基于激光扫描的机器人运动目标跟踪技术研究  

Research on Robot Moving Target Tracking Technology Based on Laser Scanning

作　　者：黄东 陈乐庚[2] HUANG Dong;CHEN Le-geng(School of Electronic Engineering,Guilin Institute of Information Technology,Guilin Guangxi 541004,China;Guilin University of electronic Technology,School of Electronle Engineering and Automation,Guilin Guangxi 541004,China)

机构地区：[1]桂林信息科技学院电子工程学院,广西桂林541004 [2]桂林电子科技大学电子工程与自动化学院,广西桂林541004

出　　处：《计算机仿真》2022年第6期439-443,共5页Computer Simulation

基　　金：桂林电子科技大学信息科技学院2020年科研项目(XJ202011)。

摘　　要：针对传统机器人运动目标跟踪方法中存在的轨迹跟踪效果较差、跟踪效率低等问题,提出激光扫描的机器人运动目标跟踪方法。分析激光扫描中噪声干扰的主要来源,并设计相应滤波方法降低噪声干扰;利用机器人三种坐标结构构建坐标系,并构建机器人运动模型,确定运动参数约束条件;为降低系统运算负担,剔除不确定因素数据,建立检测视窗;评估障碍物的运动状态,检测运动目标物体,利用运动补偿方式提高检测精准度;在卡尔曼滤波基础上,加入高斯干扰白噪声,获取运动目标的跟踪状态矩阵。仿真结果表明:所提方法对机器人运行轨迹追踪的效果较好,且跟踪效率较高,可较好地适应复杂多变的环境。In traditional robot moving target tracking methods, the problems such as poor tracking effect and low tracking efficiency often occur. Therefore, a method of tracking robot moving targets based on laser scanning was proposed. Firstly, the main source of noise interference in laser scanning was analyzed. Correspondingly, some filtering methods were designed to reduce noise interference. Moreover, three kinds of coordinate structures of the robot were used to construct the coordinate system and the moving model of the robot and thus to determine the constraint condition of the motion parameters. In order to reduce the computational burden of the system, the data about uncertain factors were eliminated, and the detection window was built. Furthermore, the motion state of obstacles was evaluated. In the meantime, moving targets were detected. Finally, the method of motion compensation was used to improve the detection accuracy. On the basis of the Kalman filter, Gaussian white noise was added to obtain the matrix of the tracking state of the moving target. Simulation results show that the proposed method has a good effect on tracking robots and high tracking efficiency, which can make good adaptability to complex and changeable environments.

关 键 词：激光扫描 机器人 运动目标 跟踪技术 卡尔曼滤波 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]