基于动态和静态环境对象观测一致性约束的移动机器人多传感器标定优化方法
发布时间：2018-10-05 19:13
【摘要】：为了解决机器人未知环境导航过程中的多源、异构传感器空间一致性观测问题,提出了基于动态和静态环境对象观测一致性约束的摄像机与激光测距传感器联合标定优化方法。利用协方差交集方法实现运动目标图像平面方向状态融合,同时采用卡尔曼滤波和概率数据关联滤波实现一对一和一对多信息源的静态角点特征图像平面方向状态融合;在此基础上,利用动态和静态物体融合前与融合后状态误差构造优化目标函数,并利用非线性优化方法实现标定参数优化。实验结果表明,该设计方法能够提高多传感器环境观测的一致性水平,验证了该方法的有效性。
[Abstract]:In order to solve the problem of spatial consistency observation of multi-source and heterogeneous sensors during robot navigation in unknown environment, a joint calibration optimization method of camera and laser rangefinder based on observation consistency constraints of dynamic and static environment objects is proposed. The covariance intersection method is used to realize the plane direction state fusion of moving target image. At the same time, Kalman filter and probabilistic data correlation filter are used to realize the plane direction state fusion of one-to-one and one-to-many information sources. On this basis, the optimization objective function is constructed by using the state errors before and after the fusion of dynamic and static objects, and the calibration parameters are optimized by nonlinear optimization method. The experimental results show that the design method can improve the consistency of multi-sensor environmental observation and verify the effectiveness of the method.
【作者单位】： 火箭军工程大学;
【基金】：国家自然科学基金项目(61503389) 陕西省自然科学基金项目(2016JM6061)
【分类号】：TP212;TP242
本文编号：2254554
[Abstract]:In order to solve the problem of spatial consistency observation of multi-source and heterogeneous sensors during robot navigation in unknown environment, a joint calibration optimization method of camera and laser rangefinder based on observation consistency constraints of dynamic and static environment objects is proposed. The covariance intersection method is used to realize the plane direction state fusion of moving target image. At the same time, Kalman filter and probabilistic data correlation filter are used to realize the plane direction state fusion of one-to-one and one-to-many information sources. On this basis, the optimization objective function is constructed by using the state errors before and after the fusion of dynamic and static objects, and the calibration parameters are optimized by nonlinear optimization method. The experimental results show that the design method can improve the consistency of multi-sensor environmental observation and verify the effectiveness of the method.
【作者单位】： 火箭军工程大学;
【基金】：国家自然科学基金项目(61503389) 陕西省自然科学基金项目(2016JM6061)
【分类号】：TP212;TP242
【相似文献】
相关期刊论文 前1条
1 刘蓉;陈晓红;;新的大群体一致性学习修正决策方法[J];系统工程与电子技术;2008年05期
，本文编号：2254554
本文链接：https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2254554.html
