Unmanned aerial vehicle intelligent flight path planning method in multi-type scene

Abstract

The invention discloses an intelligent flight path planning method for an unmanned aerial vehicle in a multi-type scene, and the method comprises the following steps: 1, building task input and task type recognition: setting the unmanned aerial vehicle to be in a fixed-height flight mode, and obtaining space information required by a task, including a take-off point position, a task point coordinate set and a task region boundary, the system identifies the task type through analysis; if the input is a discrete point set, the task is judged as a point task; if the structure is a broken line structure, judging the task as a line task; if the boundary is a task area boundary, the task is judged to be a surface task, and if the boundary width L is small, the task is still judged to be a line task. The method has the advantages that a fitness function is improved, the shortest total time of multiple unmanned aerial vehicle tasks is used as an optimization target, and task real-time weight parameters are introduced, so that a planning result meets task timeliness and energy consumption optimization at the same time; the coverage of planar tasks is realized in combination with a cattle farming algorithm, dynamic planning of electric quantity and a battery swap station is performed in combination with a greedy algorithm, and path optimization and improvement of energy management are realized.

Description

Technical Field

[0001] This invention belongs to the field of unmanned aerial vehicles (UAVs), specifically relating to an intelligent flight path planning method for UAVs in various scenarios. Background Technology

[0002] With the widespread application of drone technology in fields such as power line inspection, agricultural plant protection, urban security, and disaster relief, drones are gradually evolving towards "driverless," automated, and intelligent operations. Faced with diverse and complex operational scenarios, traditional path planning methods struggle to balance flight efficiency, energy consumption control, and full coverage requirements, often leading to issues such as overlapping coverage, path redundancy, and even missed inspections. Especially in long-duration, high-load inspection missions, drones need to fly continuously for extended periods, thus necessitating solutions to power consumption and battery replacement issues.

[0003] In some large-scale, long-distance inspection missions, traditional path planning algorithms cannot simultaneously address the requirements of flight efficiency, energy consumption control, and mission coverage, resulting in the inability to achieve efficient global path optimization. These problems are particularly prominent in long-distance inspection missions such as power line inspection, oil and gas pipeline inspection, and agricultural plant protection. Therefore, how to combine mission type and power management to achieve intelligent path planning to ensure the successful completion of missions is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0004] To address the problems and shortcomings of existing technologies, the purpose of this invention is to provide an intelligent flight path planning method for unmanned aerial vehicles (UAVs) in various scenarios. This method can intelligently select the optimal path planning strategy according to the mission scenario, taking into account path shortestization, energy consumption optimization, and comprehensive mission coverage. It not only improves the operational efficiency of UAVs in complex environments but also enhances the intelligence and adaptability of the system, providing key technical support for UAVs to perform various types of missions.

[0005] A method for intelligent flight path planning of unmanned aerial vehicles (UAVs) in various scenarios includes the following steps: Step 1, Establish Task Input and Task Type Recognition: Set the UAV to fixed altitude flight mode and acquire the spatial information required for the task, including the takeoff point location, task point coordinate set, and task area boundary. The system analyzes and identifies the task type: if the input is a discrete point set, it is identified as a point task; if it is a polyline structure, it is identified as a line task; if it is a task area boundary, it is identified as a surface task. If the boundary width L is small, it is still identified as a line task. Step 2, Execute Path Planning Calculation: The system automatically schedules the appropriate path planning algorithm model based on the identified task type: for point tasks and line tasks, the genetic algorithm module is called to sequentially perform population initialization, path encoding, crossover and mutation, and fitness function calculation. The system performs numerical calculations (using path length and energy consumption as indicators) and finally outputs the optimized optimal path sequence. For area tasks, it calls the ox-plowing-style coverage algorithm module to map the region into the minimum bounding rectangle and generates equidistant grid routes according to the set flight interval. The system achieves full coverage path output for the region through a "back-and-forth" strategy. Step 3: Output the path results and deploy the flight mission: The system outputs the optimized flight path results as a waypoint sequence, including the GPS coordinates and execution order of each waypoint. The path can be directly sent to the UAV terminal to start the automatic flight mission. If it is a multi-aircraft collaborative operation mission, the system can divide the mission path into segments and deploy them to multiple flight platforms according to the area division logic to improve the operation efficiency.

[0006] Furthermore, in step 2, the initialization phase of the genetic algorithm uses the optimal point set method to generate a portion of the population. The principle of the optimal point set is as follows: Let Gs be a unit cube in s-dimensional Euclidean space. If τ∈Gs, then the optimal point set... Represented as: ; The optimal point set is mapped to the search space, and the formula for calculating the point coordinates is: ; In the formula and Let represent the upper and lower bounds of the j-th dimension, respectively; Gs: a unit cube in s-dimensional Euclidean space, and τ is a point in the s-dimensional Euclidean unit cube; : Best point set, representing the set of coordinates of the k-th point generated by the best point set method; n: The total number of prime concentration points (population size); s: The dimension of the problem (i.e., the dimension of the chromosome or the dimension of the task point coordinates); : A positive integer constant coprime to n, used to generate a uniformly distributed sequence (s=1,2,…,s); : The actual coordinates of the i-th individual in the j-th dimension; : Upper bound of the j-th dimension; : Lower Bound of the j-th dimension; k: The index of the point, with a value range of 1≤k≤n.

[0007] Furthermore, in step 2, the fitness function also incorporates the real-time requirements of the task scenario. The optimization objective is to minimize the total task completion time under multi-UAV collaborative operation. During the calculation process, tasks are assigned weights according to their real-time priority. Tasks with higher weights are prioritized in path planning, and the flight segment tasks of each UAV are dynamically adjusted to shorten the response time of high-priority tasks. The fitness function can be expressed as: ; in: m represents the number of drones; The time required to complete the task assignment for the i-th drone; Let be the real-time weight for task j; The difference between the actual completion time and the expected completion time of task j; This is the real-time penalty coefficient.

[0008] Furthermore, path planning also needs to incorporate a greedy algorithm to select the path with the most suitable estimated power consumption. The greedy algorithm includes: establishing a power consumption prediction model, power consumption prediction in path planning, greedy path selection considering power consumption, battery prediction model and battery swapping decision, and power recovery at the battery swapping station.

[0009] The advantages of this invention are: To adapt to the needs of different operational scenarios, this invention classifies task types into various forms such as point tasks (e.g., fixed-point monitoring), linear tasks (e.g., pipeline inspection), and area tasks (e.g., farmland spraying, post-disaster survey). It proposes a multi-scenario adaptive fusion algorithm for UAV path planning: a best-point set method is introduced into the genetic algorithm to initialize the population, significantly improving the uniformity of the initial solution and the coverage of the search space, reducing the risk of getting trapped in local optima; the fitness function is improved, using the shortest total time of multiple UAV tasks as the optimization objective, and a task real-time weight parameter is introduced, ensuring that the planning results simultaneously satisfy task timeliness and energy consumption optimization; the ox-plowing algorithm is combined to achieve efficient coverage of area tasks, and a greedy algorithm is integrated for dynamic planning of power consumption and battery swapping stations, achieving synergistic improvement in path optimization and energy management.

[0010] This method can intelligently select the optimal path planning strategy according to the task scenario, taking into account the shortest path, the best energy consumption, and the comprehensiveness of task coverage. It not only improves the operating efficiency of UAVs in complex environments, but also enhances the intelligence and adaptability of the system, providing key technical support for UAVs to perform multiple types of tasks. Attached Figure Description

[0011] Figure 1 This is a flowchart illustrating a method for intelligent flight path planning of unmanned aerial vehicles (UAVs) in multiple scenarios according to the present invention. Detailed Implementation

[0012] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0013] A method for intelligent flight path planning of UAVs in multiple scenarios, step 1: task input and task type identification.

[0014] The mission input module includes tasks such as acquiring mission space parameters, setting flight altitude, and inputting mission target coordinates. It automatically determines the mission type based on the input structure. The specific steps are as follows: Step 1.1: Enter the location of the drone's nest (takeoff point) The system first inputs or retrieves the initial position coordinates of the drone, denoted as: (x, y) Step 1.2: Set the mission altitude for flight. The user sets the mission flight altitude H, and the theoretical position is (x, y, h[start]). Step 1.3: Input the coordinates of the task target (in multiple formats) Depending on the task type, the input coordinates may take the following forms: (1) Point task input: consists of multiple discrete target points, represented as follows: {(x1, y1), (x2, y2), ..., (x n , y n )} Line task input: A polyline segment consisting of the starting and ending points of the path, represented as: (x_start, y_start), (x_end, y_end) (3) Surface task input: consists of a set of boundary points that enclose a closed region, represented as: P = {(x1, y1), (x2, y2), ..., (x n , y n )} Step 1.4: Type Recognition and Task Tag Generation The system automatically determines the task type based on the coordinate input structure and boundary features: - If the input consists of several discrete points, it is identified as a point task; -If the input is a straight line, it represents two discrete points and is identified as a line task; - If the input is the boundary of the task region, it will be identified as a surface task.

[0015] The identification results are recorded in the form of task type labels and passed to the path planning module.

[0016] Step 2: Perform path planning calculation.

[0017] Genetic algorithms include the following steps: 1.1 Task Input Planning point locations: {(x1,y1),(x2,y2),...,(xn,yn)} 1.2 Line Task Input Planning point locations: {(xstart, ystart), (xend, yend)} 1.3 Input Structure of Genetic Algorithm Points={p1=(x1,y1),p2=(x2,y2),...,pn=(xn,yn)} 1.4 Path Arrangement C=(c1,c2,…,cn), Path arrangement: p3→p1→p4→p2→p5 1.5 Path Length Formula

[0018]

[0019] The final output of the genetic algorithm

[0020] Ox-plowing method: Calculation of ox-plowing-style flight paths;

[0021] Generate path coordinates for ox-plowing style;

[0022]

[0023] Final path sequence representation;

[0024] Greedy Algorithm: Power estimation model: When planning the route, the drone's power consumption can be estimated using a power consumption model. Power consumption It is determined by the flight distance D and the flight efficiency k (the amount of electricity consumed per unit distance):

[0025] Power consumption estimation in route planning: When planning a route, greedy algorithms or A* algorithms are typically used to select the optimal path. Energy consumption plays a significant role in route selection. At each step of the route selection process, the estimated energy consumption to reach the destination can be calculated.

[0026] Greedy path selection considering power consumption: In each step of the greedy algorithm, when selecting a path, power consumption and distance can be considered together to choose the path with the most estimated power consumption. For example, among multiple possible paths, the path with the lowest estimated power consumption can be selected:

[0027] Battery prediction models and battery swapping decisions: When planning a route, the estimated power consumption can be used to determine whether battery swapping is necessary. This also includes determining whether the total power consumption is below a certain threshold. The battery level at the start of the trip. Total power consumption, set threshold power consumption :

[0028] If this condition is met, then the battery swapping station (cell) needs to be included as part of the route planning.

[0029] Power restoration at the battery swapping station: If the route planning includes a battery swapping station (cell), the battery capacity after recovery can be considered:

[0030] Then, continue with the remaining tasks.

[0031] The above description discloses only preferred embodiments of the present invention and should not be construed as limiting the scope of the present invention. Therefore, equivalent variations made in accordance with the claims of the present invention are still within the scope of the present invention.

Claims

1. A method for intelligent flight path planning of unmanned aerial vehicles (UAVs) in multiple scenarios, characterized in that: Includes the following steps: Step 1, Establish task input and task type recognition: Set the UAV to fixed altitude flight mode and obtain the spatial information required for the task, including the takeoff point position, task point coordinate set and task area boundary. The system identifies the task type by analysis: if the input is a discrete point set, it is judged as a point task; if it is a polyline structure, it is judged as a line task; if it is a task area boundary, it is judged as a surface task. If the boundary width L is small, it is still judged as a line task. Step 2, Execute path planning calculation: The system automatically schedules the appropriate path planning algorithm model according to the identified task type: For point tasks and line tasks, the genetic algorithm module is called to perform population initialization, path encoding, crossover and mutation, and fitness function calculation (based on path length and energy consumption), and finally outputs the optimized optimal path sequence; For area tasks, the ox-plowing cover algorithm module is called to map the area to the minimum bounding rectangle, and generate equidistant grid routes according to the set flight interval, and achieve full coverage path output of the area through the "back and forth" strategy; Step 3, Output the path results and deploy the flight mission: The system outputs the optimized flight path results as a sequence of waypoints, including the GPS coordinates and execution order of each waypoint. The path can be directly sent to the UAV terminal to start the automatic flight mission. If it is a multi-aircraft collaborative operation mission, the system can divide the mission path into segments and deploy it to multiple flight platforms according to the regional division logic to improve the operation efficiency.

2. The intelligent flight path planning method for unmanned aerial vehicles (UAVs) in multiple scenarios according to claim 1, characterized in that: In step 2, the initialization phase of the genetic algorithm uses the optimal point set method to generate a portion of the population. The principle of the optimal point set is as follows: Let Gs be a unit cube in s-dimensional Euclidean space. If τ∈Gs, then the optimal point set... Represented as: ； The optimal point set is mapped to the search space, and the formula for calculating the point coordinates is: ； In the formula and Let represent the upper and lower bounds of the j-th dimension, respectively; Gs: a unit cube in s-dimensional Euclidean space, and τ is a point in the s-dimensional Euclidean unit cube; : Best point set, representing the set of coordinates of the k-th point generated by the best point set method; n: The total number of prime concentration points (population size); s: The dimension of the problem (i.e., the dimension of the chromosome or the dimension of the task point coordinates); : A positive integer constant coprime to n, used to generate a uniformly distributed sequence (s=1,2,…,s); : The actual coordinates of the i-th individual in the j-th dimension; : Upper bound of the j-th dimension; : Lower Bound of the j-th dimension; k: The index of the point, with a value range of 1≤k≤n.

3. The intelligent flight path planning method for UAVs in multiple scenarios according to claim 2, characterized in that: In step 2, the fitness function also incorporates the real-time requirements of the task scenario. The optimization objective is to minimize the total task completion time under multi-UAV collaborative operation. During the calculation process, tasks are assigned weights according to their real-time priority. Tasks with higher weights are prioritized in path planning, and the flight segment tasks of each UAV are dynamically adjusted to shorten the response time of high-priority tasks. The fitness function can be expressed as: ； in: m represents the number of drones; The time required to complete the task assignment for the i-th drone; Let be the real-time weight for task j; This is the difference between the actual completion time and the expected completion time of task j. This is the real-time penalty coefficient.

4. The intelligent flight path planning method for UAVs in multiple scenarios according to claim 3, characterized in that: Path planning also needs to be combined with a greedy algorithm to select the path with the most suitable estimated power. The greedy algorithm includes: establishing a power estimation model, power estimation in path planning, greedy path selection considering power, battery estimation model and battery swapping decision, and power recovery at the battery swapping station.

Images