Multi-target tracking planning method and system based on fixed-wing unmanned aerial vehicle

Abstract

The invention provides a multi-target tracking planning method and system based on a fixed-wing unmanned aerial vehicle, and the method comprises the steps: obtaining the dynamic target data of the fixed-wing unmanned aerial vehicle, carrying out the Kalman filtering of the dynamic target data, and outputting the filtered dynamic target data; based on the filtered dynamic target data, target screening and hierarchical processing are carried out to obtain a plurality of target subsets, Hough transform is adopted to carry out spatial distribution feature extraction on each target subset in sequence, and a target tracking sequence is generated; according to the relative position of the target tracking sequence and the fixed-wing unmanned aerial vehicle, the guide point is dynamically calculated, the optimal flight path from the fixed-wing unmanned aerial vehicle to the guide point is generated through a preset path planning algorithm, the fixed-wing unmanned aerial vehicle is controlled to track the dynamic target, and the tracking of the fixed-wing unmanned aerial vehicle on multiple dynamic targets is realized. The calculation and tracking efficiency of multi-target tracking is further improved, and the autonomy and the intelligent level of the fixed-wing unmanned aerial vehicle in the multi-target tracking process are improved to a great extent.

Description

Technical Field

[0001] This invention relates to the field of multi-target tracking and planning technology for fixed-wing unmanned aerial vehicles (UAVs), and specifically to a multi-target tracking and planning method and system based on fixed-wing UAVs. Background Technology

[0002] Unmanned aerial vehicle (UAV) swarms are primarily used in the military field. Based on the combat capabilities of individual UAVs, they are supported by autonomous collaboration among UAVs. Integrating based on an open architecture, they possess high levels of intelligence, autonomy, and functional distribution, enabling them to perform specific combat missions. UAV swarm technology, with its low-cost units, can asymmetrically offset the capabilities of traditional large, multi-functional platforms, attracting the attention of militaries and military organizations worldwide and leading to its rapid development. In this scenario, facing complex dynamic multi-target situations, how to effectively destroy and counter these targets has become a key issue. This problem can generally be decomposed into multi-target identification, path planning for fixed-wing UAVs targeting multiple dynamic targets, and autonomous tracking. Among these, path planning for fixed-wing UAVs targeting multiple dynamic targets is a significant technical challenge. This involves planning the path for a UAV to traverse all target points given their motion information.

[0003] Researchers have studied drone path planning in UAV / UGV collaborative tasks under a messenger mechanism. They modeled the UAV path planning problem as a Dynamic Dubins Travelling Salesman Problem with Neighborhood (DDTSPN), where multiple dynamic target points represent multiple moving UGVs. Using a decoupling method, DDTSPN was decomposed into two sub-problems: the first sub-problem is solving for the access order of target points, which uses a heuristic algorithm to solve the static TSP; the second sub-problem is optimizing the access point locations within the UGV communication neighborhood, employing a sampling-based optimization method. The algorithm employs a rolling optimization strategy during implementation. However, when solving the access order of dynamic target points, the algorithm can obtain a good access order when the number of UGVs is small and the distribution structure is simple. But if the UGV distribution structure is more complex, it cannot obtain a reasonable access order. In addition, the sampling method used to determine the UAV / UGV meeting point has a result that largely depends on the number of samplings. If the number of samplings is too small, the optimization result may be unsatisfactory. On the other hand, if a more complex method is used, the calculation time is too long, which makes it impossible to meet the need for fast calculation.

[0004] Researchers also addressed the path planning problem of UAVs transmitting information to UGVs across all moving objects. They modeled this problem as a DDTSPN with priority constraints and proposed a decoupling strategy to quickly determine the order in which the UAV visits the task points and the UGVs, as well as the UAV's location within the communication neighborhood of each task point and each UGV, generating a high-quality solution within a reasonable timeframe. Therefore, current research on UAV path planning in scenarios with multiple dynamic target points is insufficient, especially regarding the case where the target point motion information is non-cooperative.

[0005] Therefore, the existing technology still needs further development. Summary of the Invention

[0006] The purpose of this invention is to overcome the above-mentioned technical deficiencies and provide a multi-target tracking planning method and system based on fixed-wing unmanned aerial vehicles (UAVs) to solve the problems existing in the prior art.

[0007] To achieve the above-mentioned technical objectives, according to a first aspect of the present invention, the present invention provides a multi-target tracking planning method based on a fixed-wing unmanned aerial vehicle, comprising: S100: Acquire dynamic target data of a fixed-wing UAV, perform Kalman filtering on the dynamic target data, and output the filtered dynamic target data; S200. Based on the filtered dynamic target data, target screening and hierarchical processing are performed to obtain multiple target subsets. The Hough transform is used to extract the spatial distribution features of each target subset in turn to generate a target tracking sequence. S300. Based on the relative position of the target tracking sequence and the fixed-wing UAV, dynamically calculate the guidance point, generate the optimal flight path from the fixed-wing UAV to the guidance point through a preset path planning algorithm, and control the fixed-wing UAV to track the dynamic target.

[0008] Specifically, based on the filtered dynamic target data, target filtering and hierarchical processing are performed to obtain multiple target subsets, including: Based on the filtered dynamic target data, a preset target filtering mechanism is used to remove false alarm targets, and the target set after removing false alarm targets is output. The target set after removing false alarm targets is then processed into multiple target subsets according to altitude.

[0009] Specifically, the step of extracting spatial distribution features from each target subset using the Hough transform to generate a target tracking sequence includes: Transform the dynamic target point set in the target subset to Hough space; The first Hough region with the highest density in the Hough space was identified by cluster analysis, and the optimal fitted line of the first Hough region was analyzed. The dynamic targets in the target subset are projected onto the optimal fitted line, and a target tracking sequence is generated according to the order of the projection points. Based on the relative motion direction between the fixed-wing UAV and the target tracking sequence, the tracking mode of the fixed-wing UAV is determined.

[0010] Specifically, determining the tracking mode of the fixed-wing UAV based on the relative motion direction between the fixed-wing UAV and the target tracking sequence includes: If the fixed-wing UAV moves in the opposite direction to the target tracking sequence, then the tracking mode of the fixed-wing UAV is a head-on tracking sequence. If a fixed-wing UAV moves in the same direction as the target in the tracking sequence, then the tracking mode of the fixed-wing UAV is a tail-chase tracking sequence.

[0011] Specifically, the step of dynamically calculating the guidance point based on the relative position of the target tracking sequence and the fixed-wing UAV includes: Calculate the relative distance between the center point of the target tracking sequence and the fixed-wing UAV; Determine whether the relative distance is greater than a first preset threshold, determine the guidance method of the fixed-wing UAV based on the determination result, and then calculate the guidance point based on the guidance method.

[0012] Specifically, determining the guidance method for the fixed-wing UAV based on the judgment result includes: When the relative distance is greater than the first preset threshold, the guidance mode of the fixed-wing UAV is predictive guidance mode. When the relative distance is less than or equal to the first preset threshold, the guidance mode of the fixed-wing UAV is real-time guidance mode.

[0013] Specifically, calculating the guidance point according to the guidance method includes: If the guidance method of the fixed-wing UAV is predictive guidance, then the meeting point between the fixed-wing UAV and the target tracking sequence is used as the guidance point. If the guidance method of the fixed-wing UAV is real-time guidance, then the spatial projection point of the current target tracking sequence is used as the guidance point.

[0014] Specifically, the prediction guidance method includes: Based on the Kalman filter algorithm, the motion state of each target in the target tracking sequence is estimated, the motion state of the target tracking sequence is predicted at multiple future time points, and the target prediction trajectory is generated. Based on the current motion status data of the fixed-wing UAV, calculate the flight time required for the fixed-wing UAV to fly from its current position to any point on the predicted trajectory of the target. Based on the flight time, the coordinates of the preceding predicted points of the fixed-wing UAV on the target predicted trajectory are obtained by interpolation on the target predicted trajectory. The preceding predicted points are used as the guidance target points of the fixed-wing UAV. Least square straight line fitting is performed on multiple consecutive preceding predicted points to generate a predicted guidance straight line.

[0015] Specifically, generating the optimal flight path from the fixed-wing UAV to the guidance point using a preset path planning algorithm includes: The Dubins path planning method is adopted. Starting from the current position of the fixed-wing UAV and ending at the guidance point, the lengths of multiple Dubins paths are calculated based on the minimum turning radius of the fixed-wing UAV. The Dubins path set is output, and the shortest Dubins path in the set is taken as the optimal Dubins path for the current fixed-wing UAV.

[0016] According to a second aspect of the present invention, a multi-target tracking and planning system based on a fixed-wing unmanned aerial vehicle is provided, comprising: Target filtering module: used to acquire dynamic target data of fixed-wing UAV, perform Kalman filtering on the dynamic target data, and output filtered dynamic target data; Target filtering module: Used to filter and stratify targets based on the filtered dynamic target data to obtain multiple target subsets. The Hough transform is used to extract the spatial distribution features of each target subset in turn to generate a target tracking sequence. Path planning module: It is used to dynamically calculate the guidance point based on the relative position of the target tracking sequence and the fixed-wing UAV, generate the optimal flight path of the fixed-wing UAV to the guidance point through a preset path planning algorithm, and control the fixed-wing UAV to track the dynamic target.

[0017] Beneficial effects: This invention provides a multi-target tracking planning method and system based on a fixed-wing UAV. It involves performing Kalman filtering on dynamic target data, outputting filtered dynamic target data, and then performing target filtering and hierarchical processing to obtain multiple target subsets. Hough transform is then used to extract spatial distribution features from each target subset, generating a target tracking sequence. Based on the relative position of the target tracking sequence and the fixed-wing UAV, a guidance point is dynamically calculated. A preset path planning algorithm generates the optimal flight path from the fixed-wing UAV to the guidance point, controlling the fixed-wing UAV to track dynamic targets. This achieves the tracking of multiple dynamic targets by the fixed-wing UAV, further improving the computational and tracking efficiency of multi-target tracking. This invention enables flexible responses to complex and changing flight environments. By updating target data and path planning in real time, it ensures that the fixed-wing UAV can accurately track targets, improving the accuracy and efficiency of data processing and significantly enhancing the autonomy and intelligence level of the fixed-wing UAV in the multi-target tracking process. Attached Figure Description

[0018] Figure 1 This is a flowchart of a multi-target tracking and planning method based on a fixed-wing UAV provided in a specific embodiment of the present invention; Figure 2 This is a schematic diagram of the composition of a multi-target tracking and planning system based on a fixed-wing UAV provided in a specific embodiment of the present invention; Figure 3 This is a schematic diagram of the calculation process of the Kalman filter algorithm provided in a specific embodiment of the present invention; Figure 4 This is a flowchart of the screening process for false alarm targets provided in a specific embodiment of the present invention; Figure 5 This is a flowchart illustrating the selection of target tracking sequence tracking modes provided in a specific embodiment of the present invention; Figure 6 This is a schematic diagram of Dubins path planning provided in a specific embodiment of the present invention; Figure 7 This is a schematic diagram of the planar coordinate to Dubins coordinate conversion provided in a specific embodiment of the present invention. Detailed Implementation

[0019] To enable those skilled in the art to better understand the technical solutions of the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Based on the embodiments in this application, other similar embodiments obtained by those skilled in the art without creative effort should all fall within the scope of protection of this application. Furthermore, directional terms mentioned in the following embodiments, such as "up," "down," "left," and "right," are only for reference to the directions in the accompanying drawings; therefore, the directional terms used are for illustrative purposes and not for limiting the invention.

[0020] The present invention will be further described below with reference to the accompanying drawings and preferred embodiments.

[0021] Example 1 Please see Figure 1 This embodiment provides a multi-target tracking planning method based on a fixed-wing unmanned aerial vehicle (UAV), and the specific implementation method is as follows: S100: Acquire dynamic target data of a fixed-wing UAV, perform Kalman filtering on the dynamic target data, and output the filtered dynamic target data; It is understood that this invention primarily addresses path planning for tracking dynamic targets by fixed-wing UAVs. Dynamic targets refer to multiple targets being tracked (such as enemy UAV swarms, vehicles, etc.) that are in real-time motion, with their positions and speeds changing over time, rather than being stationary targets. Furthermore, the target's motion pattern is unknown, thus requiring real-time prediction of the target's trajectory. Simultaneously, the fixed-wing UAV platform performing the tracking task itself has strong motion constraints. The fixed-wing UAV must fly continuously, cannot hover or stop abruptly, and is limited by a minimum turning radius, preventing instantaneous turns and requiring smooth maneuvers. Therefore, path planning must meet certain physical constraints. This invention achieves efficient and robust multi-target tracking in complex scenarios where both dynamic targets and fixed-wing UAVs are in motion.

[0022] Furthermore, the dynamic target data in this embodiment can be acquired through the high-altitude positioning and aerial photography system built into the fixed-wing UAV. The dynamic target data includes target position data and target motion state data. The target position data is the position of the dynamic target in the planar navigation coordinate system, and the target motion state data is the flight speed of the dynamic target. Kalman filtering is performed on the dynamic target data, and the specific filtering process is as follows: Step S1.1: Establish the target state vector according to the specific scenario of target tracking, ignoring the change of the target in the height direction. The target state can be represented as: ; In the formula, The position of a dynamic target in a planar navigation coordinate system can be detected by detection equipment. The target's flight speed; Step S1.2, according to Kalman filter theory, assuming the target's motion state conforms to the CV (constant velocity) motion model, the target's state transition matrix Φ can be expressed as: ; ; Among them, T S This is the interval for filter updates. Let K be the state vector of the target at time K. The state vector of the target at time K+1 is obtained based on the state transition matrix Φ.

[0023] The target's velocity is obtained by measuring the target's velocity and updating it using the Kalman filter algorithm, thus estimating the target's state.

[0024] The principle of the Kalman filter algorithm is as follows: Figure 3 As shown: the target state estimate at the previous time step. To the next moment The calculation process is an iterative loop process, and the filter output is... State estimate. From Prediction mean square error to The calculation of the mean squared error is another iterative process, mainly involving the calculation of the filter gain. The entire filtering process requires providing the initial values ​​for the filter. and Using measured and estimated values, the state estimates of navigation parameters can be obtained through recursive calculation.

[0025] Specifically, in the Kalman filter algorithm, the state estimate from the previous time step is used to predict the prior estimate for the next time step through the state transition matrix. This process considers the dynamic characteristics of the system. The output of the filter is the state estimate for this time step, which integrates the prior estimate and the current observation information. The calculation from the mean square error of the prior estimate (prediction mean square error) to the mean square error of the posterior estimate (estimation mean square error) aims to assess the uncertainty of the prediction and adjust the state estimate accordingly to make it closer to the true value. This process is mainly achieved by calculating the filter gain, which reflects the degree to which the observation information corrects the state estimate. Before the entire filtering process begins, the initial values ​​of the filter need to be set, including the initial estimated value of the state and the initial mean square error estimate. The selection of these initial values ​​has an important impact on the convergence speed and accuracy of the filter. During the filtering process, the state information of the system can be continuously updated by using the continuously obtained measurement values ​​and the state estimates obtained through recursive calculation, so as to achieve accurate tracking of the target state and improve the real-time performance and accuracy of tracking.

[0026] S200. Based on the filtered dynamic target data, target screening and hierarchical processing are performed to obtain multiple target subsets. The spatial distribution features of each target subset are extracted sequentially using Hough transform to generate a target tracking sequence.

[0027] Specifically, based on the filtered dynamic target data, target filtering and hierarchical processing are performed to obtain multiple target subsets, including: Based on the filtered dynamic target data, a preset target screening mechanism is used to remove false alarm targets, and the target set after removing false alarm targets is output. The target set after removing false alarm targets is then processed in layers according to altitude to form multiple target subsets at different altitudes. The target subsets are processed sequentially in descending order of altitude, with priority given to processing the highest level target subset.

[0028] In this embodiment, Kalman filtering is applied to the filtered dynamic target data to screen and stratify the targets. This mainly includes autonomous target screening and target stratification, prioritizing the tracking of the highest-level targets. Specifically, this includes: Step S2.1: Based on the Kalman filter algorithm, relatively accurate dynamic target data is first obtained. Then, through the target filtering module, a preset target filtering mechanism is used to remove errors and false alarm targets caused by the instability of the detection equipment. The preset target filtering mechanism includes velocity gate filtering, time gate filtering, and cluster position filtering to obtain the priority group targets, that is, the highest-level target subset, such as... Figure 4 As shown, the implementation method of the preset target filtering mechanism is as follows; (1) Speed ​​gate screening: Assume the maximum speed of the trackable target is denoted as . The velocity of each detected target is denoted as , If the total number of detected targets is , then if If so, the target is considered valid; (2) Time gate filtering: Since false alarm targets are generally caused by cloud cover or detectors themselves, they usually do not exist for a long time. Therefore, it is assumed that the longest time a false alarm target exists is... The time of the first appearance of each target is counted as The target duration is ,like If so, the target is considered valid; (3) Cluster location filtering: Using the clustering method, targets with the same state information and direction of movement are grouped into a small tracking group in the target cluster, and the target with the largest number of groups is processed first. Step S2.2, cluster target hierarchical classification: Assuming the entire cluster can be discovered completely before entering the target locking phase, and the highest and lowest target heights can be obtained, the targets are then hierarchically clustered from high to low to obtain a pre-tracked single-layer group of targets, i.e., a subset of targets, as follows: Assuming the entire cluster can be discovered intact before entering the target locking phase, the highest target height can be obtained. With the lowest target height This allows the entire cluster to be divided into The system first stratifies the target data into layers, then performs hierarchical clustering based on the hierarchy from highest to lowest level. In the formula... This represents the maximum thickness of a single target layer after considering the target's effective range.

[0029] Specifically, in this embodiment, the step of extracting spatial distribution features from each target subset sequentially using the Hough transform to generate a target tracking sequence includes: The dynamic target point set in the target subset is transformed into Hough space in turn; The first Hough region with the highest density in the Hough space was identified by cluster analysis, and the optimal fitted line of the first Hough region was analyzed. The dynamic targets in the target subset are projected onto the optimal fitted line, and a target tracking sequence is generated according to the order of the projection points. Based on the relative motion direction between the fixed-wing UAV and the target tracking sequence, the tracking mode of the fixed-wing UAV is determined.

[0030] Specifically, determining the tracking mode of the fixed-wing UAV based on the relative motion direction between the fixed-wing UAV and the target tracking sequence includes: If the fixed-wing UAV moves in the opposite direction to the target tracking sequence, then the tracking mode of the fixed-wing UAV is a head-on tracking sequence. If a fixed-wing UAV moves in the same direction as the target in the tracking sequence, then the tracking mode of the fixed-wing UAV is a tail-chase tracking sequence.

[0031] Understandably, according to the above technical solution, the discrete dynamic target group is efficiently condensed into the optimal fitted straight line path through the Hough transform, and anti-interference tracking commands are generated based on the projection point sequence. At the same time, the head-on / tail-on tracking mode is dynamically switched by combining the relative motion direction of the UAV and the target group. In complex battlefield environments, the real-time performance and robustness of multi-target tracking are significantly improved. It effectively overcomes the dependence of traditional methods on target cooperation information, embeds fixed-wing UAV maneuver constraints, and achieves the unity of engineering practicality and algorithm efficiency. It provides an innovative solution for the dual-motion coupling scenario of dynamic targets and dynamic platforms.

[0032] Furthermore, the specific implementation steps for generating a target tracking sequence are illustrated below with concrete examples: After obtaining the pre-tracked single-layer group of targets, i.e., each target subset, a straight-line target planning algorithm is designed to meet the requirements of motion process and real-time high-precision guidance for fixed-wing UAVs. The algorithm mainly includes the following steps: Step S3.1, assuming the positions of all dynamic target points are... , where n is the total number of targets in a single-layer group (a subset of targets), resulting in multiple Hough space target points; ; in, This represents the Hough space target point corresponding to the dynamic target point. This represents the constant coefficient.

[0033] Step S3.2 involves performing Hough voting, which clusters multiple points in the obtained Hough space to obtain clusters within a certain error range. The Hough region with the most cluster points, which is also the Hough region with the highest density, is denoted as the first Hough region. The center point of this region is calculated, and the spatial line corresponding to the center point is the line with the most targets in the single-layer group of targets to be tracked. Step S3.3, target tracking sequence generation. The target tracking sequence considers how the fixed-wing UAV and the target cluster move to ensure that the tracking time is minimized. That is, the dynamic targets in the target subset are projected onto the fitted straight line, and the target tracking sequence is generated according to the order of the projection points. Specifically, the target tracking sequence is obtained by calculating the relative motion state between the fixed-wing UAV and the target in the straight line and sorting the target in the straight line.

[0034] Step S3.4: Determine the planar positional relationship between the fixed-wing UAV and the entire target tracking sequence. Assume the target tracking sequence obtained after target detection via linear band tracking is as follows: , This refers to the number of targets that are expected to be tracked in this plan, and each target... The two-dimensional position in the plane is: The speed is Fixed-wing UAV position The speed is First, calculate the center point coordinates of the tracking target sequence points. ,by Let O be the origin of the coordinate system. Using the coordinate axes, calculate Based on the position in the sequence coordinate system, determine the planar positional relationship between the fixed-wing UAV and the entire tracking sequence: (1) If the fixed-wing UAV is ahead of the entire tracking sequence, that is, when the fixed-wing UAV moves in the opposite direction to the target tracking sequence, then the head-on tracking sequence is selected; (2) If the fixed-wing UAV is behind the target in the entire tracking sequence, and if the fixed-wing UAV moves in the same direction as the target tracking sequence, then the tail-chase tracking sequence is selected.

[0035] The methods for determining head-on and tail-on tracking include: calculation. The average velocity vector of each target in the velocity vector The initial tracking direction vector for target tracking is: Calculate the average velocity vector With the initial tracking direction vector The included angle: ; remember for like Then it means head-on pursuit. Then it is called tail tracking.

[0036] S300. Based on the relative position of the target tracking sequence and the fixed-wing UAV, dynamically calculate the guidance point, generate the optimal flight path from the fixed-wing UAV to the guidance point through a preset path planning algorithm, and control the fixed-wing UAV to track the dynamic target.

[0037] Specifically, the step of dynamically calculating the guidance point based on the relative position of the target tracking sequence and the fixed-wing UAV includes: Calculate the relative distance between the center point of the target tracking sequence and the fixed-wing UAV; Determine whether the relative distance is greater than a first preset threshold, determine the guidance method of the fixed-wing UAV based on the determination result, and then calculate the guidance point based on the guidance method.

[0038] Furthermore, when the relative distance is greater than the first preset threshold, the guidance mode of the fixed-wing UAV is predictive guidance; when the relative distance is less than or equal to the first preset threshold, the guidance mode of the fixed-wing UAV is real-time guidance.

[0039] Specifically, if the guidance method of the fixed-wing UAV is predictive guidance, the meeting point between the fixed-wing UAV and the target tracking sequence is used as the guidance point; if the guidance method of the fixed-wing UAV is real-time guidance, the spatial projection point of the current target tracking sequence is used as the guidance point.

[0040] It should be further explained that, according to the above technical solution, the guidance mode is dynamically switched by a relative distance threshold. When the distance is long, predictive guidance is used to calculate the future meeting point in advance, and when the distance is short, real-time guidance is switched to directly capture the current position of the target. This effectively balances the foresight of path planning and the accuracy of terminal tracking, and significantly improves the autonomous tracking and adaptability of fixed-wing UAVs to dynamic target groups in complex battlefield environments.

[0041] Specifically, in this embodiment, the prediction guidance method includes: Based on the Kalman filter algorithm, the motion state of each target in the target tracking sequence is estimated, the motion state of the target tracking sequence is predicted at multiple future time points, and the target prediction trajectory is generated. Based on the current motion status data of the fixed-wing UAV, calculate the flight time required for the fixed-wing UAV to fly from its current position to any point on the predicted trajectory of the target. Based on the flight time, the coordinates of the preceding predicted points of the fixed-wing UAV on the target predicted trajectory are obtained by interpolation on the target predicted trajectory. The preceding predicted points are used as the guidance target points of the fixed-wing UAV. Least square straight line fitting is performed on multiple consecutive preceding predicted points to generate a predicted guidance straight line.

[0042] Specifically, in this embodiment, generating the optimal flight path from the fixed-wing UAV to the guidance point using a preset path planning algorithm includes: using the Dubins path planning method, taking the current position of the fixed-wing UAV as the starting point and the guidance point as the ending point, calculating the lengths of multiple Dubins paths based on the minimum turning radius of the fixed-wing UAV, outputting a set of Dubins paths, and taking the shortest Dubins path in the set as the optimal Dubins path for the current fixed-wing UAV.

[0043] The following example illustrates the specific steps for obtaining the optimal flight path of a fixed-wing UAV. After calculating the target tracking sequence, a predetermined guidance line is given by predicting the motion state of the dynamic target in the target tracking sequence to improve the guidance accuracy of the platform. The flight path control of the fixed-wing UAV platform is implemented based on the Dubins path planning method, mainly including the following steps: Step S4.1 employs the idea of ​​filtering prediction to predict the future position and state of the cluster target, aiming to output a relatively accurate predetermined guide line at the guidance end. However, the prediction algorithm's error increases with the prediction time, showing a positive correlation. Therefore, it is necessary to select a guidance method. The selection process is as follows: Figure 5 As shown, the main steps include: (1) Calculate the position of the center point of the target tracking sequence; (2) Determine the relative distance D between the center point and the fixed-wing UAV platform; (3) The prediction guidance and real-time guidance modes are selected by judging the relative distance D and the first preset threshold d that meets the guidance accuracy requirements. The calculation process of the guidance mode is as follows: a. Real-time guidance Real-time guidance involves using data from a data source that has undergone simple preprocessing as guidance data. The main processing steps are as follows: The target tracking sequence is obtained by using the least squares method. The target line is obtained by performing least squares fitting. ; With the first tracking target point Calculate for the origin of the coordinate system , On the target line Projection on , ; Will , As the target point is redirected to the planar navigation coordinate system, the positioning coordinates of the first guide point are taken as the endpoint of the Dubins path planning, and the vector connecting the two points is taken as the direction of the path. The guidance parameters are then calculated. .

[0044] b. Predictive Guidance The predictive guidance algorithm calculates the target's position at the moment when the fixed-wing UAV platform encounters the dynamic target, using the predicted target tracking point as the target. It then uses this position as the actual guidance point for path planning. The main processing steps are as follows: First, calculate the time required to reach the real-time guidance point based on the path obtained from the real-time guidance of the fixed-wing UAV platform. ; The motion state of each target is obtained based on the Kalman filter model, which for the uniform motion model is the velocity of each target. , Assume the location of each target is: The predicted position ; Based on the prediction point Calculate the distance between the future position of each target and the future position of the first target. Assuming the platform has a fixed speed, the magnitude of which is... This gives the time difference between the fixed-wing UAV reaching the first target and then reaching each subsequent target. ; The coordinates of the preceding prediction point are then obtained as follows: ; The least squares method was used to analyze the sequence of previous prediction points. The target line is obtained by performing least squares fitting. ; With the first tracking target point Calculate for the origin of the coordinate system , On the target line Projection on , ; Will , As the guiding target point is transferred to the planar navigation coordinate system, the GPS coordinates of the first guiding point are taken as the endpoint of the Dubins path planning, and the vector of the line connecting the two points is taken as the direction of the path. The guidance parameters are then calculated. .

[0045] After inputting the parameters for Dubins path planning, the flight path of the fixed-wing UAV is calculated, mainly including the following steps: Step S5.1: Under the condition of maintaining the fixed wing's uniform motion, establish a planning model based on the fixed wing's motion constraints. Generally, the input parameters for Dubins path planning are as follows: Given the initial Dubins state , This is the initial state, where , The coordinates of the starting position. The initial heading angle, The state is terminated. Let β be the coordinates of the termination position, β be the termination heading angle, and β be the minimum turning radius. The goal is to calculate the length of the Dubins path.

[0046] Step S5.2: Based on several feasible paths for the fixed-wing UAV, calculate the arc length and straight-line distance of the circular motion, and compare these paths to select the shortest path. In this embodiment, four paths are selected for calculation: LSL (Left-Straight-Left), RSR (Right-Straight-Right), LSR (Left-Straight-Right), and RSL (Right-Straight-Left). Figure 6 As shown, Figure 6 The flight paths of the four routes mentioned above are shown.

[0047] The following uses LSL (Left-Straight-Left) as an example to illustrate the path calculation process: First, a coordinate system transformation is performed, converting the original coordinate system to the new coordinate system. axis, axis to With the origin as the starting point, arrive The vector is the new The axis, the vector perpendicular to the new x-axis is the new Axis, such as Figure 7 As shown; After coordinate transformation, in the new coordinate system, the initial Dubins state and the target Dubins state are respectively... .in, From arrive The angles of the vector in the original coordinate system have the following relationship: ; The length of the Dubins path is defined as three consecutive parts, namely: ; Where L is the total length of the Dubins path, t is the length of a straight segment of the three motion paths, and p and q are the lengths of the two turning segments of the three motion paths; Solve the following three equations simultaneously: ; After solving, we get ; get: ; Finally, the lengths of the four paths were calculated, and the shortest path was obtained by comparison. The shortest path was then used as the optimal path for the fixed-wing UAV to track dynamic targets.

[0048] Step S5.3 involves outputting the path information to the fixed-wing UAV at a frequency of 1 second for flight control, thereby achieving flight control of the fixed-wing UAV. By calculating the shortest Dubins path that satisfies the minimum turning radius constraint in real time and outputting route commands at a certain frequency, the trajectory smoothness and system stability of dynamic target tracking are significantly improved while ensuring the physical maneuverability of the fixed-wing UAV, further realizing efficient and reliable autonomous flight control.

[0049] It should be noted that this embodiment provides a multi-target tracking planning method based on a fixed-wing UAV. Kalman filtering is applied to dynamic target data to output filtered dynamic target data. Then, target filtering and hierarchical processing are performed to obtain multiple target subsets. Hough transform is used to extract spatial distribution features from each target subset sequentially, generating a target tracking sequence. Based on the relative position of the target tracking sequence and the fixed-wing UAV, a guidance point is dynamically calculated. An optimal flight path from the fixed-wing UAV to the guidance point is generated using a preset path planning algorithm, controlling the fixed-wing UAV to track dynamic targets. This achieves the tracking of multiple dynamic targets by the fixed-wing UAV, further improving the computational and tracking efficiency of multi-target tracking. This invention enables flexible responses to complex and changing flight environments. By updating target data and path planning in real time, it ensures that the fixed-wing UAV can accurately track targets, improving the accuracy and efficiency of data processing and significantly enhancing the autonomy and intelligence level of the fixed-wing UAV in the multi-target tracking process.

[0050] Example 2 Please see Figure 2 This embodiment provides a multi-target tracking and planning system based on a fixed-wing unmanned aerial vehicle (UAV), the system comprising: Target filtering module 100: used to acquire dynamic target data of fixed-wing UAV, perform Kalman filtering on the dynamic target data, and output filtered dynamic target data; Target filtering module 200: It is used to filter and stratify targets based on target location data and target motion state data to obtain multiple target subsets. The Hough transform is used to extract spatial distribution features of each target subset in turn to generate a target tracking sequence. Path planning module 300: It is used to dynamically calculate the guidance point based on the relative position of the target tracking sequence and the fixed-wing UAV, generate the optimal flight path from the fixed-wing UAV to the guidance point through a preset path planning algorithm, and control the fixed-wing UAV to track the dynamic target.

[0051] It should be noted that this embodiment provides a multi-target tracking and planning system based on a fixed-wing UAV. By setting up a target filtering module 100, a target selection module 200, and a path planning module 300, the system enables the fixed-wing UAV to track dynamic multiple targets, further improving the calculation and tracking efficiency of multi-target tracking. This allows the invention to flexibly cope with complex and ever-changing flight environments. By updating target data and path planning in real time, the system ensures that the fixed-wing UAV can accurately track the target, improving the accuracy and efficiency of data processing and greatly enhancing the autonomy and intelligence level of the fixed-wing UAV in the multi-target tracking process.

[0052] In a preferred embodiment, this application also provides an electronic device, the electronic device comprising: The computer device includes a memory and a processor. The memory stores computer-readable instructions that, when executed by the processor, implement the multi-target tracking and planning method based on a fixed-wing UAV. This computer device can be broadly categorized as a server, terminal, or any other electronic device with the necessary computing and / or processing capabilities. In one embodiment, the computer device may include a processor, memory, network interface, communication interface, etc., connected via a system bus. The processor of the computer device can be used to provide the necessary computing, processing, and / or control capabilities. The memory of the computer device may include a non-volatile storage medium and internal memory. The non-volatile storage medium may store an operating system, computer programs, etc. The internal memory can provide an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface and communication interface of the computer device can be used to connect and communicate with external devices via a network. When the computer program is executed by the processor, it performs the steps of the method of the present invention.

[0053] It should be noted that this invention provides a multi-target tracking planning method and system based on a fixed-wing UAV. The method involves performing Kalman filtering on dynamic target data, outputting filtered dynamic target data, and then performing target filtering and hierarchical processing to obtain multiple target subsets. Hough transform is then used to extract spatial distribution features from each target subset, generating a target tracking sequence. Based on the relative position of the target tracking sequence and the fixed-wing UAV, a guidance point is dynamically calculated. A preset path planning algorithm generates the optimal flight path from the fixed-wing UAV to the guidance point, controlling the fixed-wing UAV to track dynamic targets. This achieves the tracking of multiple dynamic targets by the fixed-wing UAV, further improving the computational and tracking efficiency of multi-target tracking. This invention enables flexible responses to complex and changing flight environments. By updating target data and path planning in real time, it ensures that the fixed-wing UAV can accurately track targets, improving the accuracy and efficiency of data processing and significantly enhancing the autonomy and intelligence level of the fixed-wing UAV in the multi-target tracking process.

[0054] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0055] The technical features described above can be combined arbitrarily. Although not all possible combinations of these technical features are described, any combination of these technical features should be considered to be covered by this specification, provided that such combination does not contain contradictions.

[0056] The specific embodiments of the present invention described above do not constitute a limitation on the scope of protection of the present invention. Any other corresponding changes and modifications made in accordance with the technical concept of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A multi-target tracking planning method based on a fixed-wing unmanned aerial vehicle (UAV), characterized in that, include: S100: Acquire dynamic target data of a fixed-wing UAV, perform Kalman filtering on the dynamic target data, and output the filtered dynamic target data; S200. Based on the filtered dynamic target data, target screening and hierarchical processing are performed to obtain multiple target subsets. The spatial distribution features of each target subset are extracted sequentially using Hough transform to generate a target tracking sequence. S300. Based on the relative position of the target tracking sequence and the fixed-wing UAV, dynamically calculate the guidance point, generate the optimal flight path from the fixed-wing UAV to the guidance point through a preset path planning algorithm, and control the fixed-wing UAV to track the dynamic target.

2. The multi-target tracking planning method based on a fixed-wing UAV according to claim 1, characterized in that, The filtered dynamic target data is then subjected to target selection and hierarchical processing to obtain multiple target subsets, including: Based on the filtered dynamic target data, a preset target filtering mechanism is used to remove false alarm targets, and the target set after removing false alarm targets is output. The target set after removing false alarm targets is then processed into multiple target subsets according to altitude.

3. The multi-target tracking planning method based on a fixed-wing UAV according to claim 1, characterized in that, The step of extracting spatial distribution features from each target subset using Hough transform to generate a target tracking sequence includes: Transform the dynamic target point set in the target subset to Hough space; The first Hough region with the highest density in the Hough space was identified by cluster analysis, and the optimal fitted line of the first Hough region was analyzed. The dynamic targets in the target subset are projected onto the optimal fitted line, and a target tracking sequence is generated according to the order of the projection points. Based on the relative motion direction between the fixed-wing UAV and the target tracking sequence, the tracking mode of the fixed-wing UAV is determined.

4. The multi-target tracking planning method based on a fixed-wing UAV according to claim 3, characterized in that, The step of determining the tracking mode of the fixed-wing UAV based on the relative motion direction between the fixed-wing UAV and the target tracking sequence includes: If the fixed-wing UAV moves in the opposite direction to the target tracking sequence, then the tracking mode of the fixed-wing UAV is a head-on tracking sequence. If a fixed-wing UAV is in the same direction of motion as the target in the tracking sequence, then the tracking mode of the fixed-wing UAV is a tail-chase tracking sequence.

5. The multi-target tracking planning method based on a fixed-wing UAV according to claim 1, characterized in that, The step of dynamically calculating the guidance point based on the relative position of the target tracking sequence and the fixed-wing UAV includes: Calculate the relative distance between the center point of the target tracking sequence and the fixed-wing UAV; Determine whether the relative distance is greater than a first preset threshold, determine the guidance method of the fixed-wing UAV based on the determination result, and then calculate the guidance point based on the guidance method.

6. The multi-target tracking planning method based on a fixed-wing UAV according to claim 5, characterized in that, The process of determining the guidance method for the fixed-wing UAV based on the judgment result includes: When the relative distance is greater than the first preset threshold, the guidance mode of the fixed-wing UAV is a predictive guidance mode. When the relative distance is less than or equal to the first preset threshold, the guidance mode of the fixed-wing UAV is real-time guidance mode.

7. The multi-target tracking planning method based on a fixed-wing UAV according to claim 6, characterized in that, The calculation of the guidance point according to the guidance method includes: If the guidance method of the fixed-wing UAV is predictive guidance, then the meeting point between the fixed-wing UAV and the target tracking sequence is used as the guidance point. If the guidance method of the fixed-wing UAV is real-time guidance, then the spatial projection point of the current target tracking sequence is used as the guidance point.

8. The multi-target tracking planning method based on a fixed-wing UAV according to claim 6, characterized in that, The prediction guidance method includes: Based on the Kalman filter algorithm, the motion state of each target in the target tracking sequence is estimated, the motion state of the target tracking sequence is predicted at multiple future time points, and the target prediction trajectory is generated. Based on the current motion status data of the fixed-wing UAV, calculate the flight time required for the fixed-wing UAV to fly from its current position to any point on the predicted trajectory of the target. Based on the flight time, the coordinates of the preceding predicted points of the fixed-wing UAV on the target predicted trajectory are obtained by interpolation on the target predicted trajectory. The preceding predicted points are used as the guidance target points of the fixed-wing UAV. Least square straight line fitting is performed on multiple consecutive preceding predicted points to generate a predicted guidance straight line.

9. The multi-target tracking planning method based on a fixed-wing UAV according to claim 1, characterized in that, The step of generating the optimal flight path from the fixed-wing UAV to the guidance point using a preset path planning algorithm includes: The Dubins path planning method is adopted. Starting from the current position of the fixed-wing UAV and ending at the guidance point, the lengths of multiple Dubins paths are calculated based on the minimum turning radius of the fixed-wing UAV. The Dubins path set is output, and the shortest Dubins path in the set is taken as the optimal Dubins path for the current fixed-wing UAV.

10. A multi-target tracking and planning system based on a fixed-wing unmanned aerial vehicle, characterized in that, include: Target filtering module: used to acquire dynamic target data of fixed-wing UAV, perform Kalman filtering on the dynamic target data, and output filtered dynamic target data; Target filtering module: Used to filter and stratify targets based on the filtered dynamic target data to obtain multiple target subsets. The Hough transform is used to extract the spatial distribution features of each target subset in turn to generate a target tracking sequence. Path planning module: It is used to dynamically calculate the guidance point based on the relative position of the target tracking sequence and the fixed-wing UAV, generate the optimal flight path of the fixed-wing UAV to the guidance point through a preset path planning algorithm, and control the fixed-wing UAV to track the dynamic target.

Images