摘要
巡检机器人行进中的多传感数据融合过程可能会受到环境因素的影响,导致二维栅格地图中的坐标不准确,无法实现有效避障控制,为此,提出基于自适应RRT与SLAM视觉的室内巡检机器人避障控制方法。标定了深度相机和激光雷达传感器的外参矩阵,采用SLAM视觉方法组建二维栅格地图;基于自适应RRT算法设置目标节点采样率,自适应生成初始巡检路径,动态调整步长,结合SLAM视觉从传感器数据中识别出障碍物,并映射到栅格地图中进行标记,基于局部地图,选择窗口内最优轨迹,实现避障控制。实验结果表明,应用所提方法进行避障控制下,路径长度较短,节点数量较小,避障控制偏差仅为0.02%,拐点数最少,可精准识别障碍物并进行有效避障,应用效果较好。
The multi sensor data fusion process during the movement of inspection robots may be affected by environmental factors,resulting in inaccurate coordinates in two dimensional grid maps and ineffective obstacle avoidance control.Therefore,an indoor inspection robot obstacle avoidance control method based on adaptive RRT and SLAM vision is proposed.The extrinsic matrices of the depth camera and LiDAR sensor are calibrated,and a two dimensional raster map using SLAM visual method is constructed.Based on the adaptive RRT algorithm,the target node sampling rate is set.The adaptive initial inspection path is generated and the step size is dynamically adjusted.Combined with the SLAM vision,obstacles are identified from sensor data and mapped to a grid map for marking.Based on the local map,the optimal trajectory is selected within the window to achieve obstacle avoidance control.The experimental results show that by using the method proposed in this paper to conduct the obstacle avoidance control,the path length is relatively short,the number of nodes is small,the deviation of obstacle avoidance control is only 0.02%,and the number of turning points is minimal.It can accurately identify obstacles and effectively avoid them,and the application effect is good.
作者
周彦飞
赵雨
秦鹏
王景峰
舒容
ZHOU Yanfei;ZHAO Yu;QIN Peng;WANG Jingfeng;SHU Rong(Dongying Power Supply Company,State Grid Shandong Electric Power Company,Dongying 257000,China)
出处
《机械与电子》
2025年第6期55-60,67,共7页
Machinery & Electronics
基金
国网山东省电力公司东营供电公司科技项目(520616240005)。
