Radar data processing method for updating clutter map through sliding window
The method of updating the clutter graph through sliding windows solves the problem of unsatisfactory detection effect caused by clutter interference when detecting drones, and achieves high detection accuracy and stability in a dynamic environment and reduces the false detection rate.
Patent Information
- Application Number
- CN202411889015.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-13
- Publication Date
- 2025-03-11
Smart Images
Figure CN119596267A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of radar clutter suppression, in particular to a radar data processing method for updating a clutter map with a sliding window. Background Art
[0002] With the gradual opening of the country to the low-altitude area, drones are widely used in various fields of society due to their small size, relatively low price and easy operation. However, due to the lack of unified management regulations, many drones are in a "black flying" state. This disordered flight state poses a considerable threat to public safety, airspace management, and even national security. This requires radar to be able to successfully and accurately detect drones, but for a long time, radar detection of drones has not been very ideal, mainly because of excessive clutter interference during the detection process. The clutter map in data processing can effectively reduce the impact of clutter. In a dynamic environment, the clutter map provides a real-time updated background model, enabling the system to adapt to changing environmental conditions. This dynamic update mechanism ensures that the detection algorithm still has high robustness and response speed in complex and changing scenes, which helps to reduce the system's false detection rate. Compared with the detection mode without clutter map, the system using clutter map can more accurately identify the difference between targets and non-targets, thereby maintaining a low false detection rate in complex backgrounds and improving the overall stability of detection. Summary of the invention
[0003] The present invention aims to provide a radar data processing method for updating clutter maps with a sliding window. The clutter map is updated using the idea of a sliding window, and the gradual changes in the environment in a short period of time are effectively captured. The clutter map is continuously updated by the latest data in the window, and the system can track the clutter characteristics with small changes, thereby maintaining a high detection accuracy when the environment gradually changes. The length of the sliding window can be adjusted according to actual conditions to balance the real-time performance and the stability of the clutter map. Short windows are suitable for rapidly changing environments, while longer windows are suitable for relatively stable scenes, so it has high flexibility.
[0004] Step 1: Set the sliding window length to m, the window accumulation number to n, and obtain the scanning data of m×n radar scanning circles;
[0005] Step 2: Set the window cumulative scanning circle number i to 1, set the sliding window r to 0, and establish a clutter map based on the above scanning data;
[0006] Step 3: The radar performs the m×n+ith circle scanning to obtain the scanning data of this circle;
[0007] Step 4: Based on the clutter map established by the radar m×n-circle scanning data and the scanning data after the m×n+i-th circle scanning, clutter detection is performed, and the number of targets in the current cycle is counted, and finally the scanning data that has not been filtered out is output;
[0008] Step 5: If the window cumulative scanning number i=n, set i=1, r plus 1, and update the clutter map based on the i-th round of unfiltered scanning data and the scanning data with a sliding window range of r~m+r. If the window cumulative scanning number i<n, set i plus 1, and repeat steps 3 and 4.
[0009] Step 1: Set the sliding window length to m, the window accumulation number to n, and obtain the scanning data of m×n radar scanning circles;
[0010] Step 2: Set the window cumulative scanning circle number i to 1, set the sliding window r to 0, and establish a clutter map based on the above scanning data. The specific steps include:
[0011] First, divide the radar azimuth of 90 degrees into M equal parts, divide the radar pitch of 32 degrees into N equal parts, and divide the sampled distance unit into X equal parts, M>0, N>0, X>0; secondly, the radar performs m×n circles of scanning and counts the number of points num1, num2,..., num of the current azimuth, pitch, and distance of m windows. m And the echo amplitudes AMP1, AMP2, ..., AMP m , and the calculated current echo amplitude mean newmap is used as the initial detection threshold of the clutter map and the oldmap for the next clutter map update. The following method is used to establish the amplitude clutter map:
[0012]
[0013] Finally, the window cumulative scanning circle number i is set to 1 and the sliding window r is set to 0.
[0014] Step 3: The radar performs the m×n+ith circle scanning to obtain the scanning data of this circle;
[0015] Step 4: Based on the clutter map established by the radar m×n scanning data and the scanning data after the m×n+i scanning, clutter detection is performed, and the number of targets in the current cycle is counted. Finally, the scanning data that has not been filtered out is output, including:
[0016] If the current echo amplitude is greater than A times the current data stored in the clutter map, the target is determined to exist, and the target number targetnum is increased by 1; if the current echo amplitude is less than A times the current data stored in the clutter map, the target is determined to not exist, and the target number targetnum remains unchanged. Finally, the scan data that has not been filtered out is output.
[0017] Step 5: If the cumulative number of window scanning circles \(i = n\), then set \(i = 1\), increment \(r\) by 1, and update the clutter map based on the scanning data that has not been filtered out in \(i\) circles and the scanning data within the sliding window range from \(r\) to \(m + r\). If the cumulative number of window scanning circles \(i < n\), increment \(i\) by 1, and repeat Steps 3 and 4, including:
[0018] If the cumulative number of window scanning circles \(i < n\), then repeat Steps 3 and 4 until the cumulative number of window scanning circles \(i = n\), and update the clutter map, where the value range of \(i\) is \(1\leq i\leq m\);
[0019] The clutter map is updated according to the following steps:
[0020] First, increment the sliding window \(r\) by 1; second, based on the current echo amplitude \(AMP\) of the \(i\)-th circle of radar scanning m+r , the number of targets \(num\) m+r and the scanning data of the \(m\times n\) circles of the radar, iteratively update the amplitude clutter map; third, add the stored current data \(oldmap\) in the clutter map and the echo amplitude mean value of the scanning data within the sliding window range from \(r\) to \(m + r\); finally, obtain the updated clutter map \(newmap\). The echo amplitude mean value \(A\) m+r of the scanning data within the sliding window range from \(r\) to \(m + r\) is calculated by the formula
[0021]
[0022] where \(num\) m+r = \(targetnum\). Then the clutter map update formula is:
[0023]
[0024] where \(oldmap\) is the data stored in the clutter map before update; is the clutter map update coefficient, and its value range is \(newmap\) is the data stored in the clutter map after iterative update, and it is also used as \(oldmap\) for the next clutter map update.
[0025] Compared with the prior art, the remarkable advantages of the present invention are:
[0026] 1) The sliding window method allows real-time calculation of the results within a specific time window while the data stream is continuously updated. By sliding the window, the results of the current clutter map can be quickly updated without recalculating the entire data sequence, improving the efficiency of real-time processing. The sliding window method can save a large amount of computing resources when dealing with large-scale data, which is helpful for target detection and handling of clutter with small fluctuations. Description of the Drawings
[0027] Figure 1It is an implementation flow chart.
[0028] Figure 2 This is a schematic diagram of the sliding window processing process.
[0029] Figure 3 Scanning data before clutter map processing
[0030] Figure 4 The scanned data processed by the present invention DETAILED DESCRIPTION
[0031] In order to facilitate the understanding of the present invention, the present invention will be described more comprehensively with reference to the accompanying drawings. The present invention can be implemented in different forms and is not limited to the embodiments described in the text. On the contrary, the embodiments are provided to make the content disclosed in the present invention more thorough and comprehensive. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.
[0032] See also Figure 1-2 This example provides a radar data processing method for updating a clutter map using a sliding window. The feature of the present invention is that the sliding window concept is used to complete the iterative update of the clutter map, thereby reducing the impact of small fluctuations in clutter and improving the probability of target detection.
[0033] Step 1: Set the sliding window length to m, the window accumulation number to n, and obtain the scanning data of the radar scanning m×n circles:
[0034] Step 1 includes:
[0035] The sliding window length is set to 3 and the window accumulation number is set to 8, and the scanning data of 24 radar scans are obtained.
[0036] Step 2: Set the window cumulative scanning circle number i to 1, the sliding window r to 0, and establish a clutter map based on the above scanning data:
[0037] Step 2 includes:
[0038] First, the clutter map is stored in the form of azimuth plus pitch plus distance, dividing the radar azimuth of 90 degrees into 24 equal parts, the radar pitch of 32 degrees into 4 equal parts, and the sampled distance unit into 5 equal parts; secondly, the radar scans 3×8 circles, and the number of scan data num1, num2, num3 of the current azimuth, pitch, and distance after the radar scans 3×8 circles, as well as the echo amplitude AMP1, AMP2, AMP3 of the point trace are counted, and the calculated current echo amplitude mean newmap is used as the initial value of the detection threshold of the clutter map at the current azimuth, pitch, and distance, as well as the oldmap for the next clutter map update. The following method is used to establish the initial clutter map:
[0039]
[0040] Finally, set the window cumulative scan circle number i to 1, and set the sliding window r to 0.
[0041] Step 3: The radar performs the (m×n + i)-th circle of scanning to obtain the scanning data of this circle;
[0042] Step 3 includes:
[0043] The radar performs the (m×n + i)-th circle of scanning to obtain the scanning data of this circle.
[0044] Step 4: Based on the clutter map established from the radar's m×n circle scanning data and the scanning data after the (m×n + i)-th circle of scanning, perform clutter detection, count the number of targets in the current period, and finally output the unfiltered scanning data;
[0045] Step 4 includes:
[0046] Compare the current echo amplitude obtained in Step 3 with 4 times the newmap of the current data stored in the clutter map. If the current echo amplitude obtained in Step 3 is greater than 4 times the newmap of the current data stored in the clutter map, it is determined that a target exists, the target count targetnum is incremented by 1, and the amplitude values of the echoes of the current trace are added; if the current echo amplitude obtained in Step 3 is less than 4 times the newmap of the current data stored in the clutter map, it is determined that no target exists, and the target count targetnum remains unchanged.
[0047] Step 5: If the window cumulative scan circle number i = n, then set i = 1, increment r by 1, and update the clutter map according to the scanning data not filtered in the i-th circle and the scanning data within the sliding window range from r to m + r. If the window cumulative scan circle number i < n, increment i by 1, and repeat Steps 3 and 4;
[0048] Step 5 includes:
[0049] If the window cumulative scan circle number i has not reached the set window cumulative circle number 8, then increment i by 1, keep r unchanged, and repeat Steps 3 and 4. The radar continuously performs scanning until i reaches the window cumulative circle number 8, set i to 1, increment r by 1, and then update the clutter map. The clutter map update includes:
[0050] As Figure 2 (a) shows that according to the echo amplitudes after the radar's 3×8 circle scanning, the cumulative echo amplitude values AMP1, AMP2, AMP3 of the scanning data of Window 1, Window 2, and Window 3 and the target counts num1, num2, num3 are obtained respectively;
[0051] The sliding window slides one window to the right, as Figure 2(b) As shown in the figure, when the radar window cumulative scanning number i is equal to 8, the current echo amplitude AMP4 and the number of statistical targets num4 are obtained respectively;
[0052] The sliding window range is r~3+r scanning data echo amplitude mean A 3+r The calculation formula is,
[0053]
[0054] Among them, num 3+r =targetnum. The clutter map is updated, and the update formula is:
[0055]
[0056] Among them, oldmap is the current data stored in the clutter map in step 2; is the clutter map update coefficient, and its value range is In this paper, the clutter map update coefficient is Newmap is the data stored in the clutter map after iterative update, and also serves as the oldmap for the next clutter map update.
[0057] Figure 3 and Figure 4 They are respectively simulation diagrams before and after clutter suppression using the radar data processing method for updating the clutter map by sliding window processing of the present invention. In the diagrams, the blue dots are significantly reduced, and the scanning results are clearer.
Claims
1. A radar data processing method for updating clutter maps using a sliding window, characterized in that: The method includes the following steps: Step 1: Set the sliding window length to m and the window cumulative number of turns to n, and obtain the scan data of the radar scanning m×n turns. Step 2: Set the window cumulative scan number of turns i to 1, and set the number of sliding windows of the sliding window (hereinafter referred to as the sliding window of the sliding window) r to 0, and establish a clutter map according to the above scan data. Step 3: The radar performs the (m×n + i)-th turn of scanning and obtains the scan data of this turn. Step 4: Perform clutter detection based on the clutter map established from the radar's m×n turns of scan data and the scan data after the (m×n + i)-th turn of scanning, and count the number of targets in the current period, and finally output the unfiltered scan data. Step 5: If the window cumulative scan number of turns i = n, then set i = 1, increment r by 1, and update the clutter map according to the unfiltered scan data of i turns and the scan data with the sliding window range from r to m + r. If the window cumulative scan number of turns i < n, increment i by 1, and repeat Steps 3 and 4.
2. The radar data processing method for updating the clutter map based on sliding window processing according to claim 1 is characterized in that: Step 2 includes: (2a) Divide 90 degrees of the radar azimuth into M equal parts, divide 32 degrees of the radar elevation into N equal parts, and divide the sampled range cells into X equal parts, where M > 0, N > 0, and X > 0. (2b) The radar performs an m×n circle scan and counts the number of points of the current azimuth, elevation, and distance in the m windows, which are recorded as num1, num2, ..., num m , the echo amplitudes of the points falling into the current azimuth, elevation, and distance are recorded as AMP1, AMP2, ..., AMP m , add up these echo amplitudes (hereinafter referred to as the current echo amplitude), calculate the echo amplitude mean of the current azimuth, pitch, and distance (hereinafter referred to as the current echo amplitude mean), and use the current echo amplitude mean newmap as the initial detection threshold of the clutter map at the current azimuth, pitch, and distance, and as the oldmap for the next clutter map update. The following method is used to establish the amplitude clutter map: (2c) Set the window cumulative scan number of turns i to 1, and set the sliding window of the sliding window r to 0.
3. The radar data processing method for updating the clutter map based on sliding window processing according to claim 1, characterized in that: Step 4 includes: If the current echo amplitude is greater than A times the current data (hereinafter referred to as the current data) of the clutter map stored at the current azimuth, elevation, and range, it is determined that a target exists, and the number of targets counted targetnum is incremented by 1; if the current echo amplitude is less than A times the current data of the clutter map stored, it is determined that no target exists, and the number of targets counted targetnum remains unchanged.
4. The radar data processing method for updating clutter maps based on sliding window processing according to claim 1, characterized in that: Step 5 includes: (4a) If the window cumulative scan number of turns i < n, then repeat Steps 3 and 4 until the window cumulative scan number of turns i = n, and update the clutter map, where the value range of i is 1 ≤ i ≤ m. (4b) If the window cumulative scan number of turns i = n, update the clutter map, including the following steps: (4c) Increment the sliding window of the sliding window r by 1. (4d) According to the current echo amplitude AMP of the radar scanning circle i m+r 、Statistical target number num m+r The amplitude clutter map is iteratively updated with the scan data of the radar m×n circles. The current data oldmap stored in the clutter map is and the sliding window range is r~m+r scanning data echo amplitude mean The updated clutter map newmap is obtained by adding them together. The sliding window range is r~m+r, the echo amplitude mean value A of the scan data m+r The calculation formula is, Among them, num m+r =t arg etnum. Then the clutter map update formula is: Among them, oldmap is the data stored before the clutter map is updated; is the clutter map update coefficient, and its value range is Newmap is the data stored in the clutter map after iterative update, and also serves as the oldmap for the next clutter map update.