A method and system for trajectory planning

Abstract

The application relates to the technical field of flight path planning, and discloses a flight path planning method and system. In a flight path search space, a grid is used to represent the flight path search space, and a plurality of local windows are used to perform flight path planning in parallel. The application solves the problems that the prior art cannot efficiently and accurately complete a flight path planning task.

Description

Technical Field

[0001] This invention relates to the field of trajectory planning technology, specifically a trajectory planning method and system. Background Technology

[0002] Flight path planning is a core component of UAV mission planning. Its task, based on mission allocation and considering constraints such as flight capabilities and the flight environment, is to find the optimal feasible flight path for the UAV that best meets mission requirements, minimizing threats and obstacles. The quality of flight path planning determines whether the UAV can successfully complete its mission in complex and ever-changing environments while minimizing flight costs.

[0003] Ant colony optimization (ACO) has garnered significant attention in trajectory planning due to its advantages such as parallel processing, robustness, and strong global search capabilities. However, it also suffers from drawbacks, including susceptibility to local optima, slow convergence, and the dependence of parameter settings on the flight environment. High-efficiency and high-precision planning remains a persistent goal in trajectory planning. This invention leverages the robustness, parallelizability, and strong global search capabilities of ACO to address its limitations, such as planning accuracy restricting convergence speed and complex overall parameter adjustments. It proposes an improved trajectory planning method based on ACO. By proposing a multi-precision, multi-window planning method, based on the initial trajectory, it automatically configures the local planning window, planning precision, and planning parameters according to the characteristics of the local flight environment. This allows for parallel multi-precision, multi-window trajectory adjustments, enabling the optimization of a flight path adapted to the battlefield environment in a shorter time, thus improving the efficiency and planning accuracy of existing ACO trajectory planning algorithms. Summary of the Invention

[0004] To overcome the shortcomings of existing technologies, this invention provides a trajectory planning method and system, which solves the problems of existing technologies in completing trajectory planning tasks efficiently and with high precision.

[0005] The technical solution adopted by the present invention to solve the above problems is:

[0006] A trajectory planning method employs a grid to represent the trajectory search space and utilizes multiple local windows to perform trajectory planning in parallel.

[0007] As a preferred technical solution, it includes the following modules:

[0008] S1, Flight Environment Setup: Before track planning begins, the flight environment within the track search space is set up, and the track search space is characterized by a grid partitioning method.

[0009] S2, Initial trajectory planning: Set the initial trajectory planning parameters, the aircraft's starting point and ending point, optimize the trajectory, and obtain the initial trajectory;

[0010] S3, Parallel trajectory planning: Utilizes multiple local windows to perform trajectory planning in parallel, obtaining the flight trajectory within the local planning window;

[0011] S4, Judgment and Iteration: Determine whether the trajectory optimization meets the requirements: If not, jump to step S3 to continue the next round of parallel trajectory planning; if it meets the requirements, output the final flight trajectory.

[0012] As a preferred technical solution, step S1 includes the following steps:

[0013] S11, Grid Partitioning: The track search space is represented by a grid. A global planning window is set to represent the track search space as computer symbolic information for computer calculation and analysis.

[0014] As a preferred technical solution, step S1 includes the following steps:

[0015] S12, Space Threat Probability Calculation: Simulate the flight environment within the planning window, and calculate the probability of an aircraft being threatened when passing through each grid point based on the distribution of threats and / or obstacles.

[0016] As a preferred technical solution, in step S12, the specific method for calculating the probability of an aircraft being threatened when passing by each grid point is as follows:

[0017] If it is a radar threat, the threat probability passing through the grid points is:

[0018]

[0019] In the formula, d R d represents the distance of a grid point from the radar. R max d represents the maximum radius of the radar detection area. R min Indicates the radius of the area that is absolutely detectable by the radar;

[0020] If the threat is from air defense fire, the probability of a threat passing through a grid point is:

[0021]

[0022] d M d represents the horizontal distance between a grid point and a fire threat. M max For maximum lethal range, d M min This is the maximum range at which enemy firepower can inflict devastating damage.

[0023] If the threat is due to terrain, weather, or no-fly zones, the threat probability of passing through the grid points is:

[0024]

[0025] df max represents the maximum radius of the threat influence area, d f represents the distance of the grid point from the threat center.

[0026] As a preferred technical solution, in step S2, first set the initial planning accuracy Grid1, the initial planning window Map1, and the parameters of the trajectory optimization algorithm, then set the starting point and the ending point of the aircraft, and then call the ant colony optimization algorithm to optimize the trajectory to obtain the initial trajectory and the flight cost value.

[0027] As a preferred technical solution, the calculation method of the flight cost value is as follows:

[0028]

[0029] In the formula, F represents the flight cost; J1 represents the flight distance, which is obtained by summing the horizontal distances of the flight trajectory grid points; J2 represents the threat probability corresponding to the flight trajectory, which is obtained by summing the threat probabilities of the flight trajectory grid points; w1 and w2 are weight factor parameters, and w1 + w2 = 1, and the values are defined by the user; S represents the theoretically possible farthest flight distance, T represents the theoretically possible maximum trajectory threat probability, and both S and T are constant values.

[0030] As a preferred technical solution, step S3 includes the following steps:

[0031] S31, according to the existing flight trajectory X i and the initial planning window Map1, the user customizes the range of the local window to be divided; where, X i represents the flight trajectory obtained after the end of the i-th round of trajectory planning, i represents the current iteration number of the trajectory planning, and the initial trajectory is denoted as X1;

[0032] S32, after dividing into several local planning windows, set the grid accuracy of each local planning window to 1 / n of the local planning window size; where, 1 / n represents the ratio of the grid accuracy to the local planning window size, and n is given by the user according to preferences;

[0033] S33, according to the threat / obstacle distribution in the local planning window, set the weight factors w1 and w2 for each window: if the threats in the local window are sparsely distributed, then let w1 > w2; otherwise, w1 < w2, and the specific values of w1 and w2 are given by the user according to preferences; where, sparse distribution means that the area with a threat probability of 0 accounts for more than half;

[0034] S34, use the intersection point of the flight trajectory X i and the local window as the starting / ending point of the aircraft flight in the local planning window;

[0035] S35, multiple local windows call the ant colony optimization algorithm in step S2 in parallel, perform trajectory planning at the same time, and obtain the flight trajectory in the local planning window;

[0036] S36, finally the flight path X i The portion of the flight path corresponding to the local planning window is replaced with the local flight path to obtain the final aircraft flight path X. i+1 .

[0037] As a preferred technical solution, in step S4, it is determined whether the trajectory optimization meets the requirements. If not, then let i = i + 1 and jump to step S3 to continue the next round of parallel trajectory planning; if it meets the requirements, then the final flight trajectory X is output. i+1 .

[0038] A trajectory planning system for implementing the aforementioned trajectory planning method includes the following modules connected in sequence:

[0039] Flight Environment Setting Module: Used to set the flight environment within the trajectory search space before trajectory planning begins, and the trajectory search space is characterized by a grid division method.

[0040] Initial trajectory planning module: Used to set initial trajectory planning parameters, aircraft start and end points, optimize trajectory, and obtain initial trajectory;

[0041] Parallel trajectory planning module: used to perform trajectory planning in parallel using multiple local windows, and obtain the flight trajectory in the local planning window;

[0042] Judgment and Iteration Module: Used to determine whether the trajectory optimization meets the requirements: if not, jump to step S3 to continue the next round of parallel trajectory planning; if it meets the requirements, output the final flight trajectory.

[0043] Compared with the prior art, the present invention has the following advantages:

[0044] (1) This invention solves the problems of planning accuracy limiting convergence speed and flight environment limiting algorithm parameter settings in the existing technology, and can complete the trajectory planning task efficiently and with high accuracy.

[0045] (2) Based on the existing ant colony trajectory planning algorithm, a multi-precision multi-window parallel planning method is proposed. It can configure local planning windows, planning precision and weight parameters according to the characteristics of local flight environment, and carry out multi-precision multi-window trajectory adjustment in parallel. It can optimize a flight trajectory that adapts to the battlefield environment in a short time. Attached Figure Description

[0046] Figure 1 Example diagram of grid division for planning windows;

[0047] Figure 2 Here is a flowchart of the basic ant colony algorithm;

[0048] Figure 3 This is a flowchart illustrating the parallel trajectory planning method described in this invention.

[0049] Figure 4 This is a schematic diagram of the initial flight path in Example 2;

[0050] Figure 5 This is a schematic diagram of the parallel programming results in Example 2;

[0051] Figure 6 for Figure 5 A magnified view of a portion of the image. Detailed Implementation

[0052] The present invention will be further described in detail below with reference to the embodiments and accompanying drawings, but the embodiments of the present invention are not limited thereto.

[0053] Example 1

[0054] like Figures 1 to 6 As shown, a parallel trajectory planning method is applied to UAV trajectory planning, including the following steps:

[0055] S1, Flight Environment Settings:

[0056] It consists of two steps: grid generation and spatial threat probability calculation.

[0057] Grid Partitioning: Before trajectory planning begins, the trajectory search space needs to be represented as computer symbolic information to facilitate computer calculation and analysis. This invention uses a grid partitioning method to represent the search space, with the trajectory formed by connecting adjacent grid nodes. First, a global planning window is set, which limits the global scope of trajectory optimization. The planning window is represented by the coordinates of four vertices, and one planning window is a rectangle. The planning accuracy of each planning window is defined by the grid size; the smaller the area represented by a single grid, the higher the planning accuracy. Figure 1 Examples of planning space and grid division are given.

[0058] Spatial threat probability calculation: This involves simulating the flight environment within the planning window (i.e., the search space). Based on the threat / obstacle distribution, the probability of a UAV being threatened when passing through each grid point is calculated. The specific calculation method is as follows:

[0059] If it is a radar threat, the threat probability passing through the grid points is:

[0060]

[0061] In the formula, d RIndicates the distance of the grid points from the radar; d R max d represents the maximum radius of the radar detection area. R min This indicates the radius of the area that is absolutely detectable by radar.

[0062] If the threat is from air defense fire, the probability of a threat passing through a grid point is:

[0063]

[0064] d M d represents the horizontal distance between a grid point and a fire threat. M max For maximum lethal range, d M min This is the maximum range at which enemy firepower can inflict devastating damage.

[0065] If the threat is due to terrain, weather, or no-fly zones, the probability of a threat passing through a grid point is:

[0066]

[0067] d f max d represents the maximum radius of the area affected by the threat. f This indicates the distance of a grid point from the center of the threat.

[0068] S2, Initial trajectory planning:

[0069] Set the initial planning accuracy (Grid1), initial planning window (Map1), and trajectory optimization algorithm parameters. Set the UAV's start and end points, and call the existing ant colony optimization algorithm to optimize the trajectory, obtaining the initial trajectory (denoted as X). i Let i = 1, and the flight cost value. (See attached diagram) Figure 2 A flowchart of the existing basic ant colony algorithm is given. The method for calculating the flight cost is as follows:

[0070]

[0071] F represents the flight cost; J1 represents the flight range, obtained by summing the horizontal distances of the flight track grid points; J2 represents the threat probability corresponding to the flight track, obtained by summing the threat probabilities of the flight track grid points. w1 and w2 are weighting factor parameters, and w1 + w2 = 1, defined by the user. S represents the theoretically possible longest possible flight range, and T represents the theoretically possible maximum threat probability of the flight track. Both S and T are estimated by the user and are constant values.

[0072] S3, multi-precision multi-window parallel trajectory planning:

[0073] Based on existing flight path X iWith the initial planning window Map1, the user defines the range of the local window to be divided. After dividing into several local planning windows, the planning accuracy (i.e., grid size) of each local planning window is automatically set to 1 / n of the local window size, where n is given by the user according to preferences. According to the threat / obstacle distribution in the local planning window, the weight factors w1 and w2 under each window are set. If the threats in the local window are sparsely distributed, then w1 > w2; otherwise, w1 < w2, and the specific values are given by the user according to preferences. Taking the intersection of the flight track X i and the local window as the starting / ending point of the UAV flight in the local planning window. Multiple local windows parallelly call the basic ant colony optimization algorithm in step S2, perform trajectory planning simultaneously, and obtain the flight tracks in the local planning windows. Finally, replace the corresponding part in the flight track X i with the local flight tracks to obtain the final UAV flight track X i+1 .

[0074] S4. Judgment and iteration:

[0075] Judge whether the trajectory optimization meets the requirements. If not: let i = i + 1, and jump to S3 to continue the next round of multi-precision multi-window parallel planning; if so: output the final flight track X i+1 .

[0076] The present invention provides a trajectory planning method, which can solve the problems existing in the prior art, such as the planning accuracy restricting the convergence speed and the flight environment restricting the algorithm parameter setting, and can efficiently and highly accurately complete the trajectory planning task.

[0077] Compared with the prior art, the higher the planning accuracy, the slower the convergence. The present invention first performs a coarse-precision search and then adjusts the trajectory with high precision, which can improve the optimization speed.

[0078] Compared with the prior art, the parameter setting of the algorithm is related to the flight environment. If a single window is searched, it is difficult to obtain a satisfactory trajectory. The present invention adopts multi-window search, which can adjust the local planning parameters according to the local flight situation, so as to obtain a better and more satisfactory trajectory.

[0079] Preferably, this method can run a program on a readable storage medium to implement the multi-precision multi-window parallel planning algorithm and apply it to trajectory planning.

[0080] The present invention has the following beneficial effects compared with the prior art:

[0081] Based on existing ant colony trajectory planning algorithms, this invention proposes a multi-precision, multi-window parallel planning method that can configure local planning windows, planning precision, and weight parameters according to the characteristics of local flight environments, and carry out multi-precision, multi-window trajectory adjustments in parallel. It can optimize a flight trajectory that adapts to the battlefield environment in a short time.

[0082] Example 2

[0083] like Figures 1 to 6 As shown, as a further optimization of Embodiment 1, this embodiment also includes the following technical features based on Embodiment 1:

[0084] (1) Determine the UAV flight environment: Assume the UAV flight area is 600×600km 2 The threat distribution is shown in Table 1, with threat point coordinates and radius given randomly. Before planning begins, the UAV flight space needs to be gridded, and the intersections of the grids are the track nodes. Then, the threat locations and their radius of influence are represented in the spatial grid area. The relevant parameter settings for the ant colony algorithm used in this paper are given in Table 2. The UAV's starting point is [40, 40] km, and the target point is [560, 400] km.

[0085] Table 1 Threat Distribution Table

[0086]

[0087] Table 2 Ant Colony Algorithm Parameter Settings

[0088]

[0089] (2) Initial trajectory planning: The grid accuracy is set to 40km. The existing ant colony algorithm is used for optimization. Let w1 = w2 = 0.5. The initial flight path is as follows: Figure 4 The trajectory is shown as a solid circle.

[0090] (3) Four local planning windows are generated based on user preferences. The planning accuracy of each local planning window is 1 / 10 of its planning range. The optimization parameters of each planning window are adjusted according to the threat distribution. Parallel planning is performed using an existing ant colony algorithm. The planning results are shown in the appendix. Figures 5-6 As shown.

[0091] from Figure 4 As can be seen, if optimization is only sought within the initial planning window, the weights of flight costs can only be set uniformly. For regions with sparsely distributed threats (such as...), Figure 4 (Left half) The threat probability of the flight path is mostly 0, and the algorithm can find a path with the shortest distance. When the drone enters a threat-dense area (such as...) Figure 4(Right half) Because the weights of the threat probability cannot be adjusted in time, the algorithm can only find a neutral optimized path. This path has a shorter flight distance and avoids areas with greater threats to some extent, but its flight path does not meet the user's safety requirements. At this time, the user prefers to exchange a higher flight distance for a lower threat probability.

[0092] from Figures 4-6 As can be seen, the flight range increases after the trajectory adjustment. This increase primarily comes from areas with densely distributed threats. In these areas, through multi-precision, multi-window parallel planning, the algorithm assigns a larger weight to the threat cost, thus the optimization result tends to avoid high-threat areas at the expense of flight range. In threat-sparse areas, since a trajectory with zero threat can always be found, trajectory planning in these areas always aims to minimize flight range. The final adjusted flight trajectory shows that in threat-dense areas, the UAV always travels along the direction of least threat, meeting the user's need to minimize threats. Therefore, the method proposed in this invention configures different optimization parameters according to different local flight environments, thereby ensuring that the optimized overall trajectory meets user needs.

[0093] Figures 4-6 In the optimization shown, the highest track accuracy is approximately 5km, and the total optimization time is approximately 132 seconds. Using 5km accuracy as the planning accuracy and the parameter settings in Tables 1 and 2, a track planning operation without multi-precision, multi-window parallel processing was performed. After approximately 15 minutes of operation, no optimization results were output. This is because the initial planning window is large, and using a high planning accuracy within this window significantly increases the number of grid nodes, resulting in a huge search space and slowing down the convergence speed of the ant colony algorithm. This comparison shows that, with comparable accuracy, the method proposed in this invention can significantly improve the convergence speed.

[0094] As described above, the present invention can be implemented well.

[0095] All features disclosed in all embodiments of this specification, or steps in all methods or processes implied in the disclosure, may be combined and / or extended or replaced in any way, except for mutually exclusive features and / or steps.

[0096] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Based on the technical essence of the present invention, any simple modifications, equivalent substitutions, and improvements made to the above embodiments within the spirit and principles of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A trajectory planning method, characterized in that, Within the trajectory search space, a grid is used to represent the trajectory search space, and multiple local windows are used to perform trajectory planning in parallel. The aforementioned trajectory planning method includes the following modules: S1, Flight Environment Setup: Before track planning begins, the flight environment within the track search space is set up, and the track search space is characterized by a grid partitioning method. S2, Initial trajectory planning: Set the initial trajectory planning parameters, the aircraft's starting point and ending point, optimize the trajectory, and obtain the initial trajectory; S3, Parallel trajectory planning: Utilizes multiple local windows to perform trajectory planning in parallel, obtaining the flight trajectory within the local planning window; Step S3 includes the following steps: S31, based on existing flight paths and initial planning window Users can define the local window area to be divided; among them, Indicates the first The flight path obtained after the completion of the flight path planning. This represents the current iteration number of the trajectory planning; the initial trajectory is denoted as... ; S32, after dividing the area into several local planning windows, sets the grid precision of each local planning window to the size of the local planning window. ;in, This represents the ratio of grid accuracy to the size of the local planning window. Provided by the user based on their preferences; S33, Based on the distribution of threats / obstacles within the local planning window, set the weighting factor for each window. and If the threats in the local window are sparsely distributed, then let Conversely, , and The specific values ​​are given by the user according to their preferences; among them, sparse distribution means that more than half of the areas have a threat probability of 0. S34, with flight path The intersection with the local window serves as the start / end point of the aircraft's flight within the local planning window; S35, multiple local windows call the ant colony optimization algorithm in step S2 in parallel, perform trajectory planning at the same time, and obtain the flight trajectory in the local planning window; S36 will finally trace its flight path. The portion of the flight path corresponding to the local planning window is replaced with the local flight path to obtain the final aircraft flight path. ; S4, Judgment and Iteration: Determine whether the trajectory optimization meets the requirements: If not, jump to step S3 to continue the next round of parallel trajectory planning; if it meets the requirements, output the final flight trajectory.

2. The trajectory planning method according to claim 1, characterized in that, Step S1 includes the following steps: S11, Grid Partitioning: The track search space is represented by a grid. A global planning window is set to represent the track search space as computer symbolic information for computer calculation and analysis.

3. The trajectory planning method according to claim 2, characterized in that, Step S1 includes the following steps: S12, Space Threat Probability Calculation: Simulate the flight environment within the planning window, and calculate the probability of an aircraft being threatened when passing through each grid point based on the distribution of threats and / or obstacles.

4. The trajectory planning method according to claim 3, characterized in that, In step S12, the specific method for calculating the probability of an aircraft being threatened when passing through each grid point is as follows: If it is a radar threat, the threat probability passing through the grid points is: ； In the formula, This indicates the distance of each grid point from the radar. This indicates the maximum radius of the radar detection area. Indicates the radius of the area that is absolutely detectable by the radar; If the threat is from air defense fire, the probability of a threat passing through a grid point is: ； This indicates the horizontal distance between a grid point and a fire threat. For maximum lethal range, This is the maximum range at which enemy firepower can inflict devastating damage. If the threat is due to terrain, weather, or no-fly zones, the threat probability of passing through the grid points is: ； Indicates the maximum radius of the area affected by the threat. This indicates the distance of a grid point from the center of the threat.

5. A trajectory planning method according to any one of claims 1 to 4, characterized in that, In step S2, the initial planning accuracy is first set. Initial planning window First, the trajectory optimization algorithm parameters are set, then the aircraft's starting point and ending point are set, and then the ant colony optimization algorithm is called to optimize the trajectory, so as to obtain the initial trajectory and flight cost values.

6. The trajectory planning method according to claim 5, characterized in that, The calculation method for the flight cost is as follows: ； In the formula, Indicates the cost of flight; The flight distance is represented by the sum of the horizontal distances between the points on the flight track grid. This represents the threat probability corresponding to the flight path, which is obtained by summing the threat probabilities of the flight path grid points. and It is a weighting factor parameter, and The value is defined by the user. This indicates the theoretically longest possible flight distance. This represents the theoretically maximum probability of a flight path threatening a target. and All are constant values.

7. The trajectory planning method according to claim 6, characterized in that, In step S4, determine whether the trajectory optimization meets the requirements. If not, then let Proceed to step S3 to continue the next round of parallel trajectory planning; if satisfied, output the final flight trajectory. .

8. A trajectory planning system, characterized in that, A method for implementing a trajectory planning method according to any one of claims 1 to 7 includes the following modules connected in sequence: Flight Environment Setting Module: Used to set the flight environment within the trajectory search space before trajectory planning begins, and the trajectory search space is characterized by a grid division method. Initial trajectory planning module: Used to set initial trajectory planning parameters, aircraft start and end points, optimize trajectory, and obtain initial trajectory; Parallel trajectory planning module: used to perform trajectory planning in parallel using multiple local windows, and obtain the flight trajectory in the local planning window; Judgment and Iteration Module: Used to determine whether the trajectory optimization meets the requirements: if not, jump to step S3 to continue the next round of parallel trajectory planning; if it meets the requirements, output the final flight trajectory.

Images