Unmanned plane locus planning method based on random sampling in narrow space

Abstract

The invention relates to an unmanned plane locus planning method based on random sampling in a narrow space, which belongs to the unmanned plane locus planning field. The method comprises the steps of obtaining unmanned plane flight environment information; taking an original state of an unmanned plane, an object state and a 3D model of a flight environment as initial parameters of a locus plan; conducting road sign points sampling based on the 3D model of the flight environment; storing the road sign points in road sign point lists corresponding to a narrow space N and an open area U respectively; employing multiple rapidly-exploring random trees and connecting multiple RRT locus trees in pairs to form a fused RRT locus tree; then connecting the fused RRT locus tree with the original state and the object state of the unmanned plane, thereby generating a complete locus that connects the original state and the object state of the unmanned plane. According to the invention, the calculating efficiency of the unmanned plane locus planning in a narrow space can be effectively raised.

Description

Technical field [0001] The invention relates to an unmanned aerial vehicle trajectory planning method based on random sampling in a narrow space, and belongs to the field of unmanned aerial vehicle trajectory planning. Background technique [0002] The purpose of UAV trajectory planning is to find a continuous flight from the initial state of the UAV to the mission goal in the environment that satisfies certain optimal performance indicators and constraints according to the mission objectives, environmental information and constraints of the UAV Track. Trajectory planning is an important issue in the field of UAV navigation and control, and it is also one of the key technologies to achieve autonomous flight of UAVs. [0003] Unstructured and complex 3D environments are widely used in many practical applications such as drone environment exploration, target search and monitoring. Many mission scenarios in reality (such as complex indoor spaces, collapsed buildings, mines, etc.) be...

AI Technical Summary

Problems solved by technology

[0004] Traditional trajectory planning methods based on random sampling (such as probabilistic path map method (PRM) and rapidly expanding random tree (RRT), etc.) are widely used in solving planning problems with dynamic / kinematic constraints and high-dimensional spaces. However, since this type of method generally adopts a uniform sampling strategy in space, it is often unable to effectively identify free space from complex structures in complex 3D environments with narrow spaces; and it is difficult to realize trajectory tracking in complex spaces with poor connectivity. Extended, the overall computational efficiency of trajectory planning is low

Images