Space-time joint trajectory generation method and system for fixed-wing unmanned aerial vehicle

Abstract

The invention discloses a space-time joint trajectory generation method and system for a fixed-wing unmanned aerial vehicle, and is used for solving the space-time joint trajectory planning problem of the fixed-wing unmanned aerial vehicle in a complex environment. Comprising the steps that an airborne computer of the fixed-wing unmanned aerial vehicle acquires a three-dimensional space map of a flight area, dense obstacle information, a starting position point, a target position point and physical parameters of the unmanned aerial vehicle; calculating a path from a starting position point to a target position point by adopting a fast exploration random algorithm RRT, and processing the path to generate a Dubins track; deformation processing is carried out on the Dubins trajectory based on a MINCO trajectory generation framework, and a space-time joint optimization trajectory is obtained through calculation by using an adaptive covariance matrix improved grey wolf optimization algorithm CMAGWO. The flight path of the fixed-wing unmanned aerial vehicle generated by adopting the method is continuous, the flight stability of the fixed-wing unmanned aerial vehicle can be improved, and unstable vibration of a fuselage caused by singular point positions of the Dubins path is avoided.

Description

Technical Field

[0001] The present invention belongs to the field of real-time online trajectory planning for intelligent robots, and specifically relates to a spatiotemporal joint trajectory generation method and system for fixed-wing unmanned aerial vehicles. Background Art

[0002] Drones are categorized into two types: fixed-wing and multi-rotor. Compared to multi-rotor drones, fixed-wing drones offer longer flight times, higher speeds, greater payload capacity, and greater range. Therefore, fixed-wing drones can cover larger and longer distances, are suitable for situations requiring rapid arrival, and can carry larger payloads. These advantages have led to their widespread use in commercial, military, and research fields. However, compared to multi-rotor drones, fixed-wing aircraft require more space for takeoff, landing, and turning due to their larger takeoff and landing areas and larger turning radius. Therefore, current flight planning for multi-rotor drones is not suitable for fixed-wing drones.

[0003] At present, non-autonomous intelligent fixed-wing UAVs are limited to cruising in a stable flight state on a predetermined path. Autonomous intelligent fixed-wing UAVs, without human intervention, can effectively reduce the risk of flight collisions based on autonomous decision-making and path planning control. However, for application scenarios with complex obstacles at low altitudes, real-time online trajectory planning is challenging due to the dynamic, random and interactive nature of the environment. In some cases, such as high-speed obstacle avoidance, fixed-wing UAVs need to perform fast and flexible maneuvers. Therefore, efficient and accurate real-time online trajectory planning technology for UAVs is imminent, especially in three-dimensional avoidance maneuvers and safe trajectories. Fixed-wing UAVs face multiple challenges in planning efficient and high-quality paths: (1) It is necessary to use onboard computers to frequently process a large amount of discrete data in three-dimensional maps. High-quality path planning solutions require fine time discretization of dynamic processes, which increases the complexity of the algorithm; (2) Due to the high nonlinearity of dynamic processes, the safety of UAVs in effective flight under extreme conditions faces difficulties; (3) Traditional parameterized flight trajectory methods are not time-efficient enough to adjust and optimize trajectories under continuous time constraints. The classic Dubins curve is commonly used as the trajectory of fixed-wing UAVs. The Dubins trajectory consists of three parts: an arc, a straight line, and another arc (or arcs alternating in three directions). A disadvantage of the Dubins trajectory is that there is a singularity at the junction of the straight line and the arc, and all turns of the drone are performed at maximum curvature.

[0004] Currently, commonly used continuous-time trajectory parameterization methods include polynomial splines, Bezier curves, and B-spline curves. Polynomial spline trajectory parameterization is characterized by high dimensionality, spatiotemporal coupling, and low code implementation difficulty. Bezier splines are characterized by high dimensionality, spatiotemporal coupling, and moderate code implementation difficulty. B-splines are characterized by low dimensionality, high spatiotemporal coupling, and moderate code implementation difficulty. These methods cannot jointly optimize information from both spatial and temporal perspectives, and the continuous-time constraints are not decoupled from the trajectory parameters. Aside from the norm constraints of higher-order derivatives, these methods do not support the full state and input control constraints of the UAV's dynamics. For example, patent CN111562797B discloses a method for optimizing the optimal real-time trajectory of a drone flight time to ensure convergence; patent CN 113624237B discloses a method for range-adjusted drone trajectory planning based on the Dubins curve; patent CN118409597A discloses a tree-based multi-Dubins robot collaborative coverage path planning method; and patent US11557210B2 discloses a drone obstacle avoidance method based on RRT and Dubins trajectory planning. These disclosed methods do not involve improving the Dubins trajectory for use in scenes with dense obstacles. Fixed-wing drone flight is primarily limited by trajectory curvature.

[0005] MINCO (Minimum Control Effort) trajectories refer to a set of spatiotemporal parameterized trajectories based on necessary and sufficient conditions for optimality. Based on the principle of differential flatness, MINCO, a multirotor trajectory planning framework, generates polynomial curves from points based on a set of spatiotemporal parameterized trajectories based on necessary and sufficient conditions for optimality, taking into account geometric configuration constraints and user-defined dynamic constraints. MINCO trajectories are parameterized by space and time vectors, consisting of continuous polynomial line segments connected end-to-end that satisfy both time and space constraints. MINCO is characterized by continuous time constraints, low dimensionality, independent and uncoupled spatiotemporal optimization, and moderate code implementation difficulty. MINCO's sparse parameters allow direct control of both spatial and temporal degrees of freedom (DOF) of trajectory deformation. The MINCO framework can be extended to differentially flat fixed-wing UAVs. Trajectories are deformed to satisfy curvature constraints for fixed-wing UAVs, typically using the finite memory Newton descent method (LBFGS) to optimize the objective function. No research has yet been published on the technical details of modifying the Dubins trajectory within the MINCO trajectory generation framework for agile flight trajectory planning in complex obstacle environments for fixed-wing UAVs. Summary of the Invention

[0006] The Dubins trajectory commonly used in fixed-wing UAVs is discontinuous, which can lead to unstable control and abnormal vibration of the UAV. In order to solve the problem of discontinuous Dubins trajectory commonly used in fixed-wing UAVs, the present invention provides a spatiotemporal joint trajectory generation method for fixed-wing UAVs, comprising:

[0007] The onboard computer of the fixed-wing UAV obtains the three-dimensional spatial map of the flight area, dense obstacle information, starting and target position points, and the physical parameters of the UAV;

[0008] The fixed-wing UAV uses the rapid exploration random algorithm RRT to calculate a path from the starting position to the target position, and processes it to generate a Dubins trajectory;

[0009] The Dubins trajectory is deformed based on the MINCO trajectory generation framework, and the spatiotemporal joint optimization trajectory is calculated using the improved Grey Wolf Optimization Algorithm (CMAGWO) with adaptive covariance matrix. MINCO refers to the set of spatiotemporal parameterized trajectories based on the necessary and sufficient conditions for optimality.

[0010] Preferably, the method of using a rapid randomized reconnaissance algorithm (RRT) to calculate a path from the starting position to the target position and processing the path to generate a Dubins trajectory includes:

[0011] Based on the three-dimensional space map, the RRT algorithm is used to determine the collision-free shortest path of the fixed-wing UAV from the starting position to the target position;

[0012] Generate Dubins loci based on the shortest paths.

[0013] Preferably, the deformation processing of the Dubins trajectory based on the MINCO trajectory generation framework includes:

[0014] Based on the flight parameters of the fixed-wing UAV, the goal is to minimize the flight time from the starting position to the target position and to have a smooth trajectory; and a spatiotemporal joint planning model is constructed with time and space joint constraints as constraints.

[0015] Based on the joint time and space constraints, the MINCO trajectory generation framework is used to modify the Dubins trajectory to obtain the fixed-wing UAV trajectory from the starting position to the target position.

[0016] Modifying the Dubins trajectory using the MINCO trajectory generation framework and using CMAGWO to perform a spatiotemporal joint planning model until an optimal solution is reached; the optimal solution being the optimal trajectory;

[0017] The calculation of the parameters in the time and space joint constraints is determined by the geometric constraints and dynamic constraints, speed and acceleration constraints of the fixed-wing UAV flight environment.

[0018] Preferably, the method of modifying the Dubins trajectory using the MINCO trajectory generation framework based on the joint time and space constraints to obtain the UAV trajectory from the starting position point to the target position point includes:

[0019] In the three-dimensional space map, based on the Dubins trajectory, the MINCO trajectory generation framework is used to calculate and determine the parameter vector of each trajectory segment, considering the joint constraints of time and space. There are two parameter vectors, including: a spatial vector and a temporal vector, and the spatial vector and the temporal vector are both one-dimensional vectors.

[0020] Preferably, in the three-dimensional space map, based on the Dubins trajectory, considering the joint constraints of time and space, the MINCO trajectory generation framework is used to calculate and determine the parameter vector of each trajectory segment, including:

[0021] Based on the Dubins trajectory, it is assumed that the time interval from the starting point to the target point through each point of the Dubins trajectory is averaged, and the space vector p(t i );

[0022] Based on the calculated space vector, adjusting the time vector after considering the joint time and space constraints;

[0023] Among them, p(t i ) is the MINCO trajectory with t i is the i-th trajectory segment at the starting time; p(t i ) is the time length [0,T i ]; T i For fixed-wing UAVs from the t i The waypoint corresponding to time flies to the tth i+1 The length of time that the waypoint corresponding to the moment has passed; the waypoint vector is q = (q1…q M-1 ), q i It is the tth i The waypoint corresponding to the time, i=1,2,...,M-1;time vector is T=(T1…T M-1 ) T , M is the number of MINCO trajectory segments.

[0024] Preferably, the fixed-wing UAV trajectory is p(t), which is composed of M polynomials, which are defined as

[0025] Where, is the matrix coefficient of the i-th polynomial; β(t)=(1,t,t 2 ,…,t N ) T It is the cardinality.

[0026] Preferably, the spatiotemporal joint planning model includes the following calculation formula:

[0027]

[0028] Where p (s) (t) is the s-order differential of the MINCO trajectory, usually s = 0, 1, 2; ρ T is the time regularization coefficient; ρ T t M represents the time regularization of the cost functional; t M is the total flight time of the M-segment MINCO trajectory; p (0:s-1) (0) is the s-1 order of the MINCO trajectory; p(t i ) is the MINCO trajectory with t i is the i-th trajectory segment at the starting time; M is the number of MINCO trajectory segments; T is the total duration of the MINCO trajectory; T i is the duration of the ith segment of the MINCO trajectory; v max is the maximum allowed flight speed; a max is the maximum allowed flight acceleration; d0 is the attitude of the fixed-wing UAV at the starting point; d g is the attitude of the fixed-wing UAV at the endpoint; the i-th timestamp is t i ;DubinsPath i is the i-th Dubins locus.

[0029] Preferably, the method of modifying the Dubins trajectory using the MINCO trajectory generation framework and optimizing the spatiotemporal joint planning model using CMAGWO until an optimal solution is reached comprises:

[0030] Step 1: The coordinates of the discrete points of the Dubins trajectory are taken as population A, where the individuals in the population are divided into α, β and δ, and their positions are recorded as Q α , Q β and Q δ ;

[0031] Step 2: Based on the objective function of minimizing flight time and smoothing trajectory in the spatiotemporal joint planning model, calculate the fitness values ​​of population A and the individuals of the new population B generated based on population A.

[0032] Step 3: Select a portion of individuals with high fitness values ​​from population A and new population B and concatenate them to construct a transition vector;

[0033] Step 4: Based on the adaptive covariance matrix generated by the transition vector, an elliptical search area is generated, and Gaussian distributed sampling is performed, and multiple candidate solution vectors are randomly selected from the area;

[0034] Step 5: Determine whether the maximum number of iterations has been reached. If so, proceed to step 6. If not, go to step 2.

[0035] Step 6: Return the individual with the best fitness, and the position vector of the corresponding individual is the optimal solution; the position vector corresponds to a one-dimensional vector composed of the spatial parameter and time parameter solutions.

[0036] Based on the same inventive concept, the present application also discloses a spatiotemporal joint trajectory generation system for a fixed-wing UAV, comprising:

[0037] An information acquisition module is used to obtain a three-dimensional spatial map of the flight area, dense obstacle information, starting and target positions, and physical parameters of the UAV based on the fixed-wing UAV's onboard computer;

[0038] A Dubins trajectory generation module is used to calculate a path from the starting position to the target position using a rapid exploration random algorithm RRT based on the fixed-wing UAV, and process it to generate a Dubins trajectory;

[0039] The optimal trajectory determination module is used to deform the Dubins trajectory based on the MINCO trajectory generation framework and calculate the spatiotemporal joint optimization trajectory using the adaptive covariance matrix improved gray wolf optimization algorithm CMAGWO; where MINCO refers to the set of spatiotemporal parameterized trajectories based on the necessary and sufficient conditions for optimality.

[0040] Preferably, the system further comprises:

[0041] The trajectory management module is used to realize the operation of the fixed-wing UAV along the predetermined route through trajectory tracking;

[0042] The trajectory tracking module is used to send commands to the drone's underlying controller to control the aircraft.

[0043] Based on the same inventive concept, the present application also provides an electronic device, comprising: at least one processor and a memory; the memory and the processor are connected via a bus;

[0044] The memory is used to store one or more programs;

[0045] When the one or more programs are executed by the at least one processor, a spatiotemporal joint trajectory generation method for a fixed-wing UAV provided by the present invention is implemented.

[0046] Based on the same inventive concept, the present application also provides a readable storage medium having an execution program stored thereon. When the execution program is executed, a spatiotemporal joint trajectory generation method for a fixed-wing UAV provided by the present invention is implemented.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] The present application discloses a spatiotemporal joint trajectory generation method and system for a fixed-wing unmanned aerial vehicle (UAV), comprising: the fixed-wing UAV's onboard computer having acquired a three-dimensional spatial map of the flight area, dense obstacle information, a starting position and a target position, and the UAV's physical parameters; the fixed-wing UAV using a rapid randomized reconnaissance algorithm (RRT) to calculate a path from the starting position to the target position, and processing the calculated path to generate a Dubins trajectory; the Dubins trajectory is deformed based on the MINCO trajectory generation framework, and CMAGWO is used to perform spatiotemporal joint optimization calculations on the Dubins trajectory to obtain an optimal flight trajectory; the trajectory generated using the technical solution provided by the present application is smooth and has the shortest flight time, which is beneficial for the fixed-wing UAV to fly autonomously in complex low-altitude environments such as cities;

[0049] The flight trajectory of the fixed-wing UAV generated using the technical solution provided by this application is continuous, which can improve the flight stability of the fixed-wing UAV and avoid unstable vibration of the fuselage caused by Dubins trajectory singularities. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 This is a flow chart of the spatiotemporal joint trajectory generation method for fixed-wing UAVs of the present invention;

[0051] Figure 2 The principle of calculating MINCO trajectory parameters based on CMAGWO;

[0052] Figure 3 Side view of planning a collision-free path using RRT in the implementation case;

[0053] Figure 4 A top view of a collision-free path planned using RRT in the implementation case;

[0054] Figure 5 is the uncorrected Dubins locus in the implementation case;

[0055] Figure 6 The MINCO-modified Dubins trajectory is used in the implementation case;

[0056] Figure 7 A system architecture diagram for spatiotemporal joint trajectory generation for fixed-wing UAVs;

[0057] Figure 8 Provides a framework for spatiotemporal joint path optimization systems;

[0058] Figure 9 This is a schematic diagram of the structure of an electronic device provided in this application. DETAILED DESCRIPTION

[0059] In order to overcome the problem of discontinuous curvature of the Dubins trajectory commonly used in fixed-wing UAVs, this application proposes a joint trajectory optimization for fixed-wing UAVs from both spatial and temporal perspectives, and the continuous time constraints and spatial trajectory parameters are solved independently; at the same time, this application also applies the MINCO trajectory generation framework to the modified Dubins trajectory for fixed-wing UAVs. This application uses the random optimization algorithm CMAGWO to replace the limited memory Newton descent method to calculate the MINCO trajectory generation framework to modify the Dubins trajectory. The advantage of CMAGWO is that it does not require the calculation of gradients, and when the problem is of ultra-large scale, it reflects the algorithmic advantages of random search. The present invention is beneficial for fixed-wing UAVs to fly autonomously in low-altitude environments with dense obstacles such as cities, generating smooth trajectories and the shortest flight time. In order to better understand the present invention, the contents of the present invention are further explained below in conjunction with the drawings and examples in the specification.

[0060] Example 1:

[0061] like Figure 1 As shown, this embodiment provides a method for generating a spatiotemporal joint trajectory for a fixed-wing UAV, comprising:

[0062] S1. The onboard computer of the fixed-wing UAV has obtained the three-dimensional spatial map of the flight area, dense obstacle information, starting and target positions, and the physical parameters of the UAV;

[0063] Specifically, the fixed-wing UAV obtains a three-dimensional spatial map of the flight area, determines and marks the three-dimensional information of obstacles in the three-dimensional spatial map, and marks the starting position point and the target position point.

[0064] S2, the fixed-wing UAV uses the rapid exploration random algorithm RRT to calculate a path from the starting position to the target position, and processes it to generate a Dubins trajectory;

[0065] S3. Deform the Dubins trajectory based on the MINCO trajectory generation framework and calculate the spatiotemporal joint optimization trajectory using the modified grey wolf optimization algorithm with adaptive covariance matrix (CMAGWO); here, MINCO (Minimum Control Effort) refers to a set of spatiotemporal parameterized trajectories based on the necessary and sufficient conditions for optimality.

[0066] Furthermore, step S2 includes:

[0067] The rapid randomized exploration (RRT) algorithm is used to calculate the shortest and collision-free path from the starting point to the target point. The collision-free path is processed to generate the Dubins trajectory, and the corrected path is checked until it meets the dynamic physics of the fixed-wing UAV.

[0068] The problem of space-time joint trajectory optimization calculation for fixed-wing UAV can be abstracted into a mathematical optimization calculation problem. The flight trajectory is composed of a series of curve segments from the starting position to the target position, and the vector formed by the starting and ending points of each curve segment is recorded as the waypoint vector q = (q1…q M-1 ),in It is the tth i The starting point of the i-th curve segment at time. Waypoint q i With q i+1 The length of the curve can be set, and the waypoints can be interpolated. M-1 ) T is the vector formed by the flight time of the UAV in each curve segment, Here, t represents a moment, and T represents the length of time.

[0069] MINCO is a trajectory with space vector and time vector as parameters, that is, a continuous polynomial line segment that satisfies time and space constraints is connected end to end. i , adjacent polynomial trajectory segments p i The midpoint of (t) is recorded as The starting point state of the UAV trajectory p(t) is recorded as d0 (s-1 order differential), and the ending point state is recorded as d g (s-1 order differential).

[0070] Based on the flight parameters of the fixed-wing UAV, with the goal of minimizing the flight time from the starting position to the target position and maintaining a smooth trajectory, and considering the geometric constraints and dynamic constraints of the fixed-wing UAV flight environment, as well as the speed and acceleration constraints, the trajectory optimization calculation objective function is used to construct a spatiotemporal joint planning model, which is expressed as follows:

[0071]

[0072] Where p (s) (t) is the s-order differential of the MINCO trajectory, usually s = 0, 1, 2; ρ T is the time regularization coefficient; t M is the total flight time of the MINCO trajectory; p (0:s-1) (0) is the s-1 order of the MINCO trajectory; p(t) is a time-uniform MINCO trajectory; p(t i ) is the starting point of a certain segment (a total of M segments) of the MINCO trajectory i The trajectory segment at time (i segment); p(t i ) has a time span of [0,T i ]; Consider the i-th segment of a time-uniform MINCO trajectory The time is taken as 0 relative to the start time of the segment, which is a 2s-1 degree polynomial with s as the order; the total duration of the trajectory is T; ρ T t M represents the time regularization of the cost functional, ρ T is the time regularization coefficient; v max is the maximum allowed flight speed; a max is the maximum allowed flight acceleration; M is the number of polynomial segments, which is also the dimension of the trajectory; d0 is the starting attitude; d g is the terminal posture; DubinsPath i is the i-th Dubins trajectory segment,

[0073] The discrete point sequence is obtained by sampling the Dubins trajectory generated by the fixed-wing UAV, which is defined as follows:

[0074] P={(w1,χ1),(w2,χ2),…,(w N ,χ N )} (2)

[0075] The minimum control polynomial MINCO trajectory of the fixed-wing UAV is defined as:

[0076]

[0077] Where the dimension m, the order of the integration chain, and the initial and final boundary conditions in this definition are all arbitrary constants.

[0078] Among them, an m-dimensional trajectory p(t) has M polynomials and the degree of freedom is N=2s-1. The i-th trajectory p i (t) is defined as:

[0079]

[0080] in, is the matrix coefficient of the i-th polynomial; β(t)=(1,t,…,t N ) T is the cardinality; T i is the duration of the i-th polynomial, the time vector T=(T1,…,T M ) T .

[0081] All trajectory segments ξ MINCO There are only compact parameter vectors q and T, where q=(q1,…,q M-1 ), q i is the middle point. The linear formula for evaluating the entire trajectory through the parameter vectors q and T is as follows:

[0082] c=M(q,T) (5)

[0083] In step 3 of this application, CMAGWO is introduced to solve the spatiotemporal joint planning model and obtain the optimal trajectory. Specifically, CMAGWO is used to solve the MINCO trajectory parameters q and T. The calculated MINCO trajectory meets the spatiotemporal deformation requirements of the mission and ensures local smoothness of the trajectory. The trajectory parameters q and T are concatenated end to end to form a one-dimensional vector, which serves as the solution parameter.

[0084] A candidate solution for CMAGWO is a one-dimensional vector, representing the position vectors of individuals in a pack. The position of a single individual in the pack corresponds to an element of a one-dimensional vector consisting of the parameters q and T. The dynamic game of a pack of gray wolves hunting their target prey, as they approach, is the process of solving for the trajectory parameters q and T.

[0085] In the process of solving the objective function, the cost function is calculated until it reaches a minimum value, and finally a smooth trajectory with the shortest flight time is obtained.

[0086] In order to improve the convergence of the classic GWO for calculating the MINCO trajectory generation framework to correct the Dubins trajectory, this patent improves the classic GWO to CMAGWO to increase the convergence speed of the solution. First, a new population with evolutionary ability (here denoted as h) is added on the basis of the four population searches of the classic GWO. The random search of the evolutionary population follows a Gaussian distribution in the initial position, and an elliptical directional search direction is adopted; second, the best individuals selected from the original four populations of the classic GWO are shared with the h population, so that the exploration ability of the h population is improved by integrating high-quality individuals. CMAGWO adopts an evolutionary strategy of adaptive covariance. According to the parameter vector obtained from historical iterations, the covariance of the Gaussian distribution is calculated. The geometric shape of the covariance distribution is a directional elliptical area. The search area is dynamically updated according to each historical iteration, and the parameter vector is finally approximated to the optimal value.

[0087] Using several waypoints to determine the trajectory, we further use MINCO to deform the Dubins trajectory. We need to find the waypoints and times that satisfy the deformation. When implementing the code, we need to combine the waypoint vector and the time vector into a one-dimensional vector as the parameter to be solved to facilitate numerical calculations.

[0088] The waypoint vector and time vector are optimized calculation parameters, and the MINCO framework is used to optimize and calculate a UAV trajectory. Using the MINCO framework, a joint time and space planning model (including objective functions and constraints) is constructed. The waypoint vector and time vector are uncoupled (the space vector is calculated by assuming the time average of each trajectory segment, then the space vector is fixed and the time vector is adjusted). Therefore, the two parameters are uncoupled and can be calculated separately). CMAGWO is used to solve the problem and obtain several solution vectors.

[0089] Figure 2 The principle of solving MINCO trajectory parameters based on CMAGWO.

[0090] Step 1: The discrete points of the Dubins trajectory are taken as population A. The individuals in the population are divided into α, β and δ, and their position vectors are recorded as: Q α , Q β and Q δ ;

[0091] Step 2: Based on the objective function of minimizing flight time and smoothing trajectory in the spatiotemporal joint planning model, calculate the fitness values ​​of population A and the individuals of the new population B generated based on population A.

[0092] Step 3. Select a part of individuals with high fitness values ​​from population A and new population B and splice them to construct a transition vector; here, individuals with high fitness values ​​can be selected according to a certain proportion, for example, the top 10%, 20% or 30% with high fitness values, etc. The number of individuals to be selected can be determined according to the computing task, and this application does not make any specific restrictions.

[0093] Step 4: Based on the adaptive covariance matrix generated by the transition vector, an elliptical search area is generated, and Gaussian distributed sampling is performed, and multiple candidate solution vectors are randomly selected from the area;

[0094] Step 5: Determine whether the maximum number of iterations has been reached. If so, proceed to step 6. If not, go to step 2.

[0095] Step 6: Return the individual with the best fitness. The position vector (one-dimensional vector) of the corresponding individual is the optimal solution (a one-dimensional vector concatenated from the spatial parameter and time parameter solutions. One-dimensional vector is used here for the convenience of calculation).

[0096] Example 2

[0097] Below, taking a fixed-wing UAV performing an inspection mission in a city as an example, a spatiotemporal joint trajectory generation method for a fixed-wing UAV provided in this application is specifically introduced.

[0098] This task involves planning the shortest path from a starting point to a destination, passing through five fixed navigation points along the way. This involves combined spatial and temporal trajectory planning. The fixed-wing drone's starting coordinates are marked as the origin (0, 0, 0), and it hovers around the destination coordinates (120.647, 306.801, -100). Ten cylindrical obstacles are set up to simulate urban buildings. The geometric parameters are shown in Table 1.

[0099] Table 1 Geometric parameters of obstacles (unit: m)

[0100] Serial number north East radius high 1 316.985 -263.847 55.569 111.323 2 -149.911 -270.88 68.7 142.556 3 527.214 -173.106 18.532 45.371 4 754.445 -275.778 9.801 20.127 5 -305.869 -176.23 35.847 36.068 6 -139.892 204.732 62.841 133.492 7 -154.027 -498.076 28.409 23.014 8 558.791 -301.937 82.116 50.992 9 565.974 -102.069 52.776 215.224 10 -232.717 -88.537 18.483 135.727

[0101] Table 2 Position coordinates of navigation points that fixed-wing drones must pass through (unit: m)

[0102]

[0103]

[0104] Based on the above data, the spatiotemporal joint trajectory generation method for fixed-wing UAVs provided in this application is used for optimization. The specific steps are as follows:

[0105] Step S1: Initialize a three-dimensional spatial map within the area covered by the starting position point and the target end position of the fixed-wing UAV, and mark dense obstacle information, the starting position point and the target position point in the three-dimensional spatial map;

[0106] Step S2: using the rapid randomized reconnaissance algorithm (RRT) to calculate a path from the starting point to the target point, and processing the path to obtain a Dubins trajectory;

[0107] The coordinates of the discrete points of the trajectory are shown in Table 3. Since the generated path has not yet considered the physical performance constraints of the fixed-wing UAV, it cannot be used for the actual flight path of the fixed-wing UAV.

[0108] Figure 3 Side view of a collision-free path planned using RRT, with a cylinder as an obstacle. Figure 4 This is an overhead view of a collision-free path planned using RRT. The curve passing through the obstacle is the collision-free path, and the yellow points on the path are the waypoints along the path.

[0109] The figure shows the uncorrected Dubins trajectory. There are two intersecting lines that pass through an obstacle. The thicker curve is the path calculated using RRT, and the thinner curve is the Dubins trajectory.

[0110] Table 3 Coordinates of discrete points on a collision-free path planned using RRT (unit: m)

[0111] Serial number north East land 1 164.679 -310.639 -100 2 3.004 -353.726 -100 3 -63.133 -186.854 -100 4 -132.517 52.563 -100 5 -295.008 -38.539 -100 6 -394.257 31.631 -100 7 -380.416 110.970 -100 8 -196.470 305.178 -100 9 120.647 306.802 -100

[0112] Step S3: deform the Dubins trajectory based on the MINCO trajectory generation framework, and use CMAGWO to perform spatiotemporal joint optimization calculation on the Dubins trajectory to obtain an optimal trajectory.

[0113] The Dubins trajectory considers six possible configurations: LSL, LSR, RSL, RSR, RLR, and LRL, where L and R represent left and right turns, respectively, and S represents a straight line segment. The Dubins trajectory is discontinuous, which can lead to unstable control of the fixed-wing UAV. Occasional vibrations can be observed during cornering in the simulation.

[0114] The MINCO trajectory generation framework is used to modify the Dubins trajectory to obtain the optimal path to the target location. The modified path is checked to ensure that it meets the dynamic constraints of the fixed-wing UAV.

[0115] The minimum control polynomial MINCO trajectory of the fixed-wing UAV is defined as:

[0116]

[0117] Here, an m-dimensional trajectory p(t) is composed of M polynomials with N=2s-1 degrees of freedom.

[0118] The i-th MINCO trajectory is defined as:

[0119]

[0120] in, is the matrix coefficient of the i-th polynomial; β(t)=(1,t,…,t N ) T is the cardinality. i is the duration of the i-th polynomial, the time vector T=(T1,…,T M ) T .

[0121] MINCO track ξ MINCO There are only compact parameter vectors q and T, where q=(q1,…,q M-1 ), q i is the middle point. The linear formula for the entire trajectory is evaluated by the parameter vectors q and T:

[0122] c=M(q,T) (8)

[0123] The MINCO trajectory parameters are solved based on CMAGWO. In the process of solving the objective function, the cost function is calculated until it reaches a minimum value.

[0124] Figure 6 This is the MINCO-corrected Dubins trajectory in the implementation case. The thicker curve is the path calculated using RRT, and the thinner curve is the MINCO-corrected Dubins trajectory.

[0125] The experimental results show that the fixed-wing UAV follows the MINCO trajectory generation framework to correct the Dubins trajectory, so the MINCO trajectory generation framework to correct the Dubins trajectory is feasible.

[0126] Example 3:

[0127] In order to implement the method of the above embodiment, the present invention also provides a spatiotemporal joint trajectory generation system for fixed-wing UAV, including: an information acquisition module, a Dubins trajectory generation module, an optimal trajectory determination module, a trajectory management module and a trajectory tracking module.

[0128] like Figure 7 As shown, the information acquisition module includes map information collection and environmental obstacle sampling, which is used to initialize the three-dimensional space map within the flight area of ​​the fixed-wing UAV and mark the dense obstacle information, starting position point and target position point, and the physical parameters of the UAV in the three-dimensional space map;

[0129] The Dubins trajectory generation module generates a Dubins trajectory, specifically for calculating a path from the starting position point to the target position point using a rapid exploration random algorithm RRT, and processing the path to generate a Dubins trajectory;

[0130] The optimal trajectory determination module uses the MINCO curve generation framework to correct the Dubins trajectory. Specifically, it is used to deform the Dubins trajectory based on the MINCO trajectory generation framework and use CMAGWO to perform spatiotemporal joint optimization calculation on the Dubins trajectory to obtain an optimal trajectory.

[0131] The trajectory management module enables the fixed-wing UAV to run along the predetermined trajectory through trajectory tracking;

[0132] The trajectory tracking module sends commands to the drone's underlying controller to control the aircraft. The above modules rely on the data from the onboard sensors to filter and obtain the flight state estimation results.

[0133] like Figure 8 As shown, the system framework of the spatiotemporal joint trajectory generation system for fixed-wing UAVs constructed using the above modules is used to implement a spatiotemporal joint trajectory generation method for fixed-wing UAVs in the present application. The specific implementation process can be referred to the above example and will not be repeated here.

[0134] Example 4

[0135] like Figure 9As shown, the present invention also provides an electronic device, which may be a computer, a single-chip microcomputer, a smart mobile device, or the like. The electronic device in this embodiment may include a processor, a memory, a transceiver component, and the like. The memory, processor, and transceiver component are connected via a bus; the memory may be used to store an execution program, which may include instructions; and the processor may be used to execute the instructions stored in the memory. The memory may also be used to store data, which may be accessed and / or modified during the execution of the instructions.

[0136] The processor may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in the storage medium to implement the corresponding method flow or corresponding function, so as to implement the steps of a spatiotemporal joint trajectory generation method for a fixed-wing UAV in the above embodiment.

[0137] Example 5

[0138] Based on the same inventive concept, the present invention also provides a readable storage medium, specifically an electronic device readable storage medium (Memory), which is a memory device in the electronic device for storing programs and data. It is understandable that the storage medium here can include both built-in storage media in the electronic device and, of course, extended storage media supported by the electronic device. The storage medium provides a storage space that stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space. These instructions can be one or more execution programs (including program code). It should be noted that the storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk storage. The processor loads and executes one or more instructions stored in the storage medium to implement the steps of a method for generating a spatiotemporal joint trajectory for a fixed-wing unmanned aerial vehicle in the above embodiment.

[0139] Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0140] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0141] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0142] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0143] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0144] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention are included in the scope of the claims of the present invention to be approved.

Claims

1. A method for generating spatiotemporal joint trajectories for fixed-wing UAVs, characterized in that: include: The onboard computer of the fixed-wing UAV obtains the three-dimensional spatial map of the flight area, dense obstacle information, starting and target position points, and the physical parameters of the UAV; The fixed-wing UAV uses the rapid exploration random algorithm RRT to calculate a path from the starting position to the target position, and processes it to generate a Dubins trajectory; The Dubins trajectory is deformed based on the MINCO trajectory generation framework, and the spatiotemporal joint optimization trajectory is calculated using the improved Grey Wolf Optimization Algorithm (CMAGWO) with adaptive covariance matrix. MINCO refers to the set of spatiotemporal parameterized trajectories based on the necessary and sufficient conditions for optimality.

2. The method according to claim 1, wherein The rapid randomized exploration algorithm (RRT) is used to calculate a path from the starting point to the target point, and the path is processed to generate a Dubins trajectory, including: Based on the three-dimensional space map, the RRT algorithm is used to determine the collision-free shortest path of the fixed-wing UAV from the starting position to the target position; Generate Dubins loci based on the shortest paths.

3. The method according to claim 2, wherein The deformation processing of the Dubins trajectory based on the MINCO trajectory generation framework includes: Based on the flight parameters of the fixed-wing UAV, the goal is to minimize the flight time from the starting position to the target position and to have a smooth trajectory; and a spatiotemporal joint planning model is constructed with time and space joint constraints as constraints. Based on the joint time and space constraints, the MINCO trajectory generation framework is used to modify the Dubins trajectory to obtain the fixed-wing UAV trajectory from the starting position to the target position. Modifying the Dubins trajectory using the MINCO trajectory generation framework and solving the spatiotemporal joint planning model using CMAGWO until an optimal solution is reached; the optimal solution being the optimal trajectory; Preferably, the method of modifying the Dubins trajectory using the MINCO trajectory generation framework based on the joint time and space constraints to obtain the UAV trajectory from the starting position point to the target position point includes: In the three-dimensional space map, based on the Dubins trajectory, the MINCO trajectory generation framework is used to calculate and determine the parameter vector of each trajectory segment, considering the joint constraints of time and space. There are two parameter vectors, including: a spatial vector and a temporal vector, and the spatial vector and the temporal vector are both one-dimensional vectors.

4. The method according to claim 3, wherein In the three-dimensional space map, based on the Dubins trajectory, the parameter vector of each trajectory segment is calculated and determined using the MINCO trajectory generation framework, taking into account the joint constraints of time and space, including: Based on the Dubins trajectory, it is assumed that the time interval from the starting point to the target point through each point of the Dubins trajectory is averaged, and the space vector p(t i ); Based on the calculated space vector, adjusting the time vector after considering the joint time and space constraints; Among them, p(t i ) is the MINCO trajectory with t i is the i-th trajectory segment at the starting time; p(t i ) is the time length [0,T i ]; T i For fixed-wing UAVs from the t i The waypoint corresponding to time flies to the tth i+1 The length of time that the waypoint corresponding to the moment has passed; the waypoint vector is q = (q1…q M-1 ), q i It is the tth i The waypoint corresponding to the time, i=1,2,...,M-1;time vector is T=(T1…T M-1 ) T , M is the number of MINCO trajectory segments.

5. The method according to claim 4, wherein The fixed-wing UAV trajectory is p(t), which is composed of M polynomials, which are defined as Where, is the matrix coefficient of the i-th polynomial; β(t)=(1,t,t 2 ,…,t N ) T It is the cardinality.

6. The method according to claim 3, wherein The spatiotemporal joint planning model includes the following calculation formula: Where p (s) (t) is the s-order differential of the MINCO trajectory, usually s = 0, 1, 2; ρ T is the time regularization coefficient; ρ T t M represents the time regularization of the cost functional; t M is the total flight time of the M-segment MINCO trajectory; p (0:s-1) (0) is the s-1 order of the MINCO trajectory; p(t i ) is the MINCO trajectory with t i is the i-th trajectory segment at the starting time; M is the number of MINCO trajectory segments; T is the total duration of the MINCO trajectory; T i is the duration of the ith segment of the MINCO trajectory; v max is the maximum allowed flight speed; a max is the maximum allowed flight acceleration; d0 is the posture of the fixed-wing UAV at the starting point; d g The fixed-wing UAV is at the terminal attitude; Dubins Path i is the i-th Dubins locus; Preferably, the method of modifying the Dubins trajectory using the MINCO trajectory generation framework and optimizing the spatiotemporal joint planning model using CMAGWO until an optimal solution is reached comprises: Step 1: The coordinates of the discrete points of the Dubins trajectory are taken as population A, where the individuals in the population are divided into α, β and δ, and their positions are recorded as Q α , Q β and Q δ ; Step 2: Based on the objective function of minimizing flight time and smoothing trajectory in the spatiotemporal joint planning model, calculate the fitness values ​​of population A and the individuals of the new population B generated based on population A. Step 3: Select a portion of individuals with high fitness values ​​from population A and new population B and concatenate them to construct a transition vector; Step 4: Based on the adaptive covariance matrix generated by the transition vector, an elliptical search area is generated, and Gaussian distributed sampling is performed, and multiple candidate solution vectors are randomly selected from the area; Step 5: Determine whether the maximum number of iterations has been reached. If so, proceed to step 6. If not, go to step 2. Step 6: Return the individual with the best fitness, and the position vector of the corresponding individual is the optimal solution; the position vector corresponds to a one-dimensional vector composed of the spatial parameter and time parameter solutions.

7. A spatiotemporal joint trajectory generation system for fixed-wing UAVs, characterized in that: include: An information acquisition module is used to obtain a three-dimensional spatial map of the flight area, dense obstacle information, starting and target positions, and physical parameters of the UAV based on the fixed-wing UAV's onboard computer; A Dubins trajectory generation module is used to calculate a path from the starting position to the target position using a rapid exploration random algorithm RRT based on the fixed-wing UAV, and process it to generate a Dubins trajectory; The optimal trajectory determination module is used to deform the Dubins trajectory based on the MINCO trajectory generation framework and calculate the spatiotemporal joint optimization trajectory using the adaptive covariance matrix improved gray wolf optimization algorithm CMAGWO; where MINCO refers to the set of spatiotemporal parameterized trajectories based on the necessary and sufficient conditions for optimality.

8. The system according to claim 7, wherein: Also includes: The trajectory management module is used to realize the operation of the fixed-wing UAV along the predetermined route through trajectory tracking; The trajectory tracking module is used to send commands to the drone's underlying controller to control the aircraft.

9. An electronic device, characterized in that: include: at least one processor and memory; The memory and the processor are connected via a bus; The memory is used to store one or more programs; When the one or more programs are executed by the at least one processor, the spatiotemporal joint trajectory generation method for a fixed-wing UAV according to any one of claims 1 to 6 is implemented.

10. A readable storage medium, characterized in that: An execution program is stored thereon, and when the execution program is executed, the spatiotemporal joint trajectory generation method for a fixed-wing unmanned aerial vehicle as described in any one of claims 1 to 6 is implemented.

Images