A method for matching airborne radar multi-target detection data with true values

Abstract

The present invention proposes a method for matching airborne radar multi-target detection data and true values, including step 1, collecting data, using the airborne radar data of the local bus data as the original input of the radar detection value, and the inertial navigation data of other aircraft bus data as the original input of the true value of the aircraft position information, and forming the radar detection data set and the position true value data set of the target according to the parameter index; step 2, using the local coordinate system to build a coordinate system, taking the inertial navigation data of the airborne bus data of other aircraft, and establishing the relative position information set between the local aircraft and other aircraft, and using the information set as the true value; step 3, using the dynamic time warping algorithm to calculate the matching degree between the target radar detection position and the target true value. Step 4, optimizing the matching result according to the matching degree. This method can be used by the tester to realize the matching work between the detection target and the true target of the airborne radar in a multi-target environment, which is of key significance for evaluating the accuracy of airborne radar positioning.

Description

Technical Field

[0001] The invention belongs to the field of radio orientation and navigation, and relates to an association matching method between an airborne radar detection value and a true value applied to radar detection accuracy evaluation. Background Art

[0002] Radar is used to detect flying objects and determine their distance and direction, which is crucial to flight safety. Whether an aircraft can accurately locate multiple targets in a complex environment based on radar detection information is a prerequisite for the effective use of other onboard systems on the aircraft. The prerequisite for evaluating the positioning accuracy of airborne radar or the accuracy of radar fusion data is to correctly match the target location information detected by the radar with the target's actual location information.

[0003] Traditional radar accuracy assessment often manually calibrates the mapping relationship between detected targets and real targets. To adapt to the increasingly complex trend of air environments, reality-oriented simulation, testing, and training environments have also become more complex, with multiple platforms and multiple targets as their basic characteristics. Accurate multi-target tracking can avoid situational confusion caused by radar measurement errors and inconsistent system time between different platforms.

[0004] Current research mainly focuses on real-time target matching, target matching algorithms based on multiple feature factors and target matching algorithms based on tracks, focusing on analysis efficiency, while the evaluation of radar data accuracy requires the accuracy of the matching method. Summary of the invention

[0005] The purpose of the present invention is to propose a method for matching airborne radar detection data with target true value data, reduce the error of manual evaluation and improve matching efficiency.

[0006] The technical solution of the present invention is a method for matching airborne radar multi-target detection data with true values, step 1, collecting data, entering data into a database, using the airborne radar data of the local aircraft bus data as the original input of the radar detection value, and the inertial navigation data of other aircraft bus data as the original input of the true value of the aircraft position information, and forming the radar detection data set and the position true value data set of the target according to the parameter index;

[0007] Step 2: Establish a coordinate system with the aircraft as the origin, obtain the inertial navigation data of the airborne bus data of other aircraft, calculate the relative position of the aircraft and other aircraft, establish a relative position information set between the aircraft and other aircraft, and use the relative position information set as the true value;

[0008] Step 3, using the dynamic time warping algorithm to calculate the matching degree between the radar detection position of each radar target and the true value of each aircraft; setting the error thresholds of the radar target azimuth, pitch angle, slant range, and closing speed; selecting the aircraft true value data within the time interval of the radar target detection data, and comparing the absolute value of the difference between the true value and the detection value of the target azimuth, pitch angle, slant range, and closing speed frame by frame; calculating the deviation ratio, recursively searching the shortest path, calculating the matching degree and standard error, and inputting the aircraft number, radar target number, matching degree, and standard error into the matching degree data set;

[0009] Step 4: Sort the data set obtained in step 3 from high to low according to the matching degree to form a temporary matching degree data set. The temporary matching degree data set is initialized. The initial matching degree data set is empty. The matching degree result data set is output. The initial matching degree data set is not empty. The first data of the temporary matching degree data set is added to the matching degree result data set. The data with the same radar target number as the first data in the matching degree result data set is removed, and it is continued to be determined whether the initial matching degree data set is empty.

[0010] Preferably, the parameters of the radar detection data set of step 1 include radar target number, satellite navigation time T, target azimuth angle az, target pitch angle el, and target approach speed Vc, and the parameters of the position true value data set include aircraft number, satellite navigation time T, own-aircraft longitude LA, own-aircraft latitude LO, own-aircraft altitude AL, own-aircraft heading angle HE, own-aircraft pitch angle PI, own-aircraft roll angle RO, own-aircraft east speed VE, own-aircraft north speed VN, and own-aircraft celestial speed VU.

[0011] Preferably, the parameters of the relative position information set in step 2 include aircraft number, satellite navigation time T, slant range R, target azimuth angle az, target pitch angle el, target approach speed Vc, target entry angle A, target westward speed Vw, target northward speed Vn, and target celestial speed Vu.

[0012] Specifically, the calculation of step 2: time alignment is performed, and the true value data of the positions of other aircraft are interpolated based on the local time. The interpolation operation includes spherical linear interpolation and linear interpolation. The longitude and latitude data of the aircraft are interpolated by spherical linear interpolation; the altitude, eastward speed, northward speed, and celestial speed of the aircraft are interpolated by linear interpolation.

[0013] The matching degree calculation in step 3 is as follows: the shortest path length is divided by the number of path frames.

[0014] The standard error calculation of step 3 is as follows: taking the absolute value of the difference between the true value of the target position and the radar detection value in azimuth, pitch angle, slant range and closing speed at each step in the shortest path sequence, squaring the above absolute values ​​item by item and taking the average value, and finally taking the square root of the average value.

[0015] The beneficial effects of the present invention are as follows: by completing the matching of airborne radar detection data and target real data in a multi-target environment, making full use of airborne bus data, achieving the matching of multi-target detection values ​​and real values ​​in a large spatial range, effectively improving the efficiency of radar precision evaluation, reducing manpower while reducing human errors, and providing more objective data. At the same time, the method can be used for matching fusion data, and has a certain reference significance for intelligent target positioning algorithms. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The radar detection target and the real target matching process of the present invention;

[0017] Figure 2 The matching degree process based on dynamic time warping of the present invention;

[0018] Figure 3 The process of recursively searching for the shortest path of the present invention;

[0019] Figure 4 This is the matching degree matching process of the present invention. DETAILED DESCRIPTION

[0020] The present invention is explained and illustrated in detail below with reference to the accompanying drawings and embodiments.

[0021] The present invention provides a method for matching detection data with true values ​​in a multi-detection target environment. The specific implementation process is as follows:

[0022] Step 1: Collect data, using the airborne radar data of the local aircraft bus data as the original input of the radar detection value, and the inertial navigation data of other aircraft bus data as the original input of the true value of the aircraft position information, and forming the radar detection data set and the position true value data set of the target according to the parameter indicators;

[0023] The parameter index can be selected from the commonly used parameter indexes in the prior art according to the work requirements. The present invention provides a data set format design involving the least parameters. The radar detection data set format is shown in Table 1, and the position true value data set is shown in Table 2;

[0024] Table 1 Radar detection dataset format design

[0025]

[0026] Note: The description column is used to mark parameter indicators when certain information needs to be specially marked;

[0027] Table 2 Format design of position truth dataset

[0028]

[0029] Note: The description column is used to mark parameter indicators when certain information needs to be specially marked.

[0030] Step 101: Before collecting data, the data formats of the radar detection data set and the position true value data set are agreed upon;

[0031] Step 102: Collect data and enter it into a database.

[0032] Step 2: Establish a coordinate system with the aircraft as the origin, obtain the inertial navigation data of the airborne bus data of other aircraft, establish a relative position information set between the aircraft and other aircraft, and use the relative position information set as the true value;

[0033] Step 201, time alignment, interpolation operation is performed on the true value data of other aircraft positions based on the aircraft time: 1) The longitude and latitude data of the aircraft are interpolated by spherical linear interpolation; 2) The altitude, eastward speed, northward speed, and celestial speed of the aircraft are interpolated by linear interpolation;

[0034] Step 202, the relative positions of the aircraft and other aircraft are calculated to form a relative position information set. The data format of the relative position information set is shown in Table 3;

[0035] Table 3 Relative position dataset format design

[0036]

[0037] Note: The description column is used to mark parameter indicators when certain information needs to be specially marked.

[0038] Step 3, using the dynamic time warping algorithm to calculate the matching degree between the radar detection position of each radar target and the true value of each aircraft;

[0039] Step 301, setting radar target azimuth, pitch angle, slant range, and approach speed error thresholds;

[0040] Step 302, error calculation, selecting the true value data of the aircraft within the time interval of the radar target detection data, and comparing the absolute value of the difference between the true value and the detection value of the target azimuth, pitch angle, slant range, and closing speed frame by frame;

[0041] Step 303, deviation ratio calculation, remove the outliers from the absolute values ​​obtained in step 302, and divide the absolute value data sets of the target azimuth, pitch angle, slant range, and closing speed without outliers by the error thresholds of the target azimuth, pitch angle, slant range, and closing speed set in step 301, and then sum them up to obtain the sequence pair of the radar detection position and the true value position of the aircraft.<Tr,Tp> The distance d<Tr,Tp> , Tr represents the timestamp of the radar detection position data, and Tp represents the timestamp of the aircraft's true position data;

[0042] Step 304: Recursively search for the shortest path and calculate the sequence pair<Tr0,Tp0> To sequence pair <Tr m ,Tp n > the shortest path length D, Tr0 is the radar target detection start time, Tp0 is the earliest time of the aircraft in the radar target detection data time, Tr m is the radar target detection end time, Tp n is the last time of the aircraft within the radar target detection data time, m is the number of radar target detection data frames, and n is the number of aircraft true value data frames within the time interval of the radar target detection data;

[0043] Step 3041, initialize an empty sequence queue W={} and the shortest path length D=0, the target radar detection data number i=0, and the aircraft data number j=0 in the time interval of the radar target detection data;

[0044] Step 3042: Get sequence pair <Tr i ,Tp j > Join queue W, add D+d <Tr i ,Tp j >Assign the shortest path length D, d <Tr i ,Tp j > represents the distance between the radar detection position number i and the aircraft true position number j;

[0045] Step 3043: determine whether i and j simultaneously satisfy i=m and j=n. If so, output the queue W and the shortest path length D, and end step 304. If not, proceed to step 3044.

[0046] Step 3044, determine whether i is equal to m, if they are equal, i=i, j=j+1, go to step 3042, if they are not equal, go to step 3045;

[0047] Step 3045, determine whether j is equal to n, if they are equal, i=i+1, j=j, go to step 3042, if they are not equal, go to step 3046;

[0048] Step 3046: Get sequence pair <Tr i ,Tp j >neighboring node d <Tr i ,Tp j+1 >, d <Tr i+1 ,Tp j >, d <Tr i+1 ,Tp j+1 >minimum value among

[0049] Step 3047, take the data number with the minimum value in step 3046 to update i, j, and go to step 3042;

[0050] Step 305, calculate the matching degree, dividing the shortest path length by the number of path frames;

[0051] Step 306, calculate the standard error, take the absolute value of the difference between the true value of the target position and the radar detection value of the azimuth, the pitch angle, the slant range, and the closing speed at each step in the shortest path sequence, square the above absolute values ​​one by one, and then take the average value, and finally take the square root of the average value;

[0052] Step 307, input the matching degree data set, the data format is shown in Table 4;

[0053] Table 4. Design of matching degree dataset format

[0054]

[0055] Step 4: sorting the matching degree. Use the bubble sort method to sort the matching degree data set obtained in step 3 from high to low according to the matching degree.

[0056] Step 5: Optimize the match;

[0057] Step 501, initialize a temporary matching dataset and a matching result dataset. The structures of the two datasets are consistent with the format of the fatigue dataset in Table 4. The matching result dataset is empty. Copy the matching dataset to the temporary matching dataset.

[0058] Step 502: Check whether the temporary matching dataset is empty. If the temporary matching dataset is empty, the matching result dataset and the matching dataset are output, and step 5 ends; if it is not empty, go to step 503;

[0059] Step 503: Add the first data of the temporary matching degree data set to the matching degree result data set;

[0060] Step 504 , remove the data with the same radar target number as the first data in the temporary matching degree data set, and return to step 502 .

Claims

1. A method for matching airborne radar multi-target detection data with true values, characterized in that: Step 1: Collect data and enter it into the database. Use the airborne radar data of the local bus data as the original input of the radar detection value, and the inertial navigation data of other aircraft bus data as the original input of the true value of the aircraft position information, and form the radar detection data set and position true value data set of the target according to the parameter indicators; Step 2: Establish a coordinate system with the aircraft as the origin, obtain the inertial navigation data of the airborne bus data of other aircraft, perform time alignment, and interpolate the true value data of the positions of other aircraft based on the aircraft time: 1) The longitude and latitude data of the aircraft are interpolated by spherical linear interpolation; 2) The altitude, eastward speed, northward speed, and celestial speed of the aircraft are interpolated by linear interpolation to calculate the relative position of the aircraft and other aircraft, and establish the relative position information set between the aircraft and other aircraft. The relative position information set is used as the true value; Step 3, using the dynamic time warping algorithm to calculate the matching degree between the radar detection position of each radar target and the true value of each aircraft; Set the error thresholds of radar target azimuth, pitch angle, slant range and closing speed; select the true value data of the aircraft within the time interval of the radar target detection data, and compare the absolute value of the difference between the true value and the detection value of the target azimuth, pitch angle, slant range and closing speed frame by frame; Calculate the deviation ratio, remove the outliers from the absolute values, and divide the absolute value data sets of the target azimuth, pitch angle, slant range, and closing speed without outliers by the error thresholds of the target azimuth, pitch angle, slant range, and closing speed set previously, and then sum them up as the sequence pair of the radar detection position and the true value position of the aircraft.<Tr,Tp> The distance d<Tr,Tp> , Tr represents the timestamp of the radar detection position data, and Tp represents the timestamp of the aircraft's true position data; Recursively search for the shortest path and calculate the sequence pair<Tr0,Tp0> To sequence pair <Tr m ,Tp n > the shortest path length D, Tr0 is the radar target detection start time, Tp0 is the earliest time of the aircraft in the radar target detection data time, Tr m is the radar target detection end time, Tp n is the last time of the aircraft in the radar target detection data time, m is the number of radar target detection data frames, n is the number of aircraft true value data frames in the time interval of the radar target detection data; initialize an empty sequence queue W={} and the shortest path length D=0, the target radar detection data number i=0, the aircraft data number j=0 in the time interval of the radar target detection data; take the sequence pair <Tr i ,Tp j > Join queue W, add D+d <Tr i ,Tp j >Assign the shortest path length D, d <Tr i ,Tp j > represents the distance between the radar detection position number i and the aircraft true position number j; judge whether i and j simultaneously satisfy i equals m and j equals n. If the judgment condition is met, output the queue W and the shortest path length D; If the judgment condition is not met, determine whether i is equal to m. If they are equal, i=i, j=j+1, and continue to take sequence pairs. <Tr i ,Tp j > Join queue W, add D+d <Tr i ,Tp j >Assign the value to the shortest path length D. If they are not equal, determine whether j is equal to n. If they are equal, i=i+1, j=j, and continue to take sequence pairs <Tr i ,Tp j > Join queue W, add D+d <Tr i ,Tp j >Assign the value to the shortest path length D. If they are not equal, take the sequence pair <Tr i ,Tp j >neighboring node d <Tr i ,Tp j+1 >, d <Tr i+1 ,Tp j >, d <Tr i+1 ,Tp j+1 >, update i, j according to the data number of the minimum value, and continue to take sequence pairs <Tr i ,Tp j > Join queue W, add D+d <Tr i ,Tp j >Assign a value to the shortest path length D; To calculate the matching degree, divide the shortest path length by the number of path frames; Calculate the standard error, take the absolute value of the difference between the true value of the target position and the radar detection value in each step of the shortest path sequence, the difference in azimuth, the difference in pitch angle, the difference in slant range, and the difference in closing speed, square the above absolute values ​​one by one, and then take the average value, and finally take the square root of the average value; Input the aircraft number, radar target number, matching degree, and standard error into the matching degree data set; Step 4: Sort the data set obtained in step 3 from high to low according to the matching degree to form a temporary matching degree data set. The temporary matching degree data set is initialized. The initial matching degree data set is empty. The matching degree result data set is output. The initial matching degree data set is not empty. The first data of the temporary matching degree data set is added to the matching degree result data set. The data with the same radar target number as the first data in the matching degree result data set is removed, and it is continued to be determined whether the initial matching degree data set is empty.

2. The method for matching airborne radar multi-target detection data with true values ​​according to claim 1, characterized in that: The parameters of the radar detection data set in step 1 are radar target number, satellite navigation time T, target azimuth az, target pitch angle el, and target closing speed Vc. The parameters of the position truth data set are aircraft number, satellite navigation time T, own-aircraft longitude LA, own-aircraft latitude LO, own-aircraft altitude AL, own-aircraft heading angle HE, own-aircraft pitch angle PI, own-aircraft roll angle RO, own-aircraft eastward speed VE, own-aircraft northward speed VN, and own-aircraft celestial speed VU.

3. The method for matching airborne radar multi-target detection data with true values ​​according to claim 1, characterized in that: In step 2, the parameters of the relative position information set include aircraft number, satellite guidance time T, slant range R, target azimuth angle az, target pitch angle el, target closing speed Vc, target entry angle A, target westward speed Vw, target northward speed Vn, and target celestial speed Vu.

Images