CN113325413A

CN113325413A - Radar target locking tracking method and device

Info

Publication number: CN113325413A
Application number: CN202110450051.4A
Authority: CN
Inventors: 吴浩; 李刚
Original assignee: Tsinghua University
Current assignee: Tsinghua University
Priority date: 2021-04-25
Filing date: 2021-04-25
Publication date: 2021-08-31
Anticipated expiration: 2041-04-25
Also published as: CN113325413B

Abstract

The invention discloses a radar target locking tracking method and a device, wherein a tracking processing point is selected, radar scanning parameters are updated for scanning, whether a target exists in scanning data is calculated through a radar algorithm, and if the target exists, the target is immediately pushed into a screening strategy to carry out multistage strategy screening to determine a final tracking target; otherwise, the detected target is continuously scanned in the execution period, when no push point or no real target meeting the screening exists in the scanning period, the recovery strategy is executed to update the radar scanning parameter, and when the frequency of continuously executing the recovery strategy exceeds a threshold value, the tracking process is exited. And modifying and outputting the real target pushed after the multi-stage strategy is screened, if no target pushed by the radar or no real target meeting the screening exists in the scanning period, outputting the target which predicts the current moment by using an algorithm, performing track matching on the output target, and adding the corresponding track. Therefore, the reliability, the accuracy, the efficiency and the stability of the radar target locking tracking are improved.

Description

Radar target locking tracking method and device

Technical Field

The invention relates to the technical field of radar tracking, in particular to a radar target locking tracking method and device.

Background

In reality, various high-risk and sensitive targets seriously threaten the life, safety, privacy and other problems of people, particularly, for occasions with important people, places such as oil, dams and the like need to pay close attention to the targets in real time so as to rapidly guide striking processing equipment to strike and process the targets to prevent various dangerous conditions and accidents from being generated, so that the requirements on the reliability, accuracy, efficiency and stability of target tracking and locking are extremely high, however, the current radar target tracking and locking system or algorithm cannot effectively solve and improve the performances, and the current facing problems are as follows: target signal weakness, false target, inaccurate parameter estimation, low processing effect and large data volume, so a good locking tracking system and algorithm are urgently needed.

The existing target tracking method generally comprises the processes of data preprocessing, point track data arrangement, point track association and track filtering updating. In general, a classical target tracking method generates track data from point track data under the driving of azimuth data, like a pipeline.

With the rise of algorithm complexity and the diversity of input data, the explosion of data combinations becomes the most troublesome problem faced by the radar target tracking method. In the theories of active data processing of related radars, generally more complex associated filtering algorithms, such as joint probability distribution algorithm, multi-hypothesis algorithm and the like, are described as time consuming, large in calculation amount and high in engineering application difficulty, a series of application hypotheses are proposed aiming at the existing difficulties, and the integrity and complexity of the algorithm are reduced, so that the optimal algorithm is reduced to be suboptimal or locally optimal. But the undeniable fact is that the performance of the algorithm suffers an inevitable decline.

1. Problems encountered in a real-world target tracking lock-in environment:

2. fading of target signal: target signal strength variations due to multipath propagation, blind speed and atmospheric environment.

3. False target: false targets are scanned due to signal noise, signal clutter, and signal interference.

4. Inaccurate estimation of radar scanning parameters: due to the influences of noise, antenna instability, target indistinguishability, target splitting, multipath propagation and propagation effects, the set parameters cannot scan a real and effective target.

Disclosure of Invention

The present invention is directed to solving, at least to some extent, one of the technical problems in the related art.

Therefore, the invention aims to provide a radar target locking and tracking method, which solves the problems of low processing effect, large data volume, low accuracy and non-ideal real-time tracking during radar target tracking.

Another object of the present invention is to provide a radar target locking and tracking device.

In order to achieve the above object, an embodiment of the present invention provides a radar target locking and tracking method, including:

s1, selecting a current target point, and setting current scanning parameters according to the data information of the current target point and the scanning equipment information;

s2, scanning according to the current scanning parameters in the current period, calculating scanning data through a radar algorithm, and determining whether a tracking target exists in the scanning data;

s3, when a tracking target exists in the scanning data, judging whether the tracking target is a real target or not through a multi-level target screening strategy;

s4, when the tracking target is a real target, correcting the tracking target to be used as an output target, performing track association matching according to the output target, updating the current scanning parameters according to the data information of the output target, executing S2, and scanning in a new scanning period;

and S5, when the tracking target is not the real target, predicting the output target of the current scanning period through a prediction algorithm, performing track correlation matching on the predicted output target, updating the current scanning parameters, executing S2, and performing scanning of a new scanning period.

In order to achieve the above object, an embodiment of another aspect of the present invention provides a radar target tracking apparatus, including:

the setting module is used for selecting a current target point and setting current scanning parameters according to the data information of the current target point and the scanning equipment information;

the calculation module is used for scanning according to the current scanning parameters in the current period, calculating scanning data through a radar algorithm and determining whether a tracking target exists in the scanning data;

the judging module is used for judging whether the tracking target is a real target or not through a multi-stage target screening strategy when the tracking target exists in the scanning data;

the first output module is used for correcting the tracking target to be used as an output target when the tracking target is a real target, performing track association matching according to the output target, updating current scanning parameters according to data information of the output target, and scanning a new scanning period;

and the second output module is used for predicting the output target of the current scanning period through a prediction algorithm when the tracking target is not the real target, performing track association matching on the predicted output target, updating the current scanning parameters and performing scanning of a new scanning period.

The radar target locking tracking method and device provided by the embodiment of the invention have the following advantages:

1) the target tracking and locking efficiency is improved; and the minimum airspace scanning is used, the traditional radar scanning period is directly skipped, the time for locking a real target is greatly shortened, and the tracking and locking efficiency is improved. And high-speed calculation is carried out by using a GPU with high calculation amount, so that the processing efficiency is improved.

2) The reliability of tracking and locking the target is improved; on the basis of traditional algorithm processing models, a target behavior weighting processing mode is provided, so that real effective targets can be distinguished from false targets more accurately, and the distinguishing capability of a system on the real effective targets is improved.

3) The stability of tracking and locking the target is improved; compared with the traditional radar data processing mode, the tracking method is designed, the air maneuvering target can be stably locked, and the air maneuvering target is kept in the locking of the system all the time.

4) The method is not only suitable for mechanical radar scanning locking tracking, but also suitable for locking tracking of phased array radar, namely suitable for target scanning locking tracking.

Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.

Drawings

The foregoing and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:

FIG. 1 is a flow diagram of a radar target lock tracking method according to one embodiment of the present invention;

FIG. 2 is a block flow diagram of a radar target lock tracking method according to one embodiment of the present invention;

FIG. 3 is a block flow diagram of a radar target lock tracking method according to another embodiment of the present invention;

FIG. 4 is a detailed tracking lock flow diagram of a radar target locking tracking method according to one embodiment of the present invention;

fig. 5 is a schematic structural diagram of a radar target tracking device according to an embodiment of the present invention.

Detailed Description

Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are illustrative and intended to be illustrative of the invention and are not to be construed as limiting the invention.

The following describes a radar target locking tracking method and apparatus proposed according to an embodiment of the present invention with reference to the accompanying drawings.

First, a radar target lock tracking method proposed according to an embodiment of the present invention will be described with reference to the accompanying drawings.

FIG. 1 is a flow chart of a radar target lock tracking method according to one embodiment of the invention.

As shown in fig. 1, the radar target locking tracking method includes the following steps:

step S1, selecting the current target point, and setting the current scanning parameters according to the data information of the current target point and the scanning device information.

Specifically, a found target is selected as a reference point of system tracking locking, namely a tracking processing point, scanning parameters are dynamically set according to the tracking processing point, and the radar scans according to the set scanning parameters to detect information of the found target. According to the radar tracking target information, the operation parameters are dynamically set, so that the operation resources can be reduced, and the operation efficiency is improved.

Further, the object S1 further includes:

and according to the data information of the current target point, carrying out Kalman algorithm to predict a spatial region which is possibly generated in the period under the target, and dynamically updating the current scanning parameters through the predicted data.

Further, the scanning device information includes azimuth angle and pitch angle data of the radar turntable;

the data information of the current target point includes speed, horizontal azimuth angle, pitch angle, distance, signal amplitude, motion direction and the like.

And step S2, scanning according to the current scanning parameters in the current period, calculating the scanning data through a radar algorithm, and determining whether the tracking target exists in the scanning data.

Further, after S2, the method further includes:

and S21, when the tracking target does not exist in the scanning data of the current scanning period, automatically updating the current scanning parameters by using a recovery strategy, and scanning the new scanning period.

Further, after S21, the method further includes:

if the tracking target exists in the scanning data of the new scanning period after the current scanning parameters are updated, executing S3;

if the tracking target does not exist in the scanning data of the new scanning period after the current scanning parameters are updated, the current scanning parameters are updated again by using the recovery strategy to scan the new scanning period until the continuous updating times reach the preset threshold value, and the tracking of the current target is finished.

Further, in a scanning period without a tracking target, an output target of the current scanning period is predicted through a prediction algorithm, and track correlation matching is carried out on the predicted output target.

And further, radar algorithm calculation is carried out on the detected scanning data, and whether a tracking target exists is calculated. And if the tracking target is calculated by the radar, immediately pushing the tracking target to a screening strategy for screening and confirming the tracking target. If the target is not tracked and the scanning period is not finished, continuing to scan the task to calculate the target point, if the periodic scanning is finished and the target is not tracked, performing corresponding Kalman algorithm prediction output, and updating the parameters by using a recovery strategy.

It can be understood that, when the parameters are updated by using the recovery strategy, in order to reduce the loss rate of the target, the corresponding recovery strategy is set to expand the scanning area, and the target scanning detection is performed again. And if the target is not scanned by the retrieval strategies for several times, the target is considered to be lost, the whole tracking strategy flow is exited, and corresponding predicted values are carried out in the screening strategies each time the retrieval strategies are executed. And if the retrieval times threshold value is not exceeded, carrying out next scanning detection.

And step S3, when the tracking target exists in the scanning data, judging whether the tracking target is a real target or not through a multi-stage target screening strategy.

Specifically, a tracking target pushed by the radar is screened and confirmed by a multi-stage strategy, and a final target is output.

Further, S3 further includes:

s31, setting a target position threshold, screening the target through the target position threshold, executing S32 if the tracked target meets the condition of the target position threshold, and waiting for other tracked targets in the current scanning period if the tracked target does not meet the condition of the target position threshold;

s32, weighting the multi-dimensional feature information in the tracked target to obtain a weight, carrying out weight screening on the weight of the tracked target through a weight threshold, and if the tracked target meets the condition of the weight threshold, determining that the tracked target is a real target, and executing S4; if the tracked target does not meet the condition of the weight threshold, adding the tracked target into the data queue, continuing to scan in the current scanning period, if a real target exists in the scanning period, executing S4, if no real target exists after the current scanning period is finished, selecting the tracked target with the maximum weight as the real target in the current scanning period from the data queue in the current scanning period, clearing the data queue, and executing S4.

Further, S31 further includes:

when all the tracked targets in the current scanning period do not satisfy the condition of the target position threshold, the tracked targets are not real targets, and S5 is executed.

And step S4, when the tracking target is a real target, correcting the tracking target to be used as an output target, performing track association matching according to the output target, updating the current scanning parameters according to the data information of the output target, executing S2, and scanning in a new scanning period.

Further, S4 is followed by:

and if the output target is the correction value of the tracking target, matching and correcting the historical prediction target in the flight path according to the output target.

And step S5, when the tracking target is not the real target, predicting the output target of the current scanning period through a prediction algorithm, performing track correlation matching on the predicted output target, updating the current scanning parameters, and executing S2 to perform scanning of a new scanning period.

Further, in one embodiment of the present invention, when the number of times of continuously predicting the output target exceeds a set threshold, the tracking of the current target is ended.

Specifically, if a real target is obtained through screening, a kalman algorithm is used for correction output, and if no real target exists in the whole period, corresponding kalman algorithm prediction output is performed. Specifically, when no real target exists in the period, target information after Kalman prediction filtering is pushed outwards to represent the real target information in the current period.

And after the output target is obtained, carrying out track matching association according to the output target, and adding the track matching association into a track queue to form a real-time track. If the correction value of the tracking target is output in the period, the historical predicted point of the matched track is corrected, so that the target at the predicted moment is closer to the reality, otherwise, the historical predicted point is not corrected.

The method of the present invention will be described in detail below with reference to the accompanying drawings.

The idea of the invention is as follows: selecting a found target as a reference point of system tracking locking, namely a tracking processing point, then dynamically setting calculation parameters, scanning by a radar according to the parameters, processing data received by scanning by a radar algorithm to calculate whether the data is a target, continuously executing the scanning calculation process in the process, and immediately pushing the data to a screening strategy to screen a multi-stage strategy to determine a final tracking target if the target exists; otherwise, continuously executing the scanning detection target in the period, executing the retrieving strategy when no pushing point or target meeting the screening strategy exists in the scanning period, exiting the whole tracking process when the frequency of continuously executing the retrieving strategy exceeds a threshold value, otherwise, updating the parameters by using the retrieving strategy and continuously executing the scanning detection target. And carrying out Kalman algorithm prediction correction output on the target pushed after the multi-stage strategy is screened, and if the target pushed by the radar is not obtained or the target meeting the screening strategy is not obtained in the detection period, outputting target information for predicting the current moment by using the Kalman algorithm. And finally, carrying out track matching on the output target, adding the corresponding track, and when outputting a value for tracking target correction, carrying out matching correction on points in the track to correct historical prediction data.

Specifically, as shown in fig. 2, fig. 3 and fig. 4, the tracking method of the present invention is divided into two parts, a tracking policy and a screening policy, and the two policies are operated in parallel in the system, so that the efficiency can be improved.

1) And selecting a found target to enter a tracking locking process, and inputting current target information.

2) And according to the input target information, carrying out Kalman algorithm to predict a spatial region which possibly appears in a target lower period, dynamically updating parameters by using prediction data, and executing new parameter scanning.

3) And the radar executes a tracking scanning strategy and a radar target screening strategy according to the predicted space area.

Tracking and scanning strategies:

a) and (3) the radar scans a designated area in a designated period, data is processed by utilizing a radar algorithm, and if a target exists, target information is immediately pushed outwards to a target screening strategy for screening and confirming the actual target.

b) If the T1 periodic radar algorithm does not obtain a suspicious target or does not have a target meeting the screening strategy, the strategy is retrieved once, and parameters such as a scanning area, a period and the like are recalculated. And determining a target loss period through the set expansion search times, and exiting the tracking mode. If the system detects that the suspicious target appears, the system enters a target screening strategy, and confirms that the suspicious target is a real target, the system ends the old tracking strategy, updates the locked tracking scanning parameters, and starts the locked tracking scanning.

The scanning period is dynamically changed and is updated by the scanning result in the period T1, if the target is not found in the first period T1, the retrieval strategy needs to be executed, the scanning area needs to be enlarged, and the periodic T2 scanning corresponding to the updating of the retrieval strategy is searched again.

And (3) target screening strategy:

a) target screening strategy Using a Secondary target Filtering strategy

First-stage filtration: and (4) screening target positions, namely, setting prediction parameters by combining a reference point target distance and a system, calculating a position screening threshold Thr1 by combining time difference and speed, screening the target positions, carrying out secondary filtering on the position condition points if the conditions are met, waiting for next detection data in the current period if the conditions are not met, and predicting by using b) if the period has no detection data.

And (3) second-stage filtering: weighting the target multidimensional characteristic information (speed, distance, signal-to-noise ratio and the like) to obtain a weight value, setting a weight value threshold (Thr2) according to experience and system efficiency to carry out weight screening, carrying out b) correction operation on the target information meeting the condition, otherwise, waiting for the completion of the detection and screening of all targets in the current period, determining whether a target meeting the threshold, namely a real target, is obtained, carrying out b) correction operation, and carrying out prediction by using b) if the current period is completed and the target meeting the threshold is not obtained.

b) Kalman algorithm filtering prediction correction target

There are three cases of corresponding input data:

and in case 1, the target information meeting the filtering requirements of Thr1 and Thr2 obtains real value information of the target elimination error by correcting data, and starts a new round of tracking strategy again by using the real information to realize the locking tracking of the target.

And in case 2, the condition that Thr1 does not satisfy the target information of Thr2 filtering, namely the condition that the target of the secondary filtering is not satisfied by the threshold in the secondary filtering period, but the target is detected in the period, and the target with the maximum secondary statistical weight in the period is selected as the output target.

And in case 3, if the targets of Thr1 and Thr2 are not met in the period, prediction is carried out to obtain one piece of prediction information of the current period, and the prediction information is output.

The first two situations are output real targets, corresponding correction is carried out through a kalman algorithm to serve as a final output target, track correlation matching is carried out, the final output target is added into a track queue to form a real-time track, the output target is used for correcting a historical track prediction target, a new round of tracking strategy is started again by using the output target, and locking tracking of the target is achieved; the predicted target does not need to initiate a new round of tracking, which is performed by the corresponding tracking recovery strategy.

4) End tracking

It can be understood that the invention takes the tracking and locking target as the center, analyzes and sets the tracking and scanning detection period and finds the target strategy parameter according to the time efficiency and the accuracy in many aspects, and can improve the time efficiency, reduce the loss rate of the target and increase the accuracy of the tracking target.

Target detection is carried out by setting and retrieving target strategy parameters and dynamically updating scanning parameters. Take the 2-time recovery strategy as an example: firstly, the radar scans a designated area (a three-dimensional area formed by horizontal and pitching) in a designated period to obtain target information: a suspected target detected in a T1 period is immediately pushed into a screening strategy; if no target meeting the screening strategy is found or is not found, a scanning area is properly enlarged according to the retrieval strategy, new period scanning detection is carried out, and the suspected target is detected to be immediately pushed into the screening strategy; and if the target is not detected in the period T2, the target is considered to be lost, the tracking mode is exited, and the lock tracking is finished.

By using the target double-threshold screening strategy and the secondary target screening strategy, the speed and the accuracy can be simultaneously met, and the real target information can be obtained.

A first stage: and judging the screening of the distance based on the target speed, the time difference and the like, and performing secondary screening if the distance is met.

And a second stage: and carrying out weighted screening based on the multidimensional characteristic weight coefficient. Setting a screening threshold value for improving time efficiency and accuracy, and when the weight value of the multi-dimensional characteristic is greater than the set threshold value, considering the target as the target and immediately jumping out; otherwise, adding the target into the queue to be selected, and waiting for the next target in the period.

Selecting the target with the maximum weight as the detection target of the current period after the period is finished; if no target in the period meets one stage and the predicted period number is smaller than the target loss period number threshold, predicting a value according to a Kalman algorithm to be used as target output, otherwise, ending tracking.

And forming and correcting a flight path after an output result is obtained. And outputting target information obtained after prediction and correction of the target subjected to double-threshold screening according to a kalman algorithm, and if no target exists, taking a predicted value as output to be matched and associated with the previous output target to form a flight path. And when the output point is the real target, correcting the predicted target in the track, and updating the historical predicted target of the track to enable the track to be closer to the real track.

According to the radar target locking tracking method provided by the embodiment of the invention, the radar target tracking method is split, and high-efficiency GPU parallel processing is used by utilizing the high flexibility of a computer in the aspect of data operation; real-time dynamic parameter updating is used for radar scanning parameters to improve data accuracy and reduce false targets; using a multi-feature weight to superpose and screen the target in the aspect of associating and screening the target; the reliability, the accuracy, the high efficiency and the stability of the radar tracking locking system are improved.

Next, a radar target locking and tracking apparatus proposed according to an embodiment of the present invention is described with reference to the drawings.

As shown in fig. 5, the radar target locking and tracking apparatus includes: the device comprises a setting module 501, a calculating module 502, a judging module 503, a first output module 504 and a second output module 505.

The setting module 501 is configured to select a current target point, and set a current scanning parameter according to data information of the current target point and scanning device information.

A calculating module 502, configured to scan according to the current scanning parameter in the current period, calculate the scanning data through a radar algorithm, and determine whether a tracking target exists in the scanning data.

The determining module 503 is configured to determine whether the tracking target is a real target according to a multi-level target screening strategy when the tracking target exists in the scan data.

The first output module 504 is configured to, when the tracking target is a real target, modify the tracking target to serve as an output target, perform track association matching according to the output target, update a current scanning parameter according to data information of the output target, and perform scanning in a new scanning period.

And a second output module 505, configured to predict, by using a prediction algorithm, an output target in the current scanning period when the tracking target is not the real target, perform track association matching on the predicted output target, update the current scanning parameter, and perform scanning in a new scanning period.

It should be noted that the foregoing explanation of the method embodiment is also applicable to the apparatus of this embodiment, and is not repeated herein.

According to the radar target locking and tracking device provided by the embodiment of the invention, the radar target tracking method is split, and high-efficiency GPU parallel processing is used by utilizing the high flexibility of a computer in the aspect of data operation; real-time dynamic parameter updating is used for radar scanning parameters to improve data accuracy and reduce false targets; using a multi-feature weight to superpose and screen the target in the aspect of associating and screening the target; the reliability, the accuracy, the high efficiency and the stability of the radar tracking locking system are improved.

Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "a plurality" means at least two, e.g., two, three, etc., unless specifically limited otherwise.

In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.

Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.

Claims

1. A radar target locking tracking method is characterized by comprising the following steps:

2. The method according to claim 1, further comprising, after the S2:

3. The method of claim 2, further comprising, after S21:

4. The method according to claim 2 or 3,

and in the scanning period without the tracking target, predicting the output target of the current scanning period by a prediction algorithm, and performing track correlation matching on the predicted output target.

5. The method according to claim 1, wherein the S3 further comprises:

s31, setting a target position threshold, screening targets through the target position threshold, executing S32 if the tracked targets meet the condition of the target position threshold, and waiting for other tracked targets in the current scanning period if the tracked targets do not meet the condition of the target position threshold;

6. The method of claim 5, wherein the S31 further comprises:

7. The method according to claim 1, further comprising, after the S4:

8. The method according to claim 1, wherein the S1 further comprises:

and predicting a spatial region which is possibly generated in the period under the target by using a kalman algorithm according to the data information of the current target point, and dynamically updating the current scanning parameter through the predicted data.

9. The method of claim 1, wherein the scanning device information includes azimuth and elevation data of a radar gantry;

the data information of the current target point comprises speed, horizontal azimuth angle, pitching angle, distance, signal amplitude and motion direction.

10. A radar target-locking tracking apparatus, comprising:

the first output module is used for correcting the tracking target to be used as an output target when the tracking target is a real target, performing track association matching according to the output target, updating current scanning parameters according to data information of the output target and scanning a new scanning period;