多目视觉与激光组合导航AGV精确定位技术研究

发布时间：2018-07-01 18:25

本文选题：自动导引小车 + 组合导航　；参考：《仪器仪表学报》2017年11期

【摘要】：为进一步提高自动导引车(AGV)的定位精度,提出了一种多目视觉与激光组合导航的精确定位方法。该方法首先提出了一种基于双目视觉实时测量的AGV位姿控制技术,通过多目视觉系统来识别导引线侧的多个圆形标识点完成前瞻预判与精准定姿。采用基于代数距离方差校验的改进型最小二乘拟合椭圆算法来确定各标识中心坐标,结合激光扫描与视觉定位信息,采用无迹卡尔曼滤波算法进行多传感器信息融合,最终实现精确定位。实验结果表明:使用该方法后定位精度明显得到提高,控制曲线更为平滑,定位鲁棒性更好,姿态调整精度可达±0.5°,停车定位精度可达±1 mm。
[Abstract]:In order to further improve the positioning accuracy of automatic guided vehicle (AGV), an accurate positioning method of integrated vision and laser navigation was proposed. In this method, an AGV position and attitude control technique based on binocular vision real-time measurement is proposed, which can identify multiple circular marking points on the side of the guidance line through a multi-eye vision system to achieve forward prediction and accurate attitude determination. An improved least square fitting ellipse algorithm based on algebraic distance variance check is used to determine the coordinates of each identification center. Combined with laser scanning and visual positioning information, the unscented Kalman filter algorithm is used for multi-sensor information fusion. Finally, accurate positioning is realized. The experimental results show that the positioning accuracy is obviously improved, the control curve is smoother, the positioning robustness is better, the attitude adjustment accuracy is 卤0.5 掳, and the parking positioning accuracy is 卤1 mm.
【作者单位】：南京航空航天大学机电学院;江苏省精密与微细制造技术重点实验室;
【基金】：江苏省重点研发(BE2016004-3) 中国博士后科学基金(2015M580421) 汽车装配智能输送装备关键技术研发(BY2015003-11) 江苏省普通高校研究生科研创新计划(SJZZ16_0058)项目资助
【分类号】：TP23;TP391.41

【相似文献】