RRT (rapidly-exploring random tree) route optimizing method of unmanned aerial vehicle based on greedy algorithm

Abstract

The invention belongs to the technical field of application of unmanned aerial vehicles, and discloses an RRT (rapidly-exploring random tree) route optimizing method of an unmanned aerial vehicle based on a greedy algorithm. The RRT route optimizing method comprises the following steps of using qinit as the current point qi, and initializing n to be 0; performing collision detection on a Q focus point qm-n since qi (n is [0,1,2 until to (m-i-1)]), and uniformly selecting points on the connecting line of qi and qm-n at certain intervals; judging whether the existence point is positioned on theobstacle or outside a map; when the existence point exists, turning to step 4; or else, turning to step 5; when the obstacle exists, namely that the connecting line passes through the fly-prohibitingarea, and n is equal to n+1, judging whether qi is qm-n-1 or not; when qi is qm-n-1, enabling i to be equal to i+1; or else, returning back; when the obstacle does not exist, connecting qi and qm-n, enabling qi to be equal to qm-n, and turning to step 6; judging whether i is equal to m or not; when i equal to m, finishing the program. The RRT route optimizing method has the advantages that when the RRT algorithm route of the unmanned aerial vehicle is planned, the route is reduced, the greedy concept is utilized, the result is the current optimum result, the number of route points is reduced,and the route is greatly optimized; the certain engineering value is realized on the application of route planning of the unmanned aerial vehicle or other robots.

Description

technical field [0001] The invention belongs to the technical field of unmanned aerial vehicles, and in particular relates to a method for optimizing RRT paths of unmanned aerial vehicles based on a greedy algorithm. Background technique [0002] At present, the existing technologies commonly used in the industry are as follows: [0003] The RRT algorithm is called the rapid expansion random tree method, which is a sampling-based method that performs path planning by generating random points in the feasible space. The rapid expansion random tree method first randomly samples the space, and then uses the starting point as the root node of the expansion tree to search the surrounding nodes. Select feasible nodes as path points through collision detection and add them to the root node list. By continuously guiding the search tree to the blank area, a random tree roadmap with paths is constructed, until the extended tree reaches the end point range and ends the extended tree, ...

AI Technical Summary

Problems solved by technology

Disadvantages of RRT: due to the blindness and diversity of the outward search from the node, the routes generated by each route are different; at the same time, the planned routes are very tortuous and not optimal, most of them are not the shortest route, and random exploration is also The planning time becomes longer; when planning, you need to know the global map in advance, which is actually offline planning, and you must explore the global map before planning

[0005] (1) In the prior art, the path generated by the RRT algorithm is more tortuous, the generated path points are random, and the path is not optimal, resulting in long planning time, wasted time, and reduced work efficiency

[0006] (2) In the prior art, the UAV cannot turn smoothly and fly continuously at a certain speed on the path, so that the work does not have continuity and reduces the use effect

When performing missions in complex spaces, if you fly according to the path planned by the RRT method, there will be stop-and-go situations, and you cannot fly continuously

This makes the power consumption too fast, reducing the effective time and speed of performing tasks

Images