0021] (2) Using the average amplitude feature as the reference data will not affect the alignment result due to the abnormality of a single pulse, and the robustness is strong.

[0022] (3) The proposed method was tested using several sets of Yantai measured flight data. All the data showed that the proposed method has a good effect on target distance unit alignment. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 This is a flowchart of a method for aligning large ground-grabbing angle data units based on mean square sum maximum according to the present invention;

[0024] Figure 2 is the echo graph of alignment mismatch; in the figure, (a) is the echo graph before alignment, and (b) is the echo graph after alignment;

[0025] Figure 3 This is the effect diagram of realignment;

[0026] Figure 4 This is the second set of Yantai flying data unit alignment effect diagram; in the figure, (a) is the Yantai flying data unit effect diagram before alignment, (b) is the Yantai flying data unit effect diagram after alignment;

[0027] Figure 5 This is the third set of Yantai flying data unit alignment effect diagram; in the figure, (a) is the Yantai flying data unit effect diagram before alignment, (b) is the Yantai flying data unit effect diagram after alignment; DETAILED DESCRIPTION

[0028] For better understanding and implementation, a specific implementation detailed description of a method for identifying floating targets on the sea surface based on a recursive graph of the present invention is given below in conjunction with the accompanying drawings. It should be noted that this embodiment introduces the method proposed by the present invention by taking the identification of ship targets and buoy targets as an example.

[0029] A large ground-grabbing angle data alignment method based on maximum mean square sum in this embodiment includes the following steps:

[0030] 1) Calculation of average amplitude

[0031] High-angle hang-and-run data is typically compensated for velocity every 32 or 64 pulses. The compensated data is coherent. To maintain coherence for subsequent data processing, this embodiment uses each set of coherent data as the minimum number of units for distance alignment. This embodiment uses the average amplitude as a reference for calculating the number of offset distance units for each set of data, as calculated using the formula (1).

[0032]

[0033] x m (c) represents radar echo data, i represents the number of coherent data groups, c represents the number of range bins, and N represents the number of pulses in each coherent data group. When the average amplitude is used as a reference, a single abnormal data pulse will not significantly affect the overall data, resulting in stronger robustness.

[0034] 2) Select the reference frame data

[0035] This embodiment aligns the range units based on the fact that the relative positions of radar targets and strong scattering point echoes remain unchanged in a short period of time. Therefore, a reference frame data with relatively obvious targets and strong scattering points (high signal-to-noise ratio) is required as the initial alignment standard. However, when the radar starts working, it may be in a search state or other working mode where it is not stably tracking the target. If the data in this stage is selected as the reference frame data, it will cause alignment mismatch and thus affect the alignment effect. Figure 2 As shown in the figure, Figure (a) is the original data, and Figure (b) is the distance adaptation phenomenon caused by not selecting the stable tracking data as the reference frame for alignment. It can be seen that two bright lines appear in the latter part, and the target is not completely aligned. Therefore, it is necessary to select the starting frame data after stabilization and select the stable tracking data in the latter part as the reference frame data for realignment. Figure 3 As shown, the mismatch phenomenon disappears and the target bright line does not Figure 2 Instead of two bright lines, they are aligned in one distance unit, which also reflects the importance of selecting reference frame data.

[0036] 3) Calculate the number of offset distance units

[0037] After obtaining the reference data and reference frame data in steps 1) and 2), the number of offset range units can be calculated by calculating the maximum mean square. The mechanism is that the radar main lobe beam is always aligned near the target when tracking the target. The relative position of the target and other strong scattering points remains unchanged for a short period of time, which can be matched with the reference frame data, and then the target range unit can be aligned. The specific steps are as follows:

[0038] (1) Setting the search unit number range: Generally, an integer interval symmetrical about the zero point is selected. The size of the interval is related to the radar tracking capability. When the radar tracking capability is strong, the target range unit fluctuates less, and the required search unit number range is smaller. Conversely, a larger search range is required. It is worth noting that the upper limit of the interval should be smaller than the number of distance units from the target range unit to the data boundary, because when performing the mean square calculation in the next step, the current frame reference data will be subtracted by the search unit number to allow it to slide in the reference frame data.

[0039] (2) Calculate the mean square sum by traversing the search unit: Crop the reference data of the current frame. The number of cropped units on the left and right ends are the absolute values of the lower and upper bounds of the search unit number range, respectively. The cropped data is traversed and slid in the reference frame data with the center distance unit of the reference frame as the starting point. The mean square sum is calculated as follows:

[0040]

[0041] Among them, AA S1 Represents the numerical vector of the reference frame data within the sliding window, Represents the transpose of the cropped current frame reference data numeric vector.

[0042] (3) Calculate the maximum number of offset units: Calculate the maximum value of the mean square sum calculated by traversing the search unit for each reference data frame. The number of offset units corresponding to the maximum mean square sum is the maximum number of offset units.

[0043] 4) Offset data

[0044] The target aligned data can be obtained by moving the original data corresponding to each frame of reference data by the maximum offset unit number, except Figure 2 、 Figure 3 The data shown in this example are also aligned with two other sets of Yantai flight data, such as Figure 4 、 Figure 5 As shown in the figure, (a) and (b) represent the original data and the data after mean square alignment, respectively. It can be seen that the radar in these two data sets is in staring mode and tracking mode, respectively. Before alignment, the target range units of both data sets will be offset and fluctuate, which is not conducive to the detection algorithm based on inter-frame accumulation. However, after alignment, the targets in the data are all within the same range unit, providing a data foundation for subsequent detection.

Claims

1. A method for aligning large ground-grazing angle data units based on maximum mean square sum, characterized in that The following steps are involved: Step 1) Calculation of average amplitude: Using the echo data after pulse compression processing, the average amplitude is calculated based on the number of pulses used for radar speed compensation to obtain the reference data required for unit alignment; Step 2) Selecting reference frame data: Select the first frame of reference data as the reference frame data. If the radar is not working stably when acquiring the first frame of data, select the first frame of reference data after it is working stably as the reference frame data. Step 3) Calculate the number of offset distance units: Set the range of the number of search units, slide the current frame data within the range of the search unit number, and calculate the mean square sum with the reference frame data. The search unit number with the largest mean square sum is the number of offset distance units; Step 4) Offset data: The original data corresponding to each frame of reference data is shifted by the number of offset distance units to obtain the aligned data.

2. A method for aligning large ground-grazing angle data units based on the maximum mean square sum according to claim 1, characterized in that The specific steps for calculating the average amplitude in step 1) are: The velocity compensation is performed once every 32 or 64 pulses for the large ground-grazing angle data. The compensated data are coherent. In order to maintain the coherence to meet the needs of subsequent data processing, each set of coherent data is used as the minimum number of units for distance alignment, and the average amplitude is used as a reference for calculating the number of offset distance units for each set of data. The calculation formula is shown in (1): x m (c) represents the radar echo data, i represents the number of groups of coherent data, c represents the number of range units, and N is the number of pulses in each group of coherent data.

3. The method for aligning large ground-grazing angle data units based on the maximum mean square sum according to claim 1, characterized in that The specific steps of step 3) are: (1) Setting the search unit number range: Select an integer interval symmetrical about the zero point. The size of the interval is related to the radar tracking capability. When the radar tracking capability is strong, the target range unit fluctuates less, and the required search unit number range is smaller. Conversely, a larger search range is required. Since the current frame reference data needs to subtract the search unit number when performing the mean square calculation in the next step to enable it to slide in the reference frame data, the upper limit of the interval should be smaller than the number of distance units from the target range unit to the data boundary. (2) Calculate the mean square sum by traversing the search unit: Crop the reference data of the current frame. The number of cropped units on the left and right ends are the absolute values of the lower and upper bounds of the search unit number range, respectively. The cropped data is traversed and slid in the reference frame data with the center distance unit of the reference frame as the starting point. The mean square sum is calculated as follows: Among them, AA S1 Represents the numerical vector of the reference frame data within the sliding window, Represents the transpose of the numerical vector of the cropped current frame reference data; (3) Calculate the maximum number of offset units: Calculate the maximum value of the mean square sum calculated by traversing the search unit for each reference data frame. The number of offset units corresponding to the maximum mean square sum is the maximum number of offset units.

Images