An unmanned aerial vehicle dynamic path planning method based on hybrid intelligent optimization

Abstract

The application relates to a kind of unmanned aerial vehicle dynamic path planning methods based on hybrid intelligent optimization, introduce multi-stage hybrid intelligent optimization algorithm, on the basis of static path planning based on particle swarm optimization, consider the dynamic target condition of real-time change, when unmanned aerial vehicle gradually approaches target and needs to carry out dynamic path adjustment, utilize A* algorithm mechanism to give more accurate tracking path planning scheme, so that unmanned aerial vehicle can dynamically adjust path, effectively coordinate and complete task in overall task execution. The application comprehensively considers various complex multi-constraint conditions and coordination relationship, solves the multi-unmanned aerial vehicle dynamic target tracking path planning problem, so that unmanned aerial vehicle can quickly respond when finding out the latest position of target, dynamically and real-timely adjust the track of unmanned aerial vehicle.

Description

Technical Field

[0001] This invention belongs to the field of dynamic path planning technology for unmanned aerial vehicles (UAVs), and relates to a dynamic path planning method for UAVs based on hybrid intelligent optimization. Background Technology

[0002] Dynamic target tracking path planning for multiple unmanned aerial vehicles (UAVs) is a crucial decision-making problem in swarm collaborative mission planning. It involves planning a set of safe and reliable cooperative flight paths for multiple UAVs participating in a tracking mission, taking into account various complex constraints and collaborative relationships, to ensure the successful execution of all tracking tasks. Dynamic target tracking path planning for multiple UAVs requires consideration not only of the UAVs' own flight constraints, such as maximum climb angle, maximum turning radius, and step length, but also of spatial obstacle constraints and collaborative constraints, to find the shortest path to the target area while avoiding obstacles and threats.

[0003] Currently, commonly used methods for dynamic target tracking and path planning for multiple unmanned aerial vehicles mainly include probabilistic graphical algorithms, A* algorithms, artificial potential field methods, and path collaborative planning algorithms based on swarm intelligence.

[0004] Using probabilistic graphs to simplify spatial representation is a common path planning method. It reduces the space to be planned into a simple and safe road network based on system constraints and threat avoidance conditions, and then obtains the optimal flight route from this network. Commonly used probabilistic graphical algorithms include Voronoi diagrams and pass-through diagrams. While these algorithms are easy to construct in a two-dimensional plane, they are difficult to build in three-dimensional space.

[0005] The A* algorithm is a direct search method based on static road networks. It can handle complex constraints in the environment and for unmanned aerial vehicles (UAVs). Initially used to solve single- or multi-UAV cooperative path planning problems in two-dimensional environments, researchers have extended the sparse A* method to three-dimensional environments, enabling path planning in 3D environments. Further research has focused on improving the algorithm by combining terrain smoothing techniques and variable step sizes. Additionally, studies have further refined the A* algorithm to address multi-UAV cooperative extended problems. However, this type of algorithm involves high computational cost per node, and its search efficiency decreases as the number of nodes increases. Furthermore, the search efficiency is susceptible to the influence of the evaluation function, and the obtained result may not be the optimal solution.

[0006] Artificial potential field methods, proposed by Khatib, are also effective path planning methods for unmanned aerial vehicles (UAVs). This method mimics the influence of attractive and repulsive forces on the UAV's motion in the planning space, guiding the UAV along the direction of the steepest descent of the potential function. By moving along the direction of the resultant force, the UAV can avoid obstacles and threats and reach the target location. For example, McInnes used artificial potential field methods to plan the short-range flight path of a space servicing spacecraft near the space station. A drawback of this type of algorithm is that the existence of a rest point where repulsive and attractive forces are equal may cause the algorithm to stall, leading to planning failure.

[0007] Swarm intelligence algorithms are also suitable for solving cooperative path planning problems involving multiple unmanned aerial vehicles (UAVs). Nikolos was among the first to use genetic algorithms and differential evolution programming (DEP) to calculate flight paths for single and multi-UAV cooperatives. Subsequently, much research on path planning has focused on improved evolutionary algorithms. Given that particle swarm optimization (PSO) is more flexible and better suited for path planning than genetic algorithms, Foo J.L. et al. used PSO to achieve cooperative path planning for multiple UAVs and conducted experiments on a 3D virtual simulation platform. In recent years, research using PSO for UAV path planning has gradually increased. Roberge V et al. compared parallel evolutionary algorithms with particle swarm optimization, emphasizing their differences and respective advantages, and combined these two methods for practical cooperative path planning. Furthermore, research using intelligent algorithms such as ant colony optimization, artificial immune algorithms, and neural networks for multi-UAV cooperative path planning is also growing. However, these algorithms suffer from drawbacks such as slow optimization speed, non-unique optimal solutions, potential conflicts between multiple UAVs, and constraint violations. Summary of the Invention

[0008] The technical problem solved by this invention is to overcome the shortcomings of the prior art and propose a dynamic path planning method for unmanned aerial vehicles based on hybrid intelligent optimization. This method comprehensively considers various complex multi-constraint conditions and cooperative relationships to solve the cooperative path planning problem for multiple unmanned aerial vehicles performing dynamic target tracking tasks.

[0009] The solution of this invention is: a dynamic path planning method for unmanned aerial vehicles based on hybrid intelligent optimization, comprising the following steps:

[0010] Step 1: Establish the mission environment. After receiving the target tracking mission, the UAV clarifies the mission allocation and target location information, and starts from the starting position.

[0011] Step 2: Based on the initial position of the dynamic target, use the particle swarm optimization mechanism to perform initial static path planning, and fly towards the target along the optimized path;

[0012] Step 3: During flight, the target changes dynamically. The UAV adopts a phased target deviation judgment strategy: the target position is updated every time interval Δt during flight, and the degree of deviation between the UAV and the endpoint of the current planned path is judged.

[0013] The criteria for determining whether a target has seriously deviated are as follows: when the distance between the UAV and the original target position is greater than a set threshold L, if the angle between the original target position, the position of the UAV, and the new target position is greater than the first set threshold α1, then the target is considered to have seriously deviated; otherwise, the target is considered to have a low degree of deviation. When the distance between the UAV and the original target position is less than a set threshold L, if the angle between the original target position, the position of the UAV, and the new target position is greater than the second set threshold α2, then the target is considered to have seriously deviated; otherwise, the target is considered to have a low degree of deviation, where α1>α2.

[0014] Step 4: If the target deviation is determined to be serious, the path is adjusted using a dynamic path planning mechanism based on the A* algorithm, and an early warning mechanism is used when replanning the path; otherwise, the flight continues to fly along the current planned original trajectory Δt to the next time point.

[0015] The early warning mechanism is as follows: the path of the UAV within time Δt is defined as the inertial distance; a new path for the UAV after time Δt is planned, and the UAV continues to fly for time Δt before its trajectory is adjusted to follow the newly planned path; wherein, the newly planned path starts from the position of the UAV after flying the inertial distance along the original trajectory and ends at the position of the target, and the path is planned according to the A* algorithm.

[0016] Step 5: Determine whether the UAV has reached the final target point. If yes, the target tracking mission ends; otherwise, repeat steps 3 to 5.

[0017] Furthermore, step 1, establishing the task environment, includes:

[0018] Set up a three-dimensional mission terrain environment for dynamic path planning of multiple unmanned aerial vehicles;

[0019] Set the three-dimensional starting coordinates of N unmanned aerial vehicles;

[0020] Set N three-dimensional dynamic targets, and set the initial position and final position of each dynamic target;

[0021] Assign different priorities to N unmanned aerial vehicles based on mission time or importance.

[0022] Furthermore, when each path of N unmanned aerial vehicles consists of m+1m+1 path points and m path segments, the set of path points is:

[0023] ,

[0024] i = 1, ..., N, where This represents the coordinates of the j-th track point in three-dimensional space of the i-th unmanned aerial vehicle's trajectory;

[0025] When using particle swarm optimization and the A* algorithm for path planning, the total path length of each segment of the unmanned aerial vehicle during its flight is used as the basis. As a cost of path optimization, a dynamic path planning model for multiple unmanned aerial vehicles is established:

[0026]

[0027] in, This represents the length of the track segment between the i-th waypoint and the (i+1)-th waypoint of the k-th unmanned aerial vehicle;

[0028] Simultaneously, constraints are established for unmanned aerial vehicle path planning, including spatial obstacle avoidance constraints, collision avoidance constraints, minimum track segment constraints, maximum range constraints, minimum flight altitude constraints, maximum yaw angle constraints, and maximum climb angle constraints.

[0029] Furthermore, the spatial obstacle avoidance constraint is as follows:

[0030]

[0031] in, Let s be the distance between the j-th waypoint of the k-th UAV and the obstacle. min Preset safe distance between unmanned aerial vehicles and obstacles.

[0032] Furthermore, the anti-collision constraint is as follows:

[0033]

[0034] Among them, (x ks ,y ks ,z ks () represents the intersection of the flight paths of the k-th UAV and the s-th UAV. This indicates that the k-th unmanned aerial vehicle passes through a space with coordinates (x, y, y). ks ,y ks ,z ks At that moment, This indicates that the s-th unmanned aerial vehicle passes through a space with coordinates (x, y, y). ks ,y ks ,z ks At point t min Preset safe distances and flight times between unmanned aerial vehicles.

[0035] Furthermore, the minimum track segment constraint is as follows:

[0036]

[0037] Among them, l min This is the minimum distance an unmanned aerial vehicle must travel in its current direction before changing its current flight direction.

[0038] Furthermore, the maximum range constraint is a performance constraint of the unmanned aerial vehicle itself, in the form of:

[0039]

[0040] Among them, l max This represents the farthest distance the unmanned aerial vehicle can fly under the current load.

[0041] Furthermore, the minimum flight altitude constraint is as follows:

[0042]

[0043] in, Let h be the flight altitude of the k-th unmanned aerial vehicle on the i-th segment of its trajectory. min It is the minimum altitude at which unmanned aerial vehicles can fly safely.

[0044] Furthermore, the maximum yaw angle constraint is:

[0045]

[0046] in, Let θ be the vector of the i-th segment of the trajectory of the k-th unmanned aerial vehicle, and let θ be the preset maximum yaw angle determined by the maneuverability of the unmanned aerial vehicle itself.

[0047] Furthermore, the maximum climb angle constraint is:

[0048]

[0049] in, Let be the vector of the ascending trajectory of the k-th unmanned aerial vehicle in the i-th segment. for The projection vector on the horizontal plane, δ, is the preset maximum climb angle determined by the maneuverability of the unmanned aerial vehicle itself.

[0050] This invention introduces a multi-stage hybrid intelligent optimization algorithm to solve the problem of dynamic target tracking path planning for multiple unmanned aerial vehicles (UAVs). Based on static path planning using particle swarm optimization, it considers the dynamic target situation that changes in real time. When the UAV gradually approaches the target and needs to make dynamic path adjustments, it uses the A* algorithm mechanism to provide a more accurate tracking path planning scheme, enabling the UAV to dynamically adjust its path, effectively coordinate and complete the task in the overall mission execution.

[0051] The advantages of this invention compared to the prior art are:

[0052] (1) The present invention adopts a multi-stage hybrid intelligent optimization framework to meet different path planning requirements under different target tracking conditions, and balances the effectiveness and efficiency of path planning well.

[0053] (2) When performing global path planning in the initial stage, this invention adopts a particle swarm optimization algorithm based on swarm intelligence for static path planning. By flexibly selecting waypoints to adjust the accuracy of path planning, and by using a search mechanism that combines group information with individual information, and adjusting the performance and efficiency of the iterative termination condition to balance the algorithm, it can effectively balance global search and local refinement, and avoid the optimization complexity from exponentially increasing with the size and distance of the map.

[0054] (3) When approaching the target and needing to make dynamic path adjustments, the A* algorithm is used to provide a more accurate and detailed planning scheme by setting a higher movement accuracy. This avoids the problems of long time consumption and difficulty in applying the swarm intelligence optimization algorithm when making multiple dynamic path adjustments. This allows the unmanned aerial vehicle to quickly and in real time adjust its flight trajectory and achieve dynamic tracking of the target.

[0055] (4) The present invention considers the cooperative relationship between unmanned aerial vehicles (UAVs). Based on the mission time or importance, different priorities can be assigned to UAVs, and the flight paths of UAVs with higher priority levels are considered in the path planning in turn, so that a certain safe distance is maintained between UAVs performing the tracking mission or between UAVs and obstacles, thereby ensuring flight safety. Attached Figure Description

[0056] Figure 1 This is a flowchart of the dynamic path planning process for an unmanned aerial vehicle according to an embodiment of the present invention;

[0057] Figure 2 This is a simulation case diagram of dynamic path planning for multiple unmanned aerial vehicles in an embodiment of the present invention;

[0058] Figure 3 This is a flowchart of the dynamic path planning mechanism for unmanned aerial vehicles based on the A* algorithm in an embodiment of the present invention. Detailed Implementation

[0059] The invention will now be further described with reference to the accompanying drawings.

[0060] This invention proposes a dynamic path planning method for unmanned aerial vehicles (UAVs) based on hybrid intelligent optimization, which is used to solve the path planning problem of collaborative tracking of dynamic targets by multiple UAVs. The main process includes: building an algorithm framework, establishing a task environment, proposing a problem model, designing a static search mechanism, and designing a dynamic search mechanism.

[0061] 1. Establish the algorithm framework

[0062] like Figure 1 This invention addresses the dynamic path planning problem for multiple unmanned aerial vehicles (UAVs) by introducing a multi-stage hybrid intelligent optimization algorithm. Building upon static path planning based on particle swarm optimization, it considers real-time environmental dynamics, enabling UAVs to dynamically adjust their paths, effectively coordinate, and complete tasks. The specific dynamic path planning task flow includes:

[0063] Step 1: Establish the mission environment. After receiving the target tracking mission, the UAV clarifies the mission allocation and target location information, and starts from the starting position.

[0064] Step 2: Based on the initial position of the dynamic target, use the particle swarm optimization mechanism to perform initial static path planning, and fly towards the target along the optimized path.

[0065] Step 3: During flight, the target changes dynamically. The UAV adopts a phased target deviation judgment strategy: during flight, the target position is updated once every time interval Δt, and the degree of deviation between the UAV and the endpoint of the current planned path is judged.

[0066] The method for determining whether a target has deviated significantly is as follows: when the distance between the UAV and the original target position is greater than a set threshold L, if the angle between the original target position, the position of the UAV, and the new target position is greater than 60°, then the target is considered to have deviated significantly; otherwise, the target is considered to have deviated slightly. When the distance between the UAV and the original target position is less than the set threshold L, if the angle between the original target position, the position of the UAV, and the new target position is greater than 10°, then the target is considered to have deviated significantly; otherwise, the target is considered to have deviated slightly.

[0067] Step 4: If the target deviation is determined to be serious, the path is adjusted using the dynamic path planning mechanism based on the A* algorithm, and an early warning mechanism is used when replanning the path; otherwise, the flight continues to fly Δt according to the current planned original trajectory to the next time point.

[0068] The early warning mechanism is as follows: the path of the UAV within time Δt is defined as the inertial distance; a new path for the UAV after time Δt is planned; and the UAV continues to fly for time Δt before its trajectory is adjusted to follow the newly planned path. The newly planned path starts at the position of the UAV after flying the inertial distance along the original trajectory and ends at the position of the target, and the path is planned according to the A* algorithm.

[0069] Step 5: Determine whether the UAV has reached the final target point. If yes, the target tracking mission ends; otherwise, repeat steps 3 to 5.

[0070] 2. Establish the task environment

[0071] like Figure 2 As shown, a dynamic path planning mission execution environment for multiple unmanned aerial vehicles is established. In this implementation case, a mission terrain environment of 100km*100km*100m is set. Consider the scenario where four unmanned aerial vehicles (UAVs) fly from the starting point to the destination during a mission. The starting coordinates of the four UAVs are set as (9km, 2km, 8m), (6km, 1km, 8m), (5km, 13km, 8m), and (2km, 21km, 5m), respectively. There are four dynamic targets: target A moves from coordinates (95km, 83km, 10m) to (70km, 91km, 10m); target B moves from coordinates (77km, 61km, 9m) to (90km, 86km, 9m); target C moves from coordinates (39km, 80km, 10m) to (53km, 91km, 13m); and target D moves from coordinates (61km, 92km, 5m) to (72km, 84km, 6m). The mission execution priority decreases sequentially.

[0072] In this implementation case, a certain safe distance must be maintained between unmanned aerial vehicles (UAVs) or between UAVs and obstacles during tracking tasks to ensure flight safety. Every Δt time interval, or when the planned path is about to reach its destination, the UAV needs to update the target's latest position and determine whether a new path needs to be planned. In the diagram, □ represents the UAV's starting position, ☆ represents the target's starting position, and the smaller ★ and its connecting lines represent the movement path of the dynamic target.

[0073] 3. Propose a problem model

[0074] The flight path of an unmanned aerial vehicle (UAV) consists of several waypoints, with adjacent waypoints forming a track segment. Assume there are N UAVs in the cluster, and each track consists of m+1 waypoints and m track segments:

[0075] ,

[0076] i = 1, ..., N, where This represents the coordinates of the j-th node in the trajectory of the i-th unmanned aerial vehicle in three-dimensional space.

[0077] When N unmanned aerial vehicles (UAVs) are collaboratively tracking multiple targets, the path length determines the timeliness of completing the tracking task and is an important indicator for evaluating task performance. Therefore, the total path length of each segment of the UAV's flight path can be used as a reference. As a key factor in path optimization, among which... i = 1, 2, ..., m represents the length of the track segment between the i-th waypoint and the (i+1)-th waypoint of the k-th UAV. Meanwhile, UAV path planning needs to consider various constraints, such as: spatial position constraints (obstacle avoidance constraints, collision avoidance constraints); and the UAV's own performance constraints (minimum track segment constraints, maximum range constraints, minimum flight altitude constraints, maximum yaw angle constraints, maximum climb angle constraints, etc.).

[0078] When using particle swarm optimization and the A* algorithm for path planning, a dynamic path planning model for multiple unmanned aerial vehicles is established as follows:

[0079]

[0080] Make

[0081]

[0082] in, This refers to spatial obstacle avoidance constraints, which means that unmanned aerial vehicles (UAVs) need to avoid collisions with obstacles or targets during flight. Let s be the distance between the j-th waypoint of the k-th UAV and the obstacle. min Preset safe distance between unmanned aerial vehicles and obstacles.

[0083] This refers to collision avoidance constraints, meaning that at every moment during a multi-target tracking mission, the unmanned aerial vehicle (UAV) must avoid colliding with other UAVs and maintain a safe distance. (x) ks ,y ks ,z ks () represents the intersection of the flight paths of the k-th UAV and the s-th UAV. This indicates that the k-th unmanned aerial vehicle passes through a space with coordinates (x, y, y). ks ,y ks ,z ks At that moment, This indicates that the s-th unmanned aerial vehicle passes through a space with coordinates (x, y, y). ks ,y ks ,z ks At point tmin Preset safe distances and flight times between unmanned aerial vehicles.

[0084] Indicates the minimum path segment constraint, l min This is the minimum distance an unmanned aerial vehicle must travel in its current direction before changing its current flight direction.

[0085] This represents the maximum range constraint, which is a performance constraint inherent to the unmanned aerial vehicle (UAV). max This represents the farthest distance that can be flown under the current load.

[0086] This indicates a minimum flight altitude constraint, the purpose of which is to prevent unmanned aerial vehicles from crashing to the ground due to excessively low flight altitude. Let h be the flight altitude of the k-th unmanned aerial vehicle on the i-th segment of its trajectory. min It is the minimum altitude at which unmanned aerial vehicles can fly safely.

[0087]

[0088] This indicates the maximum yaw angle constraint, which means that due to the constraints of the UAV's own maneuverability, the UAV can only turn within a certain yaw angle range. Let θ be the vector of the i-th segment of the trajectory of the k-th unmanned aerial vehicle, and let θ be the preset maximum yaw angle determined by the maneuverability of the unmanned aerial vehicle itself.

[0089]

[0090] This indicates the maximum climb angle constraint, meaning that due to the limitations of the UAV's own maneuverability, the UAV can only fly within a certain climb angle range. Let be the vector of the ascending trajectory of the k-th unmanned aerial vehicle in the i-th segment. for The projection vector on the horizontal plane, δ, is the preset maximum climb angle determined by the maneuverability of the unmanned aerial vehicle itself.

[0091] 4. Static path planning mechanism

[0092] The particle swarm optimization (PSO) algorithm is used for initial static path planning. During the planning process, a heuristic initialization method based on average interpolation and a random initialization method are employed to establish an initial "particle" population. Each "particle" is initially assigned a random velocity, and then its position and velocity are continuously adjusted based on its individual experience (i.e., the best value in its historical records) and the population experience (i.e., the global best value of the entire population) until a final satisfactory solution is found.

[0093] Each particle in this process is defined by two d-dimensional vectors, representing its position and velocity. The position of the i-th particle in the population is denoted as X. i =(x i1 ,x i2 ,…,x id The speed is denoted as V. i =(v i1 ,v i2 ,…,v id Each particle i records its best position in its search history up to the current time t, denoted as P. i =(p i1 ,p i2 ,…,p id By comparing the historical best values ​​of all particles in the population, the global best position of the entire population up to the current time t can be obtained, denoted as P. g =(p g1 ,p g2 ,…,p gd Each particle updates its velocity based on its previous position, the distance between its current position and its historical best position, and the distance between its current position and the global best position. The specific calculation formula is as follows:

[0094] V i (t+1)=wV i (t)+c1r1(P i -X i (t))+c2r2(P g -X i (t))

[0095] Where w represents the inertia factor; c1 and c2 are constants representing acceleration coefficients; r1 and r2 are two independent random numbers drawn from a uniform distribution U[0,1]. Typically, the velocity V... i All components are limited to the range [-v] max ,v max To prevent particles from moving drastically outside the search space, the particles will move based on their updated velocities, and their updated positions will be:

[0096] X i (t+1)=X i (t)+V i (t+1)

[0097] The above process will continue to iterate until the termination condition is met.

[0098] 5. Dynamic path planning mechanism

[0099] When the unmanned aerial vehicle (UAV) gradually approaches the target and needs to make dynamic path adjustments, the A* algorithm is used for dynamic path planning. The process is as follows: Figure 3 As shown:

[0100] S1. At the start of the algorithm, clear openlist and closelist.

[0101] S2. Assign the starting position information to indexStart.

[0102] S3. Determine if indexStart is the endpoint. If indexStart is not equal to the endpoint, store indexStart in closelist, check candidate nodes in turn, update openlist, and proceed to S4; otherwise, output the path and the algorithm ends.

[0103] S4. Select the node with the lowest cost in the openlist as indexStart, and repeat steps S3-S4. In this embodiment, multiple constraints are used to select feasible nodes, and the advantage index and threat index are used as heuristic factors to design cost functions G(n) and H(n). Feasible nodes are selected by comprehensively considering the minimum flight path constraint, minimum flight altitude constraint, maximum yaw angle constraint, and maximum climb angle constraint; the advantage index is based on path length, and the threat index is based on spatial obstacle avoidance constraints and collision avoidance constraints, using Euclidean distance from obstacles; the cost function G(n) is the cost from the starting point to this candidate node, expressed as the sum of the advantage index and the threat index, and the cost function H(n) is the cost from the ending point to this candidate node, expressed as Euclidean distance.

[0104] Calculate the overall priority of each feasible node, and select the node with the highest overall priority (smallest F(n) value) as the next node to be traversed:

[0105] F(n) = G(n) + H(n)

[0106] During each dynamic path adjustment of the unmanned aerial vehicle (UAV), the trajectories of other UAVs are fixed. The cost of each candidate node in the space is calculated, including the cost from the starting point to the node and the cost from the node to the destination. The cost of passing through the node is calculated to determine whether the node is passable. The speed and accuracy of the algorithm are controlled by adjusting the heuristic function.

[0107] This invention employs a multi-stage dynamic trajectory planning method for unmanned aerial vehicles (UAVs) based on particle swarm optimization and the A* algorithm, which can meet the requirements of real-time planning and UAV flight safety. Throughout the flight process, UAVs can avoid collisions with each other and evade terrain obstacles while finding their own relatively optimal paths, ultimately tracking their respective dynamic targets.

[0108] The particle swarm optimization mechanism continuously adjusts the position and velocity of "particles" by combining group experience with individual experience until a final satisfactory solution is found, effectively balancing global search and local refinement. The dynamic path planning mechanism based on the A* algorithm meets the needs of real-time changes in target position in real tracking tasks, enabling the UAV to react quickly when it detects the latest target position and dynamically and in real time adjust the UAV's trajectory to achieve dynamic target tracking.

[0109] Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make possible changes and modifications to the technical solutions of the present invention by utilizing the methods and techniques disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall fall within the protection scope of the technical solutions of the present invention.

Claims

1. A dynamic path planning method for unmanned aerial vehicles based on hybrid intelligent optimization, characterized in that... Includes the following steps: Step 1: Establish the mission environment. After receiving the target tracking mission, the UAV clarifies the mission allocation and target location information, and starts from the starting position. Step 2: Based on the initial position of the dynamic target, use the particle swarm optimization mechanism to perform initial static path planning, and fly towards the target along the optimized path; Step 3: During flight, the target changes dynamically. The UAV adopts a phased target deviation judgment strategy: the target position is updated every time interval Δt during flight, and the degree of deviation between the UAV and the endpoint of the current planned path is judged. The criterion for determining whether the target has deviated significantly is as follows: when the distance between the UAV and the original target position is greater than a set threshold L, if the angle between the original target position, the position of the UAV, and the new target position is greater than a first set threshold... Under certain conditions, the target is judged to be severely deviated; otherwise, the target deviation is judged to be low. When the distance between the UAV and the original target position is less than a set threshold L, if the angle between the original target position, the UAV position, and the new target position is greater than a second set threshold... If the deviation is significant, the target is considered to be severely deviated from; otherwise, the deviation is considered to be minor. ; Step 4: If the target deviation is determined to be serious, the path is adjusted using a dynamic path planning mechanism based on the A* algorithm, and an early warning mechanism is used when replanning the path; otherwise, the flight continues to fly along the current planned original trajectory Δt to the next time point. The early warning mechanism is as follows: the path of the UAV within time Δt is defined as the inertial distance; a new path for the UAV after time Δt is planned, and the UAV continues to fly for time Δt before its trajectory is adjusted to follow the newly planned path; wherein, the newly planned path starts from the position of the UAV after flying the inertial distance along the original trajectory and ends at the position of the target, and the path is planned according to the A* algorithm. Step 5: Determine whether the UAV has reached the final target point. If yes, the target tracking task ends; otherwise, repeat steps 3 to 5. When each path of N unmanned aerial vehicles consists of m+1 path points and m path segments, the set of path points is: , ,in Indicates the first The first unmanned aerial vehicle flight path The coordinates of each waypoint in three-dimensional space; When using particle swarm optimization and the A* algorithm for path planning, the total path length of each segment of the unmanned aerial vehicle during its flight is used as the basis. As a cost of path optimization, a dynamic path planning model for multiple unmanned aerial vehicles is established: in, , indicating the first The first unmanned aerial vehicle The first waypoint to the second Length of track segments between waypoints; Simultaneously, constraints are established for unmanned aerial vehicle path planning, including spatial obstacle avoidance constraints, collision avoidance constraints, minimum track segment constraints, maximum range constraints, minimum flight altitude constraints, maximum yaw angle constraints, and maximum climb angle constraints.

2. The dynamic path planning method for unmanned aerial vehicles based on hybrid intelligent optimization according to claim 1, characterized in that, Step 1, establishing the task environment, includes: Set up a three-dimensional mission terrain environment for dynamic path planning of multiple unmanned aerial vehicles; Set the three-dimensional starting coordinates of N unmanned aerial vehicles; Set N three-dimensional dynamic targets, and set the initial position and final position of each dynamic target; Assign different priorities to N unmanned aerial vehicles based on mission time or importance.

3. The dynamic path planning method for unmanned aerial vehicles based on hybrid intelligent optimization according to claim 1, characterized in that, The spatial obstacle avoidance constraint is: in, For the first The first unmanned aerial vehicle The distance between each waypoint and the obstacle Preset safe distance between unmanned aerial vehicles and obstacles.

4. The dynamic path planning method for unmanned aerial vehicles based on hybrid intelligent optimization according to claim 1, characterized in that, The anti-collision constraint is: in, Indicates the first Unmanned aerial vehicles and the first The intersection of the flight paths of the unmanned aerial vehicles. Indicates the first The unmanned aerial vehicle passed through spatial coordinates as At the moment, Indicates the first The unmanned aerial vehicle passed through spatial coordinates as At the moment, Preset safe distances and flight times between unmanned aerial vehicles.

5. The dynamic path planning method for unmanned aerial vehicles based on hybrid intelligent optimization according to claim 1, characterized in that, The minimum track segment constraint is: in, This is the minimum distance an unmanned aerial vehicle must travel in its current direction before changing its current flight direction.

6. The dynamic path planning method for unmanned aerial vehicles based on hybrid intelligent optimization according to claim 1, characterized in that, The maximum range constraint is a performance constraint of the unmanned aerial vehicle itself, and takes the form of: in, This represents the farthest distance the unmanned aerial vehicle can fly under the current load.

7. The dynamic path planning method for unmanned aerial vehicles based on hybrid intelligent optimization according to claim 1, characterized in that, The minimum flight altitude constraint is as follows: in, For the first The first unmanned aerial vehicle The flight altitude of the segment's flight path It is the minimum altitude at which unmanned aerial vehicles can fly safely.

8. The dynamic path planning method for unmanned aerial vehicles based on hybrid intelligent optimization according to claim 1, characterized in that, The maximum yaw angle constraint is: in, For the first The first unmanned aerial vehicle The vector of the segment trajectory, The preset maximum yaw angle is determined by the maneuverability of the unmanned aerial vehicle itself.

9. The dynamic path planning method for unmanned aerial vehicles based on hybrid intelligent optimization according to claim 1, characterized in that, The maximum climb angle constraint is: in, For the first The first unmanned aerial vehicle The vector of the ascending trajectory. for The projection vector on the horizontal plane, The preset maximum climb angle is determined by the maneuverability of the unmanned aerial vehicle itself.

Images