Searching and tracking method for track updating with high real-time performance
By separating search and tracking in the radar system and using high-precision distance and angle measurement methods, the problem of low track update rate in traditional radar systems is solved, and high real-time track updates and accuracy improvements are achieved.
Patent Information
- Application Number
- CN202411888242.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-12
- Publication Date
- 2025-04-08
AI Technical Summary
When traditional radar systems scan and track, the track update rate is low, resulting in delays in real-time track information acquisition of targets and affecting track accuracy.
By separating search and tracking, after constructing the tracking loop, only the tracked targets are trained and tracked, the track refresh rate is improved, and high-precision distance and angle measurement is achieved through the centroid positioning algorithm and sum-to-difference amplitude angle measurement scheme.
It realizes high real-time track updates, improves the target's track refresh rate and accuracy, and can obtain the target's real-time track information in a timely manner.
Smart Images
Figure CN119780900A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of radar data processing, in particular to a search and tracking method with high real-time track updating. Background Art
[0002] Traditional mechanical scanning radars and phased array scanning radars generally use Track While Scan (TWS) technology and Track And Search (TAS) technology, but these two technologies generally update the track information while completing a circle of search, which greatly reduces the update rate of the track and improves the accuracy of the track to a certain extent. The new technology aims to separate search and tracking. Once the tracking loop is built, the system will no longer search for new targets and build a new tracking loop, so that only the n tracking loop is tracked in the tracking mode. track The system can track each target in rotation, which greatly improves the track refresh rate of the tracked target and can obtain the real-time track information of the target in a more timely manner, thereby further improving the track accuracy. Summary of the invention
[0003] The purpose of the present invention is to provide a search and tracking method with high real-time track update.
[0004] The solution to achieve the purpose of the present invention is: a search and tracking method with high real-time track update, the method comprising:
[0005] Step 1: Initialization settings. First, complete system initialization and clear all caches; second, load various system parameters and set the current working state to search mode; third, wait for the interrupt indicating that initialization is complete; finally, receive N as agreed array The target detection result information of each wave position is ready for search and tracking.
[0006] Step 2: Sidelobe interference suppression. Receive the searched point information, traverse all targets detected at the current wave position, and compare the amplitude ratio of the elevation difference channel and the sum channel. If the ratio of the amplitude of the difference channel to the amplitude of the sum channel is greater than the preset threshold G ratio , the target is considered as interference and is removed.
[0007] Step 3: Valid target selection when multiple targets are detected at the same wave position. By considering the range gate number i of the point trace x , speed gate number i v and channel IQ amplitude x(i x ), using the formula THT = ai x +b|i v |+cx(i x) Calculate the threat level of the points detected in a single wave position, where a, b, and c are the weighted proportions of the range gate number, velocity gate number, and channel IQ amplitude. Finally, according to the calculated threat level, select the valid target of the current wave position and perform subsequent operations.
[0008] Step 4: It is divided into distance measurement in step 4-1 and angle measurement in step 4-2.
[0009] Step 4-1: For high-precision distance processing, first calculate the precise distance through the centroid positioning algorithm. Specifically, for each detected target point information, take the amplitude x(k) of the three points before and after the center point and the channel IQ data, and determine whether the amplitude of the left and right points is greater than the amplitude of the center point. x times; if it is greater, it is included in the calculation; otherwise, it is discarded. Finally, through the formula Calculate more accurate range gate number The distance to the target can be expressed as Where ΔR is the distance corresponding to each range gate. Secondly, through the formula Obtain the distance error caused by Doppler frequency shift, where T p is the continuous wave pulse width, f 0 is the radar operating carrier frequency, v = i v Δv is the radial velocity of the point. Finally, the high-precision detection distance after eliminating the range-Doppler coupling is obtained as
[0010] Step 4-2: The radar high-precision angle measurement adopts the sum-difference ratio angle measurement scheme. The angle measurement only uses the IQ data of the target center point in the instruction table to calculate the size and polarity of the angle error. Suppose the sum channel IQ signal is S ∑ (k) = a ∑ (n)+ib ∑ (n), the IQ signal of the azimuth difference channel is S Δa (n) = a Δa (n)+ib Δe (n), the IQ signal of the pitch difference channel is S Δe =a Δe (n)+b Δe (n). Taking the azimuth channel as an example, the modulus of the error angle can be expressed as The polarity of the error angle can be determined by the phase polarity of the ratio of the difference channel IQ to the sum channel IQ. The ratio of the difference channel IQ to the sum channel IQ is calculated to be When the phase of the ratio is in the first or second quadrant, the polarity of the error angle is positive; otherwise, the polarity is negative.
[0011] Step 5: Point trace condensation. First, determine whether the distance difference between the current wave position detection target and the azimuth and pitch adjacent wave position targets is less than the preset threshold. If the condition is met, the subsequent judgment is executed; otherwise, the judgment of the current target is completed and recorded as a new target. Secondly, determine whether the speed difference between the current wave position detection target and the azimuth and pitch adjacent wave position targets is less than the preset threshold. If the condition is met, the subsequent judgment is executed; otherwise, the judgment of the current target is completed and recorded as a new target. Next, determine whether the azimuth angle difference between the current wave position detection target and the azimuth and pitch adjacent wave position targets is less than the preset threshold. If the condition is met, the subsequent judgment is executed; otherwise, the judgment of the current target is completed and recorded as a new target. Finally, for the target group that meets the above judgment decision, it is considered that it belongs to the same target, and the point trace condensation is performed on it, and only the target with the highest threat is retained as the detection target. Here, the determination of the threat degree is determined by the IQ amplitude of the sum channel. The larger the amplitude, the greater the threat degree.
[0012] Step 6: point trace association. Assuming that the current search number is m and m>1, first, determine whether the difference in distance between the current target and the target at the adjacent wave position and the same wave position in the m-1th circle is less than the preset threshold. If the condition is met, the subsequent judgment is executed, otherwise it is considered that the current target has not been associated with the target of the previous circle, and the target is discarded; secondly, determine whether the difference in speed between the current target and the target at the adjacent wave position and the same wave position in the m-1th circle is less than the preset threshold. If the condition is met, the subsequent judgment is executed, otherwise it is considered that the current target has not been associated with the target of the previous circle, and the target is discarded; thirdly, determine whether the difference in azimuth between the current target and the target at the adjacent wave position and the same wave position in the m-1th circle is less than the preset threshold. If the condition is met, the subsequent judgment is executed, otherwise it is considered that the current target has not been associated with the target of the previous circle, and the target is discarded. target; fourth, determine whether the difference in pitch angle between the current target and the target at the adjacent wave position and the same wave position of the m-1th round is less than the preset threshold. If the condition is met, the subsequent judgment is executed, otherwise it is considered that the current target has not been associated with the target of the previous round and the target is abandoned; fifth, for the target that meets the above judgment, it is recorded as a group with the target of the m-1th round that has been successfully matched, and a mark is set. In the case of multiple target matches, the ones with the smallest matching error are preferentially selected and grouped together; finally, traverse all the targets of the current round. If no target in the current round successfully matches the previous round, the target of the previous round is abandoned, and the result of the current round point trace condensation is used as the starting number of association rounds, and point trace association is performed from the next round. If there is a target that has been successfully associated with the previous round, the association round counter n g Add 1.
[0013] Step 7: Tracking loop construction. First, compare the associated counter n g With the preset threshold N g , when n g ≥Ng When the number of valid target groups n is reached, the current target group is considered to be a valid target group and the tracking loop construction continues; otherwise, the current target group is considered to be invalid and is discarded; secondly, when the number of valid target groups n is reached, the current target group is considered to be a valid target group and the tracking loop construction continues; otherwise, the current target group is considered to be invalid and is discarded; track Greater than the maximum number of targets to follow N track When THT=ai x +bi y +cx(i x ) Calculate the threat level of each target, sort them, and select the top N track Targets build tracking loops and consider the effective number of target groups n track =N track ; Finally, for n track A tracking loop is constructed based on a set of valid targets. The construction of the tracking loop is essentially an α-β tracking loop that initializes the range, azimuth, and elevation.
[0014] Step 8: Tracking status. First, calculate the time difference and update the target position prediction value of the current round. In the input data, select a target that is close to the target according to the target prediction value as the target measurement value of this time, and estimate the target's distance, azimuth and pitch angle; secondly, substitute the measurement value into the established tracking loop, update the loop state parameters according to the filtering algorithm, and use the predicted azimuth and pitch angle of the target tracking loop as the tracking beam pointing of the target; thirdly, if a target is not detected in the current round, predict the target according to the loop information, set the target state to the memory state, and start to memorize the counter n. mem Add 1; Fourth, if the memory count value n mem Less than threshold N mem If this batch of targets is captured again, the measurement data is brought into the tracking loop to continue tracking, and the memory counter n is set. mem Clear; Fifth, if the memory counter n of a target mem Greater than threshold N mem , the target is considered lost, the tracking loop of the target is cleared, and the tracking target number counter n is set track Minus 1; Finally, if the number of tracked targets is n track If it is less than 1, the tracking state is considered to have failed. After the system re-enters the search state, step 2 is executed. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] Figure 1 The present invention is an operational flowchart of an embodiment of the present invention.
[0016] Figure 2 This is the schematic diagram of the centroid positioning algorithm.
[0017] Figure 3 The range gate-energy amplitude diagram for distance dimension CFAR. DETAILED DESCRIPTION
[0018] The method of the present invention is further described below in conjunction with the accompanying drawings and embodiments.
[0019] The present invention is a search and tracking method with high real-time track update, such as Figure 1 As shown, the method includes:
[0020] Step 1: Initialize the settings. Confirm the number of search positions N search =48 and the maximum number of tracked targets N track =3, configure the sidelobe judgment threshold G ratio =0.8, center of mass positioning scale factor G x =10%, range gate width ΔR = 4.58m, speed gate width Δv = 0.75m, radar carrier frequency f 0 =25.125GHz, pulse width T p =31.25us, bandwidth B = 50MHz, the distance, speed, azimuth, and elevation error thresholds of the point trace condensation are x ng =45m,v ng =0.75m / s,azi ng =10°,elev ng = 10°, the distance, speed, azimuth, and elevation error thresholds of the point trace association are x gg =13m,v gg =0.75m / s,azi gg =10°,elev gg =10°, associated circle threshold N g =3, memory times threshold N mem = 10. Finally, the target detection result information of 48 wave positions is received as agreed, and stored in the instruction table, ready for search and tracking.
[0021] Step 2: Sidelobe interference suppression. Accept the search trace information and traverse each target in the command table under the current wave position, compare the difference channel ratio and the channel ratio, if one of the ratios of the azimuth difference amplitude / sum channel amplitude and the elevation difference channel amplitude / sum channel amplitude is greater than the threshold G ratio , then the target is considered to be a sidelobe interference and is removed, otherwise the target is considered to be a real target and subsequent operations are performed.
[0022] Step 3: Select the effective target of the current wave position. When there are multiple real targets in the same wave position, we need to select the target with the greatest threat. The threat level can be calculated by the formula THT = ai x +bi v +cx(i x), after obtaining the threat level of each target, retain the target with the greatest threat level for subsequent operations, and propose the remaining targets.
[0023] Step 4: Distance measurement and angle measurement.
[0024] Step 4-1: Distance measurement uses the centroid positioning algorithm and the range-Doppler coupling correlation algorithm to achieve high-precision distance measurement. Figure 2 As shown, suppose there are n anchor nodes H within the communication range of an unknown node O(x,y) 1 (x 1 ,y 1 ), H 2 (x 2 ,y 2 ), H 3 (x 3 ,y 3 ),......,H N (x n ,y n ) are weighted by ω 1 ,ω 2 ,ω 3 ,......,ω n , then the principle of weighted centroid positioning algorithm can be used to obtain the coordinates of the target node O as: like Figure 3 As shown in the figure, when doing two-dimensional CFAR, the range gate number is a positive integer. The actual target is between two range gates. The energy can be used as a weighted value to perform the centroid positioning algorithm. In the specific operation, the gate number and IQ amplitude of the target point and the three points before and after are taken from the instruction table. If the amplitude of the two points affected by the target is greater than or equal to 10% of the center point, it is considered to have reference value and the point is involved in the centroid positioning algorithm. Otherwise, the point is discarded. Finally, through the formula Get a more accurate range gate number, and further find the distance Although the distance at this time is more accurate than the distance in the point trace, it is still affected by the Doppler factor. Therefore, the formula can be used to calculate By calculating the distance error caused by the Doppler frequency shift, the high-precision distance after processing can be obtained as
[0025] Step 4-2: Angle measurement is performed by sum-difference ratio. The angle measurement only uses the IQ data of the target center point in the instruction table to calculate the magnitude and polarity of the angle error. Taking the azimuth error channel as an example, the modulus value of the azimuth error angle can be expressed as The polarity of the azimuth error angle can be obtained from the ratio of the azimuth difference channel IQ data and the phase of the channel IQ data. If the phase is in the first and second quadrants, the polarity is considered positive, and if the phase is in the third and fourth quadrants, the polarity is considered negative. In specific operations, the imaginary part of the ratio a ∑ bΔa -a Δa b ∑ Polarity P a As the polarity of the error angle, the final azimuth error angle result is Δazi=P a |Δazi|, the pitch error angle is calculated in the same way.
[0026] Step 5: Point condensation. Point condensation will traverse each target of the current wave position and the adjacent wave position in turn, and perform the following judgments in turn:
[0027] (1) Is the distance difference between the current target and the adjacent targets in azimuth and elevation less than the threshold x? ng , if the conditions are met, the subsequent judgment is executed. Otherwise, the judgment of the target is completed, and the current target is recorded as a new target for subsequent operations.
[0028] (2) Is the speed difference between the current target and the target adjacent to it less than the threshold v? ng , if the conditions are met, the subsequent judgment is executed. Otherwise, the judgment of the target is completed, and the current target is recorded as a new target for subsequent operations.
[0029] (3) Is the azimuth difference between the current target and the target adjacent to it less than the threshold azimuth? ng , if the conditions are met, the subsequent judgment is executed. Otherwise, the judgment of the target is completed, and the current target is recorded as a new target for subsequent operations.
[0030] (4) Is the elevation angle difference between the current target and the target adjacent to it less than the threshold elev? ng , if the conditions are met, the subsequent judgment is executed. Otherwise, the judgment of the target is completed, and the current target is recorded as a new target for subsequent operations.
[0031] (5) For the target group that passes the previous four judgments, it is considered that they belong to a unified target and need to be condensed. The way of condensation is to retain the target with the greatest threat for subsequent operations. Here, each target meets the distance, speed, azimuth, and pitch error thresholds. When calculating the threat level, only energy needs to be considered. The greater the energy, the greater the threat level.
[0032] Step 6: Dot trace association. When the search state is in progress and the search cycle m>1, perform dot trace association on the target detected in the previous cycle. The specific steps are as follows:
[0033] (1) Determine whether the distance difference between the current target and the target at the adjacent wave position (including the current wave position) at the m-1th round is less than the preset threshold x gg If the condition is met, the subsequent judgment is executed, otherwise it is considered that the current target cannot be associated with the target of the previous round and the target is abandoned.
[0034] (2) Determine whether the difference between the speed of the current target and the speed of the target at the adjacent wave position (including the current wave position) at the time of m-1 rounds is less than the preset threshold v gg , and whether the speed directions of the two are consistent. If this condition is met, the subsequent judgment is executed, otherwise it is considered that the current target cannot be associated with the target of the previous lap and the target is abandoned.
[0035] (3) Determine whether the azimuth difference between the current target's speed and the target at the adjacent wave position (including the current wave position) at the time of m-1 turns is less than the preset threshold azi gg If the condition is met, the subsequent judgment is executed, otherwise it is considered that the current target cannot be associated with the target of the previous round and the target is abandoned.
[0036] (4) Determine whether the difference between the azimuth angle of the current target and the pitch angle of the target at the adjacent wave position (including the current wave position) at the time of m-1 rounds is less than the preset threshold elev gg If the condition is met, the subsequent judgment is executed, otherwise it is considered that the current target cannot be associated with the target of the previous round and the target is abandoned.
[0037] (5) For the target that meets the above criteria, it is recorded as a group with the target that successfully matched m-1 times, and a mark is set. If more than one target matches, the target with the smallest total matching error is preferentially selected and recorded as a group.
[0038] (6) Traverse all targets in the current round. If no target in the current round successfully matches the previous round, the target of the previous round is discarded, and the result of the point condensation of the current round is used as the starting scanning round, and the point trace association is performed from the next round. Otherwise, the associated round counter n g Add 1. And write the associated point information into the instruction table for system call.
[0039] Step 7: Tracking loop construction. When the associated turn counter n g When ≥3, traverse all associated target groups cached in the system instruction table and perform the following operations:
[0040] (1) Select the number of associated targets greater than or equal to N g , it is considered as a valid target group, otherwise it is considered as an invalid target group and discarded.
[0041] (2) When the number of valid target groups is greater than N track The method is to calculate the average threat level of each valid target group, sort the average threat levels, and select the three groups with the largest threat levels to build a tracking loop.
[0042] (3) Construct an α-β tracking loop according to the distance, speed, azimuth, and pitch angle of each group. It can be expressed as a prediction equation: Smoothing equation: Among them, x N is the distance of the Nth circle, x PN is the predicted distance of the Nth lap, is the distance smoothing value of the Nth circle, t N is the time of the Nth lap. When constructing the tracking loop, the difference between the measured values of the first and second laps is used as the initial estimate of the first-order derivative. The three batches of measured values are brought into the above-mentioned smoothing equation and prediction equation for iterative calculation, and the track start of distance, azimuth and pitch can be realized.
[0043] Step 8: Tracking state. After the data processing establishes the tracking loop, wait for the radar system to switch to the tracking state, start receiving the measurement data of the three targets, and perform the following processing:
[0044] (1) According to the instruction table time, the time difference is calculated and the target position prediction value of the current circle is updated. In the input data, the target measurement data is extracted according to the target position prediction value, and the distance, position and pitch value of the target are estimated.
[0045] (2) Bring the measured values into the α-β tracking loop of range, azimuth and elevation established during the search state and update the loop state parameters.
[0046] (3) If a target is not detected in the current circle, the track is predicted forward based on the α-β tracking loop information, and the target state is set to memory, and the memory counter is incremented by 1.
[0047] (4) If the target memory count value is less than the threshold 10 and this batch of targets is captured again, the measurement data is brought into the tracking loop to continue tracking, and the target state is set to tracking, and the memory counter of the target is cleared.
[0048] (5) If the memory count value of a target is greater than the threshold value, the target is considered lost, the target data cache is cleared, the target tracking loop is cleared, and the tracking target count is reduced by one.
[0049] (6) If the number of tracked targets is less than 1, the tracking state is considered to have failed, and the system re-enters the search mode. At this time, the state in the instruction table is set to search, and all system caches are reset. Wait for the system to switch to the search state and then repeat step 2.
Claims
1. A search and tracking method with high real-time track update, characterized in that No age search is performed in the tracking state to ensure that the tracking state has a high accuracy and refresh rate. The method includes the following steps: Step 1: Initialize settings and confirm the number of search positions N search and the maximum number of tracked targets N track , load various parameters, complete system self-check, and set the initial system working state to search state. Step 2: Sidelobe interference suppression. Receive the searched point information and perform sidelobe interference suppression to determine whether the input point is caused by sidelobe interference. If yes, repeat step 2; otherwise, execute step 3. Step 3: Select a valid target at the current wave position. When there is only one target at the current wave position, directly obtain the range gate number i of the target x , speed gate number i v And the point trace information such as the channel amplitude x(k) corresponding to the k-th range gate, and cache it in the instruction table; when there are multiple targets, select the target with the greatest threat to perform the above operation. Step 4: Distance measurement and angle measurement. The range gate number i passing through the target x , using the amplitude x(i x -1), x(i x )、x(i x +1) Perform the centroid distance measurement algorithm and perform distance-Doppler coupling on the centroid distance measurement result to obtain the target distance measurement value R it ; Through the IQ data of the sum channel and the difference channel of the target center point, the azimuth angle Δazi and the pitch angle Δelev of the target deviation from the beam direction are obtained by using the sum and difference ratio angle measurement. Step 5: Point condensation. When there is no target in the adjacent wave position, no operation is performed; when there is a target in the adjacent wave position, the adjacent wave position is traversed, and if the target information is within the preset threshold, the point traces are condensed and only the target point traces with larger energy amplitude are retained; otherwise, no operation is performed. Step 6: Dot trace association. In search mode and when the number of scan circles m>1, dot trace association for adjacent scan circles. When there is a target in the adjacent wave position of the adjacent scan circle within the preset threshold, the association counter n of the corresponding wave position is set. g Add 1; otherwise, set the associated counter n corresponding to the wave position g Cleared. Step 7: Tracking loop construction. When the associated counter n corresponding to a certain wave position g , the preset number of association circles is N g , when n g ≥N g When the wave position N g The point information of the circle is used for smooth prediction (if there are multiple targets that meet the requirements, the N with the largest threat level is selected track The system state is changed to the tracking state and step eight is executed; otherwise, the system working state is still the searching state and step two is repeated. Step 8: After the LFMCW radar system enters the tracking state, it receives the measurement point data of the patrolling tracking target and performs corresponding smooth prediction processing to obtain real-time updated track information and tracking beam pointing configuration information. track ≥1, repeat step 8, otherwise, it is considered that the tracking state has failed and the system switches back to the search state. When the system switches to the search state, execute step 2.
2. According to claim 1, it is characterized in that: The method of system initialization in step 1 specifically includes: g ≥N g Initialize settings and confirm the number of search positions N search and the maximum number of tracked targets N track , load various parameters, complete system self-check, and set the initial system working state to search state. Finally, receive N array The target detection result information of each wave position is ready for search and tracking.
3. According to claim 1, it is characterized in that: The method for suppressing sidelobe interference in step 2 specifically includes: Receive the searched point information, traverse all the targets detected at the current wave position, and compare the amplitude ratio of the pitch difference channel and the sum channel. If the ratio of the difference channel amplitude to the sum channel amplitude is greater than the preset threshold G ratio , the target is considered as interference and is removed.
4. According to claim 1, it is characterized in that: The specific methods for selecting valid targets in the same wave position in step 3 include: By considering the range gate number, velocity gate number and channel IQ amplitude of the trace, using the formula THT = ai x +bi v +cx(i x ) Calculate the threat level of the points detected in a single wave position, where a, b, and c are the weighted proportions of the range gate number, velocity gate number, and channel IQ amplitude. Finally, according to the calculated threat level, select the valid target of the current wave position and perform subsequent operations.
5. According to claim 1, it is characterized in that: The distance and angle measurement methods in step 4 specifically include: Step 4-1: For high-precision distance processing, first calculate the precise distance through the centroid positioning algorithm. Specifically, for each detected target point information, take the amplitude x(k) of the three points before and after the center point and the channel IQ data, and determine whether the amplitude of the left and right points is greater than the amplitude of the center point. x times; if it is greater, it is included in the calculation; otherwise, it is discarded. Finally, through the formula Calculate more accurate range gate number The distance to the target can be expressed as Where ΔR is the distance corresponding to each range gate. Secondly, through the formula Calculate the distance error caused by Doppler frequency shift, where T p is the continuous wave pulse width, f0 is the radar operating carrier frequency, v = i v Δv is the radial velocity of the point. Finally, the high-precision detection distance after eliminating the range-Doppler coupling is obtained as Step 4-2: The radar high-precision angle measurement adopts the sum-difference ratio angle measurement scheme. The angle measurement only uses the IQ data of the target center point in the instruction table to calculate the size and polarity of the angle error. Suppose the sum channel IQ signal is S ∑ (k) = a Σ (n)+ib ∑ (n), the IQ signal of the azimuth difference channel is S Δa (n) = a Δa (n)+ib Δe (n), the IQ signal of the pitch difference channel is S Δe =a Δe (n)+b Δe (n). Taking the azimuth channel as an example, the modulus of the error angle can be expressed as The polarity of the error angle can be determined by the phase polarity of the ratio of the difference channel IQ to the sum channel IQ. The ratio of the difference channel IQ to the sum channel IQ is calculated to be When the phase of the ratio is in the first or second quadrant, the polarity of the error angle is positive; otherwise, the polarity is negative.
6. According to claim 1, it is characterized in that: The method of point trace condensation in step 5 specifically includes: First, determine whether the distance difference between the current wave position detection target and the azimuth and pitch adjacent wave position targets is less than the preset threshold. If the condition is met, the subsequent judgment is executed; otherwise, the judgment of the current target is completed and recorded as a new target. Secondly, determine whether the speed difference between the current wave position detection target and the azimuth and pitch adjacent wave position targets is less than the preset threshold. If the condition is met, the subsequent judgment is executed; otherwise, the judgment of the current target is completed and recorded as a new target. Next, determine whether the azimuth angle difference between the current wave position detection target and the azimuth and pitch adjacent wave position targets is less than the preset threshold. If the condition is met, the subsequent judgment is executed; otherwise, the judgment of the current target is completed and recorded as a new target. Finally, for the target group that meets the above judgment decision, it is considered that it belongs to the same target, and the point traces are condensed, and only the target with the largest threat is retained as the detection target. Here, the determination of the threat degree is determined by the IQ amplitude of the sum channel. The larger the amplitude, the greater the threat degree.
7. According to claim 1, it is characterized in that: The method of point trace association in step 6 specifically includes: Assuming that the current search number is m and m>1, first, determine whether the difference in distance between the current target and the target at the adjacent wave position and the same wave position of the m-1th circle is less than the preset threshold. If the condition is met, the subsequent judgment is executed, otherwise it is considered that the current target cannot be associated with the target of the previous circle, and the target is abandoned; secondly, determine whether the difference in speed between the current target and the target at the adjacent wave position and the same wave position of the m-1th circle is less than the preset threshold. If the condition is met, the subsequent judgment is executed, otherwise it is considered that the current target cannot be associated with the target of the previous circle, and the target is abandoned; thirdly, determine whether the difference in azimuth between the current target and the target at the adjacent wave position and the same wave position of the m-1th circle is less than the preset threshold. If the condition is met, the subsequent judgment is executed, otherwise it is considered that the current target cannot be associated with the target of the previous circle, and the target is abandoned. target; fourth, determine whether the difference in pitch angle between the current target and the target at the adjacent wave position and the same wave position of the m-1th round is less than the preset threshold. If the condition is met, the subsequent judgment is executed, otherwise it is considered that the current target has not been associated with the target of the previous round and the target is abandoned; fifth, for the target that meets the above judgment, it is recorded as a group with the target of the m-1th round that has been successfully matched, and a mark is set. In the case of multiple target matches, the ones with the smallest matching error are preferentially selected and grouped together; finally, traverse all the targets of the current round. If no target in the current round successfully matches the previous round, the target of the previous round is abandoned, and the result of the current round point trace condensation is used as the starting number of association rounds, and point trace association is performed from the next round. If there is a target that has been successfully associated with the previous round, the association round counter n g Add 1.
8. According to claim 1, it is characterized in that: The method of constructing the tracking loop in step seven specifically includes: First, compare the associated counter n g With the preset threshold N g When the associated counter n g ≥N g , the current target group is considered to be a valid target group and the tracking loop construction continues; otherwise, the current target group is considered invalid and is discarded. track Greater than the maximum number of tracked targets N track When THT=ai x +b|i v |+cx(i x ) Calculate the threat level of each target, sort them, and select the top N track Targets build tracking loops and consider the effective number of target groups n track =N track Finally, for n track A tracking loop is constructed based on a group of valid targets. The construction of the tracking loop is essentially an α-β filtering tracking process that initializes the range, azimuth, and elevation.
9. According to claim 1, it is characterized in that: The method of tracking the status in step eight specifically includes: First, the time difference is calculated and the target position prediction value of the current round is updated. In the input data, a target close to the target is selected as the target measurement value of this time according to the target prediction value, and the distance, azimuth and pitch angle of the target are estimated; secondly, the measurement value is substituted into the established tracking loop, the state parameters of the loop are updated according to the filtering algorithm, and the predicted azimuth and pitch angle of the target tracking loop are used as the tracking beam pointing of the target; thirdly, if a target is not detected in the current round, the target is predicted according to the loop information, and the target state is set to the memory state, and the memory counter n is started mem Add 1; Fourth, if the memory count value n mem Less than threshold N mem If this batch of targets is captured again, the measurement data is brought into the tracking loop to continue tracking, and the memory counter n is set. mem Clear; Fifth, if the memory counter n of a target mem Greater than threshold N mem , the target is considered lost, the tracking loop of the target is cleared, and the tracking target number counter n is set track Minus 1; Finally, if the number of tracked targets is n track If it is less than 1, the tracking state is considered to have failed. After the system re-enters the search state, execute step 2.