摘要
为了使智能车在室外未知大范围环境中实现真正的自主导航,采用激光雷达传感器获取车辆周围环境信息,基于自适应渐消扩展卡尔曼滤波（AFEKF）和改进的快速联合兼容关联方法设计了一套同时定位与建图（SLAM）系统,简称为AFEKF-SLAM。首先,采用一种基于局部关联策略和聚类分组策略改进的快速联合兼容关联方法解决SLAM中的数据关联问题,为状态估计提供准确的量测-特征匹配关系;其次,采用自适应渐消扩展卡尔曼滤波对车辆状态和环境路标的位置进行估计。基于＂Victoria dataset＂的仿真实验表明,设计的SLAM系统实时性强,获得的定位和建图结果准确,能够为智能车在大范围环境自主导航提供可靠的保障。
In order to make the intelligent vehicle achieve autonomous navigation in a large outdoor environment,laser radar sensors are used to obtain the environmental information of around vehicle. A SLAM（simultaneous localization and mapping） system,called AFEKF-SLAM,is designed based on the adaptive fading extended Kalman filter（AFEKF） and an improved fast joint compatible association method. Firstly,an improved joint compatible association method based on local association strategy,and clustering strategy is applied to solve the problem of data association in SLAM to provide accurate matching relationship of measurement-feature for state estimation. Secondly,the adaptive fading extended Kalman filter is used to estimate vehicle state and the position of environmental features. Simulation experiments based on ＂Victoria dataset＂shows that the designed SLAM system has strong real-time performance and accurate location and mapping results. It can provide a reliable guarantee for the autonomous navigation of the intelligent vehicle in a wide range of environment.
作者
刘丹
段建民
孟晓燕
LIU Dan;DUAN Jianmin;MENG Xiaoyan(Faeuhy of Information Technology,Beijing University of Technology,Beijing 100124,China)
出处
《激光杂志》
北大核心
2018年第7期76-82,共7页
Laser Journal
基金
北京市教委基金项目(No.JJ002790200802)
