A high-repetition-frequency radar signal processing method

Abstract

The application discloses a high-repetition-frequency radar signal processing method, and belongs to the technical field of radars. The method comprises the following steps: performing MTD processing on each pulse group of echoes, extracting target information of each pulse group, performing constant false alarm rate detection on frequency domain data, detecting targets, prolonging the length of detection points at the edge of a transmission shadow area, assuming and marking the detection points as targets, improving the detection capability of edge targets, applying M / N criteria to perform distance deblurring, ensuring the accuracy of detection results, removing false alarms through a connected area decision, finally using a centroid method to condense point traces, and outputting accurate target positions. The technical problem of improving the detection capability of a high-repetition-frequency or single-carrier-frequency radar system is solved. The application reduces the influence of the shadow area, improves the detection capability of targets, can effectively reduce false alarms, ensures the accuracy of target detection, and is relatively simple in algorithm implementation, low in calculation amount, and suitable for practical engineering application under a high-repetition-frequency or single-carrier-frequency radar system.

Description

Technical Field

[0001] The present invention belongs to the field of radar technology, and in particular relates to a high repetition rate radar signal processing method. Background Art

[0002] Existing high-repetition-rate pulse radars typically use linear frequency modulation (LFM) or single-carrier frequency (SCF) signals as their transmission signals. LFM signals have a high time-bandwidth product (TWP), which provides high gain when processed with matched filters. This effectively resolves the conflict between the radar's radiated power and range resolution, making them suitable for long-range target detection. However, in certain specific applications (such as short-range warning radars), the pulse repetition period is very short, resulting in a smaller TWP of the transmitted signal, which significantly reduces the pulse compression effect. Therefore, a single-carrier frequency signal with a narrower time width is typically used. In these cases, the target echo signal may appear at multiple adjacent detection points in distance, making resolution difficult.

[0003] Deficiencies of existing technology:

[0004] Small time-bandwidth product: In high-repetition-rate applications, when the pulse repetition period is short, the time-bandwidth product becomes smaller when using a single-carrier signal, and there is almost no pulse compression gain, which affects the target detection performance.

[0005] Target echo ambiguity: The echo of a single-frequency signal may be relatively smooth in the range dimension, resulting in multiple continuous detection points in the detected target echo, which poses a challenge to the deambiguation of high-repetition-rate radar.

[0006] Transmission shielding zone problem: In high-repetition-rate radars, due to the radar blind spots caused by the transmitted signals, multiple "shielded zones" may be formed within certain distance ranges. Targets in these areas cannot be detected or may be lost, especially targets at the edge of the shielding zones. It is difficult to effectively reduce this impact in the existing processing flow. Summary of the Invention

[0007] The purpose of the present invention is to provide a high repetition rate radar signal processing method, which solves the technical problem of improving the detection capability of a high repetition rate or single carrier frequency radar system.

[0008] To achieve the above objectives, the present invention adopts the following technical inventions:

[0009] A high repetition rate radar signal processing method comprises the following steps:

[0010] Step 1: The MTD module receives the IQ echo data sampled by AD, performs MTD coherent processing on each pulse group, and uses the FFT method to perform frequency domain conversion;

[0011] Step 2: The data processing module modulates the MTD results of each pulse group in turn, and uses the frequency dimension to perform CFAR constant false alarm rate detection to determine whether the target exists;

[0012] Step 3: The compensation module performs masked area compensation on the detection results of the single pulse group: searching at the edge of the masked area, if a detection point is found at the edge of the masked area, and the continuous distance unit is greater than or equal to half the pulse width, it is assumed that the target is partially masked. The detection point is extended along the masked area, and the length of the detection point is padded to the pulse width. This detection point is assumed to be a complete target and marked.

[0013] Step 4: The defuzzification module performs distance defuzzification on the result obtained in step 3, and performs M / N detection to screen out valid target points;

[0014] Step 5: The target judgment module uses the valid target points obtained in step 4 to perform connectivity zone judgment in the distance dimension to eliminate false alarms or interference and ensure the accuracy of the detection results;

[0015] Step 6: The target judgment module condenses the target points using the centroid method, calculates the target amplitude, and outputs the result.

[0016] Preferably, when executing step 1, the MTD processing module receives IQ echo data from AD sampling, divides it into pulse groups, each pulse group consists of multiple pulse periods, performs FFT transformation on each pulse group, and converts it into frequency domain data to facilitate subsequent signal processing and target detection.

[0017] Preferably, when executing step 2, the following steps are specifically included:

[0018] Step 2-1: Modulate the result of MTD coherent processing to obtain signal amplitude information;

[0019] Step 2-2: Use the CFAR algorithm to detect the target in the frequency dimension, determine whether the echo signal exceeds the set false alarm threshold, and identify the potential target location.

[0020] Preferably, when executing step 3, the following steps are specifically included:

[0021] Step 3-1: For each detected target signal, extend it in the range dimension and compare it with the obscured area based on its position and duration;

[0022] Step 3-2: If the duration of the detection point is greater than or equal to half of the pulse width, it is considered as a hypothetical target at the edge of the masked area and compensation processing is performed;

[0023] Step 3-3: Extend the detection point of the hypothetical target along the direction of the occluded area and mark it as the compensation target.

[0024] Preferably, when executing step 4, the robust Chinese remainder theorem is used to defuzzify the signal and determine the actual distance of each target. When performing M / N detection, it is ensured that the number of detection points of each target meets the preset M / N criterion. If the compensation point exceeds the limit, the target will be eliminated.

[0025] Preferably, when executing step 5, the specific steps are as follows:

[0026] Step 5-1: Determine whether the valid target point is a real target and use the minimum pulse width A to determine the distance connection area;

[0027] Step 5-2: If the number of adjacent detection points is less than A×2 / 3, the target is determined to be a false alarm; otherwise, it is retained.

[0028] The high-repetition-rate radar signal processing method disclosed in the present invention solves the technical problem of improving the detection capability of high-repetition-rate or single-carrier-frequency radar systems. By compensating for targets at the edge of a shielded area, the present invention reduces the impact of the shielded area and improves target detection capability. Through M / N detection and judgment, false alarms can be effectively reduced, ensuring the accuracy of target detection. In high-repetition-rate and single-carrier-frequency radar systems, the improvements of the present invention make them more practical in specific scenarios. The algorithm is relatively simple to implement, has a low computational complexity, and relies primarily on logical judgment, making it suitable for practical engineering applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is the main flow chart of the present invention;

[0030] Figure 2 It is a schematic diagram of the impact of the emission shielding area of ​​the present invention;

[0031] Figure 3 is a schematic diagram of the assumed compensation of the emission shielding area of ​​the present invention;

[0032] Figure 4 Schematic diagram of M / N detection and false target determination according to the present invention;

[0033] Figure 5 It is the data before the dot trace condensation of the present invention;

[0034] Figure 6 This is the data after point trace aggregation and filtering of the present invention. DETAILED DESCRIPTION

[0035] Depend on Figures 1-6 A high repetition rate radar signal processing method shown includes the following steps:

[0036] Step 1: The MTD module receives the IQ echo data sampled by AD, performs MTD coherent processing on each pulse group, and uses the FFT method to perform frequency domain conversion;

[0037] The MTD processing module receives IQ echo data from AD sampling and divides it into pulse groups. Each pulse group consists of multiple pulse cycles. FFT transform is performed on each pulse group to convert it into frequency domain data for subsequent signal processing and target detection.

[0038] In this embodiment, the radar operates in high repetition rate mode. Each frame of signal consists of multiple staggered pulse groups, each of which contains several pulse periods. These pulse groups and pulses have different period and frequency characteristics. The specific parameters are as follows:

[0039] Number of pulse groups N: One frame signal contains N pulse groups.

[0040] Number of pulses in a pulse group: The number of pulses contained in the i-th pulse group is Ni.

[0041] Pulse period T: The pulse period of each pulse group is T1, T2, ..., TN.

[0042] Pulse repetition frequency F: The pulse repetition frequency of the i-th pulse group is F1, F2, ..., FN.

[0043] Pulse width Pt: The transmitted pulse width of each pulse in the i-th pulse group is Pt1, Pt2, ..., PtN respectively.

[0044] AD sampling rate Fs: The data sampling rate is Fs, which is used to sample the echo signal.

[0045] Receiving IQ data: The MTD (pulse repetition frequency time delay) module receives IQ (in-phase / quadrature) echo data from the front end after AD sampling. Each frame of the signal contains N pulse groups, each of which consists of multiple pulses.

[0046] Pulse group processing: After receiving a complete pulse group, the MTD module converts the pulse group from the time domain to the frequency domain, that is, performs a fast Fourier transform (FFT). For each pulse group, the number of FFT points is Ni, where Ni is the number of pulses in the i-th pulse group.

[0047] MTD processing can convert time domain signals into frequency domain signals through FFT transformation, which facilitates further target detection and analysis.

[0048] In this embodiment, the MTD module receives the IQ data collected by the front end, and after receiving a complete pulse group, performs MTD processing on the entire pulse group. The specific method is to perform (N1, N2, ..., NN) point FFT in the slow time dimension.

[0049] Step 2: The data processing module modulates the MTD results of each pulse group in turn, and uses the frequency dimension to perform CFAR constant false alarm rate detection to determine whether the target exists. The specific steps include the following:

[0050] Step 2-1: Modulate the result of the MTD coherent processing to obtain signal amplitude information; after the MTD processing, perform a modulus operation on the IQ data of each pulse group to obtain the amplitude information of the pulse group in the frequency dimension.

[0051] Step 2-2: Use the CFAR algorithm to detect the target in the frequency dimension, determine whether the echo signal exceeds the set false alarm threshold, and identify the potential target location.

[0052] In this embodiment, the obtained amplitude information is input into the CFAR (Constant False Alarm Rate) detection module. The CFAR algorithm determines whether a frequency point exceeds a set false alarm threshold, thereby detecting a target. A detection result of "0" indicates that the target was not detected at that location; a detection result of "1" indicates that the target was detected at that location. CFAR detection can detect target signals in a certain amount of background noise, ensuring detection accuracy.

[0053] Step 3: The compensation module performs masked area compensation on the detection results of the single pulse group: searching at the edge of the masked area, if a detection point is found at the edge of the masked area, and the continuous distance unit is greater than or equal to half the pulse width, it is assumed that the target is partially masked. The detection point is extended along the masked area, and the length of the detection point is padded to the pulse width. This section of detection points is assumed to be a complete target and marked. The specific steps include the following:

[0054] Step 3-1: For each detected target signal, extend it in the range dimension and compare it with the obscured area based on its position and duration;

[0055] In this embodiment, the specific extension method is: the maximum detectable distance unit X of the radar and the current pulse repetition period TN are used to calculate the number of distance units for each pulse. The specific calculation method is as follows:

[0056] The number of distance units per pulse is: number of distance units = TN × Fs.

[0057] To cover a wider detection range, all distance cells of each pulse are replicated Y times and connected end to end to form an extended detection result. The length of the extended result is L, satisfying L = Y × TN × Fs, and L ≥ X.

[0058] The extension operation ensures that the detection results can cover sufficient distance dimensions, so that more complete detection results can be obtained when processing occluded areas.

[0059] The detection results of length L after range dimension extension are compensated for the "emission shielding area" hypothesis, which contains Y "emission shielding areas". These areas may not be able to detect targets due to blind spots or interference of the transmission signal. The edges of these shielding areas are compensated.

[0060] Step 3-2: If the duration of the detection point is greater than or equal to half of the pulse width, it is considered as a hypothetical target at the edge of the masked area and compensation processing is performed;

[0061] In this embodiment, the range cells on both sides of the obscured area are checked for detected targets. If a target signal is present, the next step of compensation is performed. If the duration of the detection point (i.e., the duration of the target's presence in the area) is greater than or equal to half the pulse width, the detection point is assumed to be a target.

[0062] Specifically, the detection point duration is determined. For example, the transmission signal duration distance unit is PtN×Fs. If the detection point duration is greater than or equal to PtN×Fs / 2, it is assumed that the detection point is the target.

[0063] Step 3-3: Extend the detection point of the hypothetical target along the direction of the occluded area and mark it as the compensation target.

[0064] In this embodiment, it is assumed that the detection point of the target is compensated along the direction of the occlusion area, that is, the length of the detection point is extended to the occlusion area, and the detection result is marked as 1. The extended length is equal to PtN×Fs, which is compensated to the length of a normal target. The distance unit compensated in the occlusion area is marked and judged later.

[0065] The compensation operation ensures that the target at the edge of the occlusion area can be effectively detected, making up for the detection loss caused by the emission occlusion area.

[0066] Step 4: The defuzzification module performs distance defuzzification on the result obtained in step 3, and performs M / N detection to screen out valid target points;

[0067] In this embodiment, the robust Chinese remainder theorem is used to defuzzify the signal and determine the actual distance of each target. When performing M / N detection, it is ensured that the number of detection points for each target meets the preset M / N criterion. If the compensation points exceed the limit, the target will be eliminated.

[0068] In this embodiment, the specific steps of distance deambiguation and M / N detection are as follows:

[0069] Step 4-1: Align the detection results, specifically aligning the detection results of all pulse groups to form a two-dimensional matrix; each row of the matrix represents a detected distance unit, and each column represents a different pulse group;

[0070] Specifically, the detection results of the N pulse groups after distance extension are aligned to form a two-dimensional matrix. The length of each row of the matrix is ​​the maximum detectable distance unit of the radar, which is X. The detection results greater than X are truncated to the length of X. The length of each column of the matrix is ​​N, corresponding to N pulse groups.

[0071] Step 4-2: Count the detection results, specifically counting the detection results of N pulse groups in each range unit; if M pulse groups out of N pulse groups in a range unit detect a target, it is considered that the range unit has a target;

[0072] Specifically, the detection results of the N pulse groups in each distance unit are counted, and the detection result values ​​of the N pulse groups in the current distance unit are added up.

[0073] Step 4-3: M / N criterion: Specifically, for each detection point, determine whether it meets the M / N criterion; if the detection result meets the criterion, it is considered that there is a target in the distance unit, and the ambiguity resolution and target detection are completed;

[0074] Specifically, if the statistical value of the detection result is ≥M, it means that M pulse groups out of N pulse groups have detected results, then it is considered that there is a target in the distance unit, and the row coordinates of the current target in the matrix correspond to the real distance unit of the target, and the distance deambiguation and detection are completed at the same time.

[0075] Step 4-4: Review the compensation points. Specifically, after defuzzification, the detection points that need to be compensated need to be reviewed to ensure that the number of compensation points does not exceed the preset limit. When the compensation points exceed the limit, the defuzzified points are invalidated.

[0076] Specifically, the target result detected after defuzzification is reviewed. According to the mark in step 3-3, the number of compensated points in the M pulse groups with targets cannot exceed 2. When the number of compensated points exceeds 2, the current defuzzified point is invalid.

[0077] After the distance defuzzification is completed, a detection result of length X is formed, where 0 indicates that there is no target in the distance unit, and 1 indicates that there is a target in the distance unit, and the distance unit corresponds to the actual position of the target.

[0078] The M / N criterion is used to improve the accuracy of target detection and eliminate false detections and false alarms.

[0079] Step 5: The target judgment module uses the valid target points obtained in step 4 to perform connectivity zone judgment in the distance dimension to eliminate false alarms or interference and ensure the accuracy of the detection results. The specific steps are as follows:

[0080] Step 5-1: Determine whether the valid target point is a real target and use the minimum pulse width A to determine the distance connection area;

[0081] In this embodiment, valid target points are processed by traversing the detection results, specifically traversing from left to right. When a detection point is found, it is counted whether there is a detection point in the adjacent distance unit to the right, and the number of adjacent detection points is counted.

[0082] Step 5-2: If the number of adjacent detection points is less than A×2 / 3, the target is determined to be a false alarm; otherwise, it is retained.

[0083] In this embodiment, the number of adjacent detection points is counted. Assuming that the minimum pulse width of the transmitted signal of N pulse groups is A, if the number of adjacent detection points is less than 2 / 3 of the minimum pulse width A, the target is considered to be a false alarm or interference; if the number of adjacent detection points is less than the threshold value, that is, A×2 / 3, the detection point is marked as no target and no processing is performed.

[0084] Step 6: The target judgment module condenses the target points using the centroid method, calculates the target amplitude, and outputs the result.

[0085] For the detection points after the connection area judgment, the centroid method is used to perform agglomeration processing. For each target point, its amplitude value is calculated, which is determined by the average value of N pulse groups or the amplitude values ​​of adjacent pulses.

[0086] Point aggregation helps to accurately locate the target and reduce the detection error through averaging processing, ultimately outputting high-quality target detection results.

[0087] The high-repetition-rate radar signal processing method disclosed in the present invention solves the technical problem of improving the detection capability of high-repetition-rate or single-carrier-frequency radar systems. By compensating for targets at the edge of a shielded area, the present invention reduces the impact of the shielded area and improves target detection capability. Through M / N detection and judgment, false alarms can be effectively reduced, ensuring the accuracy of target detection. In high-repetition-rate and single-carrier-frequency radar systems, the improvements of the present invention make them more practical in specific scenarios. The algorithm is relatively simple to implement, has a low computational complexity, and relies primarily on logical judgment, making it suitable for practical engineering applications.

Claims

1. A high repetition rate radar signal processing method, characterized in that: The steps include: Step 1: The MTD module receives the IQ echo data sampled by AD, performs MTD coherent processing on each pulse group, and uses the FFT method to perform frequency domain conversion; Step 2: The data processing module modulates the MTD results of each pulse group in turn, and uses the frequency dimension to perform CFAR constant false alarm rate detection to determine whether the target exists; Step 3: The compensation module performs masked area compensation on the detection results of the single pulse group: searching at the edge of the masked area, if a detection point is found at the edge of the masked area, and the continuous distance unit is greater than or equal to half the pulse width, it is assumed that the target is partially masked. The detection point is extended along the masked area, and the length of the detection point is padded to the pulse width. This detection point is assumed to be a complete target and marked. Step 4: The defuzzification module performs distance defuzzification on the result obtained in step 3, and performs M / N detection to screen out valid target points; Step 5: The target judgment module uses the valid target points obtained in step 4 to perform connectivity zone judgment in the distance dimension to eliminate false alarms or interference and ensure the accuracy of the detection results; Step 6: The target judgment module condenses the target points using the centroid method, calculates the target amplitude, and outputs the result.

2. The high repetition rate radar signal processing method according to claim 1, wherein: When executing step 1, the MTD processing module receives IQ echo data from AD sampling, divides it into pulse groups, each pulse group consists of multiple pulse periods, performs FFT transformation on each pulse group, and converts it into frequency domain data to facilitate subsequent signal processing and target detection.

3. The high repetition rate radar signal processing method according to claim 1, wherein: When executing step 2, the specific steps include: Step 2-1: Modulate the result of MTD coherent processing to obtain signal amplitude information; Step 2-2: Use the CFAR algorithm to detect the target in the frequency dimension, determine whether the echo signal exceeds the set false alarm threshold, and identify the potential target location.

4. The high repetition rate radar signal processing method according to claim 1, wherein: When executing step 3, the specific steps include: Step 3-1: For each detected target signal, extend it in the range dimension and compare it with the obscured area based on its position and duration; Step 3-2: If the duration of the detection point is greater than or equal to half of the pulse width, it is considered as a hypothetical target at the edge of the masked area and compensation processing is performed; Step 3-3: Extend the detection point of the hypothetical target along the direction of the occluded area and mark it as the compensation target.

5. The high repetition rate radar signal processing method according to claim 1, wherein: When executing step 4, the robust Chinese remainder theorem is used to defuzzify the signal and determine the actual distance of each target. When performing M / N detection, ensure that the number of detection points for each target meets the preset M / N criterion. If the compensation points exceed the limit, the target will be eliminated.

6. The high repetition rate radar signal processing method according to claim 1, wherein: When executing step 5, the specific steps are as follows: Step 5-1: Determine whether the valid target point is a real target and use the minimum pulse width A to determine the distance connection area; Step 5-2: If the number of adjacent detection points is less than A×2 / 3, the target is determined to be a false alarm; otherwise, it is retained.

Images