CN-116126026-B - Unmanned aerial vehicle three-dimensional path planning method based on risk map

CN116126026BCN 116126026 BCN116126026 BCN 116126026BCN-116126026-B

Abstract

The invention provides an unmanned aerial vehicle three-dimensional path planning method based on a risk map, which comprises the steps of discretizing a low-altitude airspace based on a rasterization technology, obtaining urban airspace building data, planning an unmanned aerial vehicle geofence, building a risk assessment model according to the risk of the unmanned aerial vehicle to the ground, generating a probability-based risk map, building a path planning model based on the risk map, and carrying out unmanned aerial vehicle three-dimensional path planning under a complex environment through a trip point search algorithm. The urban low-altitude airspace is used as a research object, and the unmanned aerial vehicle autonomous path planning at a strategic stage can be performed under a complex urban environment in which multiple factors coexist by combining a risk map and a path planning algorithm, so that reasonable static collision prevention is realized while global optimization is considered, and the potential casualty risk of unmanned aerial vehicle operation on ground pedestrians is effectively reduced, so that unmanned aerial vehicle safe operation is realized.

Inventors

ZHOU LONG
Han Site
TANG MIAO
WEI JINJUN
XIONG WEIDONG
MA CHAO

Assignees

南京智慧航空研究院有限公司

Dates

Publication Date: 20260512
Application Date: 20230306

Claims (7)

1. The unmanned aerial vehicle three-dimensional path planning method based on the risk map is characterized by comprising the following steps of: step one, discretizing a low-altitude airspace; Acquiring urban airspace building data and planning a unmanned aerial vehicle to forbidden to enter a geofence; step three, establishing a risk assessment model according to the ground risk of the unmanned aerial vehicle; generating a risk map based on the casualty probability; Establishing a path planning model based on a risk map; step six, planning a three-dimensional path of the unmanned aerial vehicle; the risk assessment model in the third step is as follows: Judging the area of the area affected by the collision on the ground based on the model of the unmanned aerial vehicle descent type; The unmanned aerial vehicle accident causes casualties to be respectively three conditions that the unmanned aerial vehicle falls out of control and collides with ground personnel to cause death, and the following formula is adopted: Wherein, therein ; Refers to the geographic location The population density at the location(s), Refers to the area of the ground affected by the impact, Representing geographical location The shading coefficient at the point of the frame, , Refers to the expected value of the value to be obtained, Representing a geographic location of impact of a drone Kinetic energy at the location; Is the probability of casualties; the probability of occurrence of the uncontrolled descent event of the unmanned aerial vehicle is determined; probability of collision of unmanned aerial vehicle with ground personnel; the probability of death caused by falling of the unmanned aerial vehicle is increased; Is that When the unmanned aerial vehicle falls, the collision kinetic energy threshold value which causes death is defined to be 34J according to the death limit; Is that When the unmanned aerial vehicle falls, the collision kinetic energy threshold value with the death probability of 50% is 10 6 J; The following formula is used to define the casualties: Wherein, the Is the probability of casualties; the probability of casualties corresponding to the occurrence of ballistic drop; To the probability of casualties corresponding to uncontrolled taxiing.
2. The unmanned aerial vehicle three-dimensional path planning method based on the risk map of claim 1, wherein in the first step, the low-altitude airspace discretization adopts a rasterization technology.
3. The method for planning a three-dimensional path of an unmanned aerial vehicle based on a risk map according to claim 1, wherein the urban airspace building data in the second step at least comprises one of static building, unmanned aerial vehicle flight performance, airspace information and population density information; and based on the acquired urban airspace building data, the unmanned aerial vehicle forbidden geofence is marked on the position of the forbidden flying area, the temporary take-off and landing point and the obstacle.
4. The unmanned aerial vehicle three-dimensional path planning method based on the risk map according to claim 1, wherein the model of the unmanned aerial vehicle descent type comprises a trajectory descent mode and a runaway taxiing descent mode, and the impact area of the two descent modes is respectively: ; ; Wherein, the , , The radius of the outer ball is the size of the unmanned aerial vehicle, For the average width of the human body, Taking 1.7m for the average height of human beings, The horizontal distance moved after the unmanned aerial vehicle falls to the height of the pedestrian; the glide angle is the angle formed by the velocity vector and the ground.
5. The three-dimensional path planning method of an unmanned aerial vehicle based on a risk map according to claim 1, wherein the risk map generated in the fourth step adopts the following method: counting probability values of the risk assessment model in the third step to generate an casualty probability map; generating a population density layer, an obstacle layer and a shielding factor layer of a corresponding airspace according to the obstacle data, the population density data and the airspace information; and merging the casualty probability map with the population density layer, the barrier layer and the shielding factor layer to obtain a risk map.
6. The three-dimensional path planning method of unmanned aerial vehicle based on risk map according to claim 1, wherein the path planning model in the fifth step comprises: designing heuristic functions and constraint conditions based on grid technology; combining element values of the risk map into heuristic functions, and updating constraint conditions; the path planning model is as follows: Wherein, the , Is a weight coefficient; Is unmanned plane An average value of the probability of casualties per flight hour in the current environment; The probability of casualties per flight hour generated for pedestrians on the ground at the current position; The reward function value corresponding to the casualty probability; is a heuristic function; the actual cost from the initial position to the current position; Estimating a cost for a remaining path from the current location to the target location; Is the coordinates of the previous position; coordinates of the latter position; The coordinates of the current jump point; Is the target position coordinate; Is a reward parameter.
7. The three-dimensional path planning method of the unmanned aerial vehicle based on the risk map according to claim 1, wherein the three-dimensional path planning of the unmanned aerial vehicle in the sixth step adopts the following method: Performing three-dimensional path planning by adopting a jump point searching algorithm to obtain an initial reference path; the method of dividing various static barriers and limited areas into unmanned aerial vehicle geofences is used for replacing a path diagonalization method, and a reference path is optimized to obtain a final unmanned aerial vehicle three-dimensional path.

Description

Unmanned aerial vehicle three-dimensional path planning method based on risk map Technical Field The invention relates to the field of unmanned aerial vehicles, in particular to an unmanned aerial vehicle three-dimensional path planning method based on a risk map. Background Along with the continuous development of unmanned aerial vehicle technology and the rapid promotion of application prospect, unmanned aerial vehicle's flexibility and high mobility have fully been embodied in urban airspace, and the technological development makes urban aerial maneuver UAM (Urban Air Mobility) become possible. However, in the environment and the complex urban low-altitude airspace, such as CBD (Central Business District) area, how to plan a safe, collision-free and as short as possible flight path in a strategic stage so as to ensure the safe operation of the unmanned aerial vehicle is one of the important problems to be solved; At present, the unmanned aerial vehicle field researches on urban airspace management are focused on conflict resolution and scene application, safe operation is realized in an idealized scene, unmanned aerial vehicle path planning is one of research emphasis, and mainly applied algorithms can be roughly divided into two types, namely a graph-based search method and an optimal control-based method. The searching method of the graph often faces the dimension disaster problem, while the method based on the optimal control usually depends on a numerical solution and faces the minimum point problem. In 2011, adolf F.M and Andert.F published Rapid Multi-query path planningfor A VERTICAL TAKE-off AND LANDING unmanned AERIAL VEHICLE, and proposed an online multi-query path planning method, which combines a motion planning method based on sampling with path searching at any time, and realizes overall maneuver in unpredictable obstacle change scenes so as to improve the operation autonomy of the unmanned aerial vehicle. In 2013, wang published "unmanned aerial vehicle path planning algorithm based on PH curve", and herein a method for directly using PH with continuous curvature for unmanned aerial vehicle path planning is provided, and by utilizing continuous curvature, smooth curve and rational characteristics of PH curve, the advantages of genetic algorithm and simulated annealing algorithm are combined, global searching capability is enhanced, and basis is provided for unmanned aerial vehicle flight control. In 2020, li Xianjiang published "improved design of ant colony algorithm and application in flight path planning", a new ant colony algorithm is designed, which can avoid the problem that the traditional ant colony algorithm is easy to mature and falls into local optimum; It is not difficult to see that the above research on unmanned aerial vehicle path planning focuses on an ideal environment, and huge risks caused by complexity of obstacle shapes and distribution on unmanned aerial vehicle path planning are not considered, so that collision can be caused by deviation of unmanned aerial vehicle paths and actual operation scenes. In addition, the conventional three-dimensional path planning algorithm does not consider the casualty risk of the ground pedestrians caused by the faults of the unmanned aerial vehicle, and cannot well meet the requirement of safe operation. Disclosure of Invention Aiming at the technical defects, the invention aims to provide an unmanned aerial vehicle three-dimensional path planning method based on a risk map. The invention adopts the following technical scheme: a three-dimensional path planning method of an unmanned aerial vehicle based on a risk map comprises the following steps: step one, discretizing a low-altitude airspace; Acquiring urban airspace building data and planning a unmanned aerial vehicle to forbidden to enter a geofence; step three, establishing a risk assessment model according to the ground risk of the unmanned aerial vehicle; generating a risk map based on the casualty probability; Establishing a path planning model based on a risk map; And step six, carrying out three-dimensional path planning of the unmanned aerial vehicle. And in the first step, the low-altitude airspace discretization adopts a rasterization technology. The urban airspace building data in the second step at least comprises one of static building, unmanned aerial vehicle flight performance, airspace information and population density information; and based on the acquired urban airspace building data, the unmanned aerial vehicle forbidden geofence is marked on the position of the forbidden flying area, the temporary take-off and landing point and the obstacle. The risk assessment model in the third step is as follows: Judging the area of the area affected by the collision on the ground based on the model of the unmanned aerial vehicle descent type; The unmanned aerial vehicle accident causes casualties to be respectively three conditions that the unmanned aerial vehicle falls out