High repetition frequency radar signal processing method

Abstract

The invention discloses a high repetition frequency radar signal processing method, which belongs to the technical field of radars, and comprises the following steps: carrying out MTD processing on each pulse group of an echo, extracting target information of each pulse group, carrying out constant false alarm rate detection on frequency domain data, detecting a target, and supposing and marking the target by prolonging the length of a detection point at the edge of an emission shielding area. The detection capability of an edge target is improved, the M / N criterion is applied to carry out distance ambiguity resolution, the accuracy of a detection result is ensured, false alarms are removed through communication region judgment, and finally a centroid method is used to condense plots and output an accurate target position, so that the technical problem of improving the detection capability of a high-repetition-frequency or single-carrier-frequency radar system is solved. According to the method, the influence of a shielding area is reduced, the target detection capability is improved, false alarms can be effectively reduced, the target detection accuracy is ensured, and under a high repetition frequency and single carrier frequency radar system, the algorithm is relatively simple to implement, the calculated amount is relatively low, and the method is suitable for practical engineering application.

Description

Technical Field

[0001] The 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 usually use linear frequency modulation signals or single carrier frequency signals as transmission signals. Linear frequency modulation signals have a high time-bandwidth product and can provide a high gain when processed by matched filtering, thereby effectively solving the contradiction between the radar's radiation power and distance resolution, and are suitable for long-range target detection. However, in some specific applications (such as short-range warning radars), the pulse repetition period is very short, resulting in a smaller time-bandwidth product of the transmitted signal, which will significantly reduce the pulse compression effect. Therefore, a single carrier frequency signal with a narrower time width is usually used. In these cases, the target echo signal may appear at multiple adjacent detection points in distance, making it difficult to resolve ambiguity.

[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 gain from pulse compression, which affects the detection performance of the target.

[0005] Target echo ambiguity: The echo of a single-carrier frequency signal may be relatively smooth in the distance 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] Transmitter 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 object 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 purpose, 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 the 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 assumption compensation on the detection results of the single pulse group: search 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 of the pulse width, it is assumed that the place is a target with a part of the distance unit 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 the 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 determines the effective target point obtained in step 4 in the distance dimension, removes false alarms or interference, and ensures the accuracy of the detection result;

[0015] Step 6: The target judgment module condenses the target points through the centroid method, calculates the amplitude of the target, 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: Calculate the modulus of 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 there is an echo signal exceeding 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 distance dimension and compare it with the shielding area according to 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 regarded 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 assumed 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 the valid target point to determine whether it 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] A high repetition rate radar signal processing method disclosed in the present invention solves the technical problem of improving the detection capability of a high repetition rate or single carrier frequency radar system. The present invention reduces the influence of the shielding area and improves the detection capability of the target by performing compensation processing on the target at the edge of the shielding area. Through M / N detection and judgment, false alarms can be effectively reduced to ensure the accuracy of target detection. In a high repetition rate and single carrier frequency radar system, the improvement of the present invention makes it more practical in specific scenarios. The algorithm is relatively simple to implement, the calculation amount is low, and it mainly relies on logical judgment, which is suitable for actual 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 influence of the emission shielding area of ​​the present invention;

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

[0032] Figure 4 It is a schematic diagram of M / N detection and false target determination of 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 Figure 1-Figure 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 the 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, divides it into pulse groups, each pulse group consists of multiple pulse cycles, performs FFT transformation on each pulse group, and converts it into frequency domain data to facilitate subsequent signal processing and target detection.

[0038] In this embodiment, the radar operates in high repetition rate mode. Each frame signal consists of multiple staggered pulse groups, each pulse group contains several pulse cycles. These pulse groups and pulses have different period and frequency characteristics, and 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 respectively.

[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 the IQ (In-phase / Quadrature) echo data sampled by AD from the front end. Each frame signal contains N pulse groups, and each pulse group consists of multiple pulses.

[0046] Pulse group processing: After receiving a complete pulse group, the MTD module converts the pulse group from time domain to frequency domain, that is, performs 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 is convenient for 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; specifically, the following steps are included:

[0050] Step 2-1: Modulo the result of MTD coherent processing to obtain signal amplitude information; after MTD processing, modulo 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 there is an echo signal exceeding the set false alarm threshold, and identify the potential target location.

[0052] This embodiment inputs the obtained amplitude information into the CFAR (Constant False Alarm Rate) detection module. The CFAR algorithm is used to determine whether a certain frequency point exceeds the set false alarm threshold, thereby detecting the target: the detection result of "0" indicates that the target is not detected at this position; the detection result of "1" indicates that the target is detected at this position. CFAR detection can detect the target signal in a certain background noise, ensuring the accuracy of the detection.

[0053] Step 3: The compensation module performs masked area assumption compensation on the detection results of the single pulse group: search 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 of the pulse width, it is assumed that the place is a target with a part of the distance unit masked. The detection point is extended along the masked area, and the length of the detection point is filled to the pulse width. This section of the detection point is assumed to be a complete target and marked; specifically, the following steps are included:

[0054] Step 3-1: For each detected target signal, extend it in the distance dimension and compare it with the shielding area according to 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, calculate the number of distance units of each pulse, and the specific calculation method is as follows:

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

[0057] In order to cover a wider detection range, all distance units of each pulse will be copied Y times, 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 with a length of L after the range dimension extension are compensated for the "transmission shielding area", which contains Y "transmission shielding areas". These areas may not be able to detect the target due to blind spots or interference of the transmission signal, and 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 regarded as a hypothetical target at the edge of the masked area and compensation processing is performed;

[0061] In this embodiment, the distance cells on both sides of the shielding area are checked to see if there is a detected target. If there is a target signal, the next step of compensation is performed. If the duration of the detection point (i.e., the existence time of the target in this area) is greater than or equal to half of the pulse width, it is assumed that the detection point is 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 assumed 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 occluded area, that is, the length of the detection point is extended to the occluded 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 occluded area is marked for subsequent judgment.

[0065] The compensation operation ensures that the targets 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 method 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.

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

[0069] Step 4-1: aligning 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 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 a 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 count the detection results of N pulse groups in each distance unit; if M pulse groups out of N pulse groups in a certain distance unit detect a target, it is considered that there is a target in this distance unit;

[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.

[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, and 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 a 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 compensation points in the M pulse groups with targets cannot exceed 2. When the compensation points exceed 2, the current defuzzified point is invalid.

[0077] After the distance defuzzification is completed, a detection result with a length of 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 determines the effective target point obtained in step 4 in the distance dimension, removes false alarms or interference, and ensures the accuracy of the detection result; the specific steps are as follows:

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

[0081] In this embodiment, the valid target point is 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 on 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. It is assumed that the minimum pulse width of the transmitted signal of the 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 through the centroid method, calculates the amplitude of the target, and outputs the result.

[0085] For the detection points after the Unicom area judgment, the centroid method is used to condense them. 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] A high repetition rate radar signal processing method disclosed in the present invention solves the technical problem of improving the detection capability of a high repetition rate or single carrier frequency radar system. The present invention reduces the influence of the shielding area and improves the detection capability of the target by performing compensation processing on the target at the edge of the shielding area. Through M / N detection and judgment, false alarms can be effectively reduced to ensure the accuracy of target detection. In a high repetition rate and single carrier frequency radar system, the improvement of the present invention makes it more practical in specific scenarios. The algorithm is relatively simple to implement, the calculation amount is low, and it mainly relies on logical judgment, which is suitable for actual 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 the 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 assumption compensation on the detection results of the single pulse group: search 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 of the pulse width, it is assumed that the place is a target with a part of the distance unit 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 the 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 determines the effective target point obtained in step 4 in the distance dimension, removes false alarms or interference, and ensures the accuracy of the detection result; Step 6: The target judgment module condenses the target points through the centroid method, calculates the amplitude of the target, and outputs the result.

2. A high repetition rate radar signal processing method as claimed in claim 1, characterized in that: 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. A high repetition rate radar signal processing method as claimed in claim 1, characterized in that: When executing step 2, the specific steps include: Step 2-1: Calculate the modulus of 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 there is an echo signal exceeding the set false alarm threshold, and identify the potential target location.

4. A high repetition rate radar signal processing method as claimed in claim 1, characterized in that: When executing step 3, the specific steps include: Step 3-1: For each detected target signal, extend it in the distance dimension and compare it with the shielding area according to 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 regarded 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 assumed target along the direction of the occluded area and mark it as the compensation target.

5. A high repetition rate radar signal processing method as claimed in claim 1, characterized in that: 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 criteria. If the compensation points exceed the limit, the target will be eliminated.

6. A high repetition rate radar signal processing method as claimed in claim 1, characterized in that: When executing step 5, the specific steps are as follows: Step 5-1: Determine the valid target point to determine whether it 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