A target point tracking algorithm based on Kalman filter
The Kalman-based target point tracking algorithm addresses dynamic aircraft mission challenges by predicting and correcting target positions, enhancing adaptability and efficiency in complex environments.
Patent Information
- Application Number
- CN202210001286.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-01-04
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2042-01-04
AI Technical Summary
Traditional point trajectory algorithms struggle to meet timeliness and reliability requirements in dynamic aircraft mission environments, particularly in predicting and adapting to dynamic threats and moving targets.
A target point tracking algorithm based on Kalman filtering, incorporating a point trace processing module with submodules for association, filtering, starting, and termination, utilizing the Kalman filter to predict and correct target positions and speeds, and adaptively track targets in complex environments.
Enhances the aircraft's ability to predict and avoid dynamic threats, ensuring timely flight trajectory adjustments and improving mission efficiency with higher accuracy and reduced computational load.
Smart Images

Figure CN114299122B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a point trace algorithm, in particular to a target point trace algorithm based on Kalman filtering. Background Art
[0002] The aircraft estimates the motion parameters of the dynamic factors in the mission environment. Through pre-estimation, the aircraft can perceive the changes in the dynamic environment in advance, adjust the flight trajectory in time, avoid dynamic threats in the mission environment, shorten the time to reach the target area, and improve the efficiency of the aircraft in completing the given mission. The point trajectory estimation method based on filtering technology has a small amount of calculation. At the same time, the estimated value obtained by filtering technology can make the subsequent point trajectory prediction advanced and has strong practicality. It is widely cited in engineering. The Kalman filter algorithm and its corresponding derivative algorithm are classic algorithms.
[0003] Considering the dynamics of the real mission environment of the aircraft, the aircraft trajectory planning system needs to adjust the aircraft's flight trajectory in time according to the real-time changing environment. However, the traditional point trajectory algorithm prediction is difficult to meet the timeliness and reliability requirements of the aircraft in a dynamic environment. The Anytime Restoration Sparse A* (AR-SAS) algorithm, which combines the Anytime algorithm architecture with the sparse A* algorithm, can effectively solve the above-mentioned point trajectory planning problem with time constraints.
[0004] Therefore, it is necessary and significant to study the dynamic trajectory prediction method of aircraft in view of dynamic factors such as dynamic threats and moving targets in the aircraft mission environment. In view of the uncertainty of the position of dynamic factors in the battlefield, the aircraft needs to estimate the moving targets or threats in the dynamic environment in advance. Then, the aircraft can adapt to the complex and changing mission requirements through dynamic trajectory prediction, and adjust the aircraft flight trajectory in time according to the changes in the environment to improve the overall mission execution efficiency of the aircraft system. Summary of the invention
[0005] The technical problem to be solved by the present invention is to predict the position of an aircraft and enhance the aircraft's ability to perform tasks. The purpose is to provide a target point trajectory algorithm based on Kalman filtering to solve the above-mentioned problem.
[0006] The present invention is achieved through the following technical solutions:
[0007] A target point trace algorithm based on Kalman filtering, the algorithm is modeled through a point trace processing module, determines the target heading, speed, and position information, and removes clutter data and interference data, extracts the real target point trace data, and then forms a local point trace through navigation, batching, and batch merging processing;
[0008] The point trace processing module includes a point trace association submodule, a point trace filtering submodule, a point trace starting submodule, and a point trace termination submodule;
[0009] The point trace association submodule associates and matches the updated point trace with the historically formed trajectory, and selects a point with a high degree of association as the current point of the target point trace by comparing the positional relationship between the extrapolated point and the current point selected in the current scanning cycle;
[0010] The point trace filter submodule completes the prediction of the target position and speed, and combines the predicted value with the observed value to produce a smoothed and corrected trajectory;
[0011] After selecting the starting point, the point track starting submodule determines the next point track according to the position, speed, covariance and other information of the radiation source target and the size of the starting wave gate. The detection device predicts the position of the target according to the starting point track information and continuously updates the trajectory to obtain continuous observation of the same target. When three point tracks are continuous, it means that the point track is started successfully.
[0012] The trace termination submodule terminates the lost target trace, clears the trace history, and performs extrapolation tracking based on the historical traces. If no target is found after three extrapolations, the historical traces are cleared.
[0013] Furthermore, the functions of the point trace association submodule include accepting new point traces; associating and pairing the new point traces with all target tracks and providing a point trace-track pairing design; and outputting the successfully associated point traces.
[0014] Furthermore, the point track association submodule adopts the nearest point track association algorithm to select the detection point track with the smallest statistical distance as the paired point track of the target, firstly performs a rough association to obtain two track lists in the same or adjacent sectors, and then pairs all the tracks in the two lists with all the observation points in order, and uses the statistical distance between the track and the track as the basic basis for association.
[0015] Furthermore, the functions of the point trace filtering submodule include accepting successfully associated point traces; using a Kalman filter to complete the prediction of the target position and speed, and generate smoothed and corrected point traces; and outputting the filtered point traces.
[0016] Furthermore, the point trace filtering submodule receives the target position estimation value, filter parameters, and point trace and trajectory association result of the same point trace, then initializes the Kalman filter, and then uses the Kalman filter to obtain the predicted value of the next position of the target.
[0017] Furthermore, the functions of the point track starting submodule include making a rough estimate of the target's position, speed, and covariance information to provide initial values for continuous observation; predicting the target position based on the starting point track information; and continuously updating the trajectory to form the target's initial point track after the radiation source target enters the reconnaissance range.
[0018] Furthermore, the starting of the point trace of the point trace starting submodule includes the generation of a new point trace and the confirmation of the trajectory. The workflow of the point trace starting submodule includes:
[0019] Receive the estimated value of the target position and the start gate size value; register a temporary track based on the measured value; predict the next position of the target; determine whether the detection quality parameters meet the requirements. If they do, it means that the point track is started successfully. If not, continue to execute from registering the temporary point track based on the measured value.
[0020] Compared with the prior art, the present invention has the following advantages and beneficial effects:
[0021] 1. The present invention provides a target point tracking algorithm based on Kalman filtering, which uses the Kalman filtering algorithm to pre-estimate position information, so as to avoid all dynamic threats during the mission execution of the aircraft and shorten the time to reach the target area;
[0022] 2. The target point tracking algorithm based on Kalman filtering of the present invention avoids integral operation by adopting a simplified adaptive strong tracking cubature Kalman filtering (ASTCKF) algorithm to process the integrated navigation system, with a small amount of calculation, and the positive weight of each integral point in the filtering process ensures the positive definiteness of the estimated covariance, and has higher filtering accuracy and numerical stability; BRIEF DESCRIPTION OF THE DRAWINGS
[0023] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of this application, and do not constitute a limitation of the embodiments of the present invention. In the drawings:
[0024] Figure 1 This is a flow chart of the target point trace processing module of the present invention;
[0025] Figure 2 This is a flow chart of the target point trace association submodule of the present invention;
[0026] Figure 3 This is a flow chart of the target point trace filtering submodule of the present invention;
[0027] Figure 4 This is the target point trace starting submodule flow chart of the present invention;
[0028] Figure 5 This is a flow chart of the target point disappearance submodule of the present invention;
[0029] Figure 6This is a schematic diagram of the initial simulation result of the target point trace in the second embodiment of the present invention;
[0030] Figure 7 The figure is a schematic diagram of the target point trace association simulation result in the second embodiment of the present invention. DETAILED DESCRIPTION
[0031] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with embodiments and drawings. The exemplary embodiments of the present invention and their description are only used to explain the present invention and are not intended to limit the present invention.
[0032] Example
[0033] like Figures 1 to 7 As shown, the present invention is a target point track algorithm based on Kalman filtering. The algorithm is modeled through a point track processing module, and the target heading, speed, and position information are judged. After removing clutter data and interference data, the real target point track data is extracted, and then the local track is formed through navigation, batching, and batch merging processing;
[0034] The point trace processing module includes a point trace association submodule, a point trace filtering submodule, a point trace starting submodule, and a point trace termination submodule.
[0035] The point trace association submodule associates and matches the updated point trace with the historically formed point traces, and selects a current point that is most likely to belong to the target point trace by comparing the positional relationship between the extrapolated point and the current point selected in the current scanning cycle. The functions of the point trace association submodule include: accepting new point traces; associating and pairing the new point trace with all target point traces, and providing a "point trace-track" pairing design. Outputting the successfully associated point traces;
[0036] The working principle of the trace association submodule:
[0037] The Nearest Point Association (NNP) algorithm is used. The basic principle is to select the detection point with the smallest "statistical distance" as the paired point of the target. First, "rough association" is performed to obtain two track lists in the same or adjacent sectors (for airspace division, divided according to pitch angle and azimuth angle), and all tracks in the two lists are paired with all observation points in order, and the statistical distance between the track and the track (slant range interpolation or spatial statistical distance) is used as the basic basis for association.
[0038] The workflow of the target point trace association submodule is as follows Figure 2 shown.
[0039] 1) Coarse association: select all points that are in the same sector or adjacent sectors as the current point, and add them to two lists. The old point list {T i(1)} and free trace list {T i (2)};
[0040] 2) Calculate all points in the old point list {T i (1)} and all points {P j If the point trace has completed the initialization of the Kalman filter, it is predicted according to the current simulation time to obtain the target predicted position, and then the spatial statistical distance {l ij}; If the Kalman filter has not been established for the point trace, the spatial statistical distance with all the points traces is directly calculated using the most recent observation position of the point trace {l ij In addition, when calculating the spatial statistical distance, first determine whether the detection point is within the wave gate of the trajectory. If not, let {l ij}=∞;
[0041] 3) Select the matrix {l ij The minimum value in
[0042] 4) Traverse the entire matrix {l ij}Delete the row and column corresponding to the point and trajectory. If the value is not ∞, the detection point is paired with the trajectory and the matrix {l ij} to delete the row and column corresponding to the point and trajectory;
[0043] 5) Repeat step (3) until {l ij} has no non-∞ elements, so we get all the old points {T i (1)}pairing results;
[0044] 6) Repeat the steps (2), (3), and (4) for the free point trace and the remaining points to obtain the free point trace {T i (2)} pairing results.
[0045] 7) If the successfully matched point trace is the target track for the current point trace event, the track is filtered and updated using the successfully matched detection point trace, and predicted at a certain data rate.
[0046] The point trace filter module completes the prediction of the target position and speed, and combines the predicted value with the observed value to generate a smoothed and corrected trajectory. The functions of the point trace filter module include: accepting the successfully associated point traces; using the Kalman filter to complete the prediction of the target position and speed, generate a smoothed and corrected trajectory, and output the filtered trajectory;
[0047] The commonly used filters for point trace filtering and tracking processing are Kalman filters. The filter equation group of Kalman filters is a recursive form in the time domain, which is completely consistent with the above analysis conclusions. At the same time, the Kalman filter has a prediction-correction structure and has real-time processing characteristics.
[0048] The workflow of the point trace filter submodule is as follows Figure 3 shown.
[0049] Point trace start submodule
[0050] For targets that meet the conditions for establishing a track, the track should be automatically established. After selecting the starting track, the track start submodule determines the next track based on the position, speed, covariance and other information of the radiation source target and the size of the starting wave gate. The detection and receiving equipment predicts the position of the target based on the starting track information and continuously updates the track to obtain continuous observation of the same target. When three tracks are continuous, it means that the track start is successful.
[0051] The functions of the point trace initiation submodule include: making a rough estimate of the target's position, speed, covariance and other information to provide initial values for continuous observation; predicting the target position based on the starting point trace information; continuously updating the point trace, and finally forming the target initial point trace after the radiation source target enters the reconnaissance range.
[0052] The working principle of the dot trace start submodule:
[0053] The start of a track includes the generation of a new track and the confirmation of the track. The generation of a new track may occur during the search time or during the tracking event. The confirmation of the track generally adopts the "K / M" criterion. If the number of successful pairings of the target track is greater than or equal to K in M observations, the track is confirmed successfully, otherwise the temporary track is cancelled. In addition, the M observations here refer to the observations made during the confirmation or tracking task.
[0054] The workflow of the dot trace start submodule is as follows Figure 4 shown.
[0055] Dot trace terminator module
[0056] The trace termination submodule terminates the trace of the lost target and clears the trace history. Extrapolation tracking is performed based on the historical traces. If the target is not found after three extrapolations, the historical traces are cleared.
[0057] The functions of the trace termination submodule include: extrapolating tracking based on historical traces; if no target is found after three extrapolations, clearing the historical traces.
[0058] The working principle of the trace terminator module:
[0059] The termination rule of the point trace generally uses the "K / M" criterion. In the M consecutive observations of reliable points and precise tracking points, if the number of successful association pairing is less than K times, the point trace will be terminated. In addition, after the precise tracking point trace is terminated, the missing processing will be called, and a search task will be arranged near the most recent filtering position of the precise tracking point trace in the hope of recapturing the lost target.
[0060] The workflow of the trace termination submodule is as follows Figure 5 shown.
[0061] Embodiment 2
[0062] This embodiment performs specific operations on the basis of the first embodiment, and the target point trace is processed as follows: Figure 6 and Figure 7 As shown in the figure, the result of the dot trace initiation algorithm in data processing is shown. In the figure, 5 simulation targets are set, and the total number of batches required for the dot trace initiation is 4. When 4 batches of data contain 3 batches of the same target measurement information, the dot trace of the target is established. It can be seen from 0 that the 5 preset target dot traces have been successfully established.
[0063] When the point trace starts, the point trace is filtered to predict its future state, and then the predicted result is associated with the measurement information to obtain the actual target position. Figure 7 It can be seen that this method effectively tracks the two preset targets.
[0064] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.
Claims
1. A target point tracking algorithm based on Kalman filtering, It is characterized in that The algorithm is modeled through the point trace processing module, determines the target heading, speed, and position information, and removes clutter data and interference data to extract the real target point trace data, and then forms local target point traces through navigation, batching, and batch merging processing; The point trace processing module includes a point trace association submodule, a point trace filtering submodule, a point trace starting submodule, and a point trace termination submodule; The point trace association submodule associates and matches the updated point trace with the historically formed point trace, and selects a point with a high degree of association as the current point of the target point trace by comparing the positional relationship between the externally pushed point and the current point selected in the current scanning cycle; The point trace filter submodule completes the prediction of the target position and speed, and combines the predicted value with the observed value to produce a smoothed and corrected trajectory; After selecting the starting point, the point track starting submodule determines the next point track according to the position, speed, covariance and other information of the radiation source target and the size of the starting wave gate. The detection device predicts the position of the target according to the starting point track information and continuously updates the trajectory to obtain continuous observation of the same target. When three point tracks are continuous, it means that the point track is started successfully. The point trace termination submodule terminates the point trace of the lost target, clears the point trace history, and performs extrapolation tracking based on the historical point traces. If the target is not found after three extrapolations, the historical point traces are cleared.
2. A target point tracking algorithm based on Kalman filtering according to claim 1, It is characterized in that The functions of the point trace association submodule include accepting new point traces; associating and pairing the new point traces with all target point traces and providing a point trace-track pairing design; and outputting the successfully associated point traces.
3. The target point tracking algorithm based on Kalman filtering according to claim 2, It is characterized in that The point trace association submodule adopts the nearest point trace association algorithm, selects the detection point trace with the smallest statistical distance as the paired point trace of the target, first performs a rough association to obtain two point trace lists in the same or adjacent sectors, and then pairs all the point traces in the two lists with all the observation points in order, and uses the statistical distance between the point trace and the track as the basic basis for association.
4. The target point tracking algorithm based on Kalman filtering according to claim 1, It is characterized in that The functions of the point trace filter submodule include accepting successfully associated point traces; using the Kalman filter to complete the prediction of the target position and speed, and generate a smoothed and corrected trajectory; Output the filtered trace.
5. The target point tracking algorithm based on Kalman filtering according to claim 4, It is characterized in that The point trace filtering submodule receives the target position estimation value of the same point trace, filter parameters, and the association result of the new point trace with the existing point trace, then initializes the Kalman filter, and then uses the Kalman filter to obtain the predicted value of the next position of the target.
6. The point trace algorithm based on Kalman filtering according to claim 1, It is characterized in that The functions of the point trace starting submodule include making a rough estimate of the position, speed, and covariance information of the target to provide an initial value for continuous observation; predicting the target position based on the starting point trace information; The point trajectory is continuously updated to form the initial trajectory of the target after the radiation source target enters the reconnaissance range.
7. The target point tracking algorithm based on Kalman filtering according to claim 6, It is characterized in that The starting of the point trace of the point trace starting submodule includes the generation of a new point trace and the confirmation of the point trace. The working process of the point trace starting submodule includes receiving the target position estimation value and the starting gate size value; registering a temporary track according to the measurement value; predicting the next position of the target; judging whether the detection quality parameter meets the requirements, if it meets the requirements, it means that the point trace is started successfully, if not, continue to execute from registering the temporary point trace according to the measurement value.
