基于深度学习的无人机智能避障系统设计与实现  

Design and Implementation of Intelligent Obstacle Avoidance System for UAV Based on Deep Learning

作　　者：王义和[1] WANG Yihe(Ningxia Vocational and Technical College of Industry and Commerce,Yinchuan Ningxia 750021,China)

机构地区：[1]宁夏工商职业技术学院,宁夏银川750021

出　　处：《信息与电脑》2024年第14期108-110,共3页Information & Computer

基　　金：校级横向课题——小型四旋翼无人机智能投放技术改造。

摘　　要：随着无人机(Unmanned Aerial Vehicle,UAV)技术的发展,智能避障成了提高其自主飞行能力的关键。本文提出了一种基于深度学习的无人机智能避障系统设计,该系统包括智能路径规划、障碍物实时检测和自主动态避障等功能,能够实时检测和识别障碍物,实现快速转向避障和减速悬停避障。在硬件设计方面,系统选用了高性能的传感器模块、主控芯片和动力系统。测试结果表明,该系统具有良好的功能和性能表现,能够有效提高无人机的避障能力。With the development of unmanned aerial vehicle(UAV)technology,intelligent obstacle avoidance has become the key to enhance its autonomous flight capability.This paper proposes the design of a UAV intelligent obstacle avoidance system based on deep learning,which includes intelligent path planning,real-time detection of obstacles and autonomous dynamic obstacle avoidance,etc.The system is able to detect and identify obstacles in real time,and realise rapid steering obstacle avoidance and deceleration hovering obstacle avoidance.In terms of hardware design,the system selects highperformance sensor module,main control chip and power system.The test results show that the system has good function and performance performance,and can effectively improve the obstacle avoidance ability of the UAV.

关 键 词：深度学习 无人机 智能避障 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]