摘要
室内机器人由于技术限制并不能很好地完成复杂的任务,如放在相对复杂的环境中往往需要路线规划、避障和导航等功能,选择单一的里程计定位会因为计算产生误差积累,长时间使用定位效果不理想。单一的激光雷达会因为实际运动过程中产生的机器抖动无法确定自身位置,导致地图构建效果不理想。针对这一现状,提出基于ROS操作系统整合2D激光雷达+MPU6050作为里程计定位的自动导航技术,通过激光雷达标定里程计数据,减少里程计误差,实现在全局地图中选择目标点后进行路径规划并据此运动。在相对封闭环境中帮助机器人精确感知周围环境变化同时确定自身位置,为服务类软件提供可靠且便利的硬件平台,得以实现点对点精确服务。从机器人硬件方案、激光雷达数据+里程计数据处理的基本原理、基于GMapping的SLAM算法设计和机器人自动导航测试几个角度进行分析。
Due to technical limitations,eye-catching indoor robots cannot complete complex tasks well,such as route planning,obstacle avoidance,navigation and other functions are often required in a relatively complex environment.Choosing a single odometer for positioning will cause error accumulation due to calculation,and the positioning effect is not ideal for a long time.A single LiDAR cannot determine its own position due to machine jitter in the process of actual motion,resulting in unsatisfactory map construction effect.In view of this situation,this paper proposes an automatic navigation technology based on ROS operating system integrating 2D LiDAR+MPU6050 as odometer positioning.By calibrating odometer data with LiDAR,odometer errors can be reduced,and path planning and movement can be realized after target points are selected in the global map.To help robots accurately perceive the changes of surrounding environment and determine their own position in a relatively closed environment,so as to provide a reliable and convenient hardware platform for service software,so as to realize point-to-point accurate service.This paper analyzes robot hardware scheme,laser radar data+odometer data processing basic principle,GMapping based SLAM algorithm design and robot automatic navigation test.
出处
《工业控制计算机》
2023年第10期34-35,39,共3页
Industrial Control Computer
基金
广东省科技创新战略专项资金“攀登计划”:智能交互指引机器人(pdjh2022b0630)。
