基于粒子滤波的SLAM算法并行优化与实现

doi:10.12052/gdutxb.160055

广东工业大学学报 ›› 2017, Vol. 34 ›› Issue (02): 92-96.doi: 10.12052/gdutxb.160055

基于粒子滤波的SLAM算法并行优化与实现

朱福利, 曾碧, 曹军

广东工业大学计算机学院, 广东广州 210000

收稿日期:2016-04-01 出版日期:2017-03-09 发布日期:2017-03-09
通信作者: 曾碧(1963-),女,教授,主要研究方向为智能计算与智能机器人.E-mail:272070973@qq.com E-mail:272070973@qq.com
作者简介:朱福利(1990-),男,硕士研究生,主要研究方向为机器人技术.
基金资助:
国家自然科学基金资助项目（61202267）；广东省产学研合作专项资金资助项目（2014B090904080）

Parallel Optimization and Implementation of SLAM Algorithm Based on Particle Filter

Zhu Fu-li, Zeng Bi, Cao Jun

School of Computers, Guangdong University of Technology, Guangzhou 210000, China

Received:2016-04-01 Online:2017-03-09 Published:2017-03-09

摘要/Abstract

摘要：

基于粒子滤波的即时定位与地图构建（simultaneous localization and mapping，SLAM）算法，可在完全未知的环境进行即时的定位和地图构建.该算法使用粒子集表示定位位姿的概率分布情况，计算量与粒子集的规模成正比，在一定范围内，粒子的数量越多，算法的定位准确度和抗干扰能力越好，但在增加粒子数量的同时，将增加计算时间，从而导致定位延迟，造成移动机器人的定位误差.提出一种结合粒子滤波和SLAM算法特点的GPU并行优化的方法进行加速，从而减少计算带来的定位延迟和定位误差.通过实验，证明使用GPU并行计算的算法改进有明显效果.

关键词: 即时定位与地图构建, 粒子滤波, GPU并行计算

Abstract:

Simultaneous localization and mapping is a new type of mobile robot localization method, which can obtain data through the mobile robot's own sensors and simultaneous localization and map building in a completely unknown environment. Based on PF-SLAM algorithm, the probability distribution of the location pose is expressed by the particle set, and the calculated amount is proportional to the size of the particle set, and also the number of particles determines the algorithm's location accuracy and anti-jamming capability. At the same time, increasing the number of particles will increase the computing time, which will lead to the positioning delay and the positioning error of the mobile robot. A method is presented to improve the algorithm by using GPU parallel computing, which can reduce the calculation time, thereby to reduce the positioning error caused by positioning delay. Experimental results show that the improved algorithm of GPU parallel computing has a significant effect.

Key words: simultaneous localization and mapping, particle filter, GPU parallel computing

中图分类号:

TP301

朱福利, 曾碧, 曹军. 基于粒子滤波的SLAM算法并行优化与实现[J]. 广东工业大学学报, 2017, 34(02): 92-96.

Zhu Fu-li, Zeng Bi, Cao Jun. Parallel Optimization and Implementation of SLAM Algorithm Based on Particle Filter[J]. Journal of Guangdong University of Technology, 2017, 34(02): 92-96.

参考文献

[1] DURRANT-WHYTE H, BAILEY T. Simultaneous localization and mapping (SLAM):Part I The essential algorithms[J]. Robotics and Automation Magazine, 2006, 13(2):99-110.
[2] SMITH R, SELF M, CHEESEMAN P. Estimating uncertain spatial relationships in robotics[J]. Autonomous Robot Vehicles. 1990, 4:167-193.
[3] SHOWDONG H, GAMINI D. Convergence and consistency analysis for extended Kalman filter based SLAM[J]. IEEE Transactions on Robotics, 2007, 23(5):1036-1049.
[4] 宋宇, 李庆玲, 康轶非, 等. 平方根容积Rao-Blackwillised粒子滤波SLAM算法[J]. 自动化学报, 2014, 40(2):357-367.SONG Y, LI Q L, KANG Y F, et al. SLAM with square-root cubature rao-blackwillisedparticle filter[J]. Acta Automatica Sinica, 2014, 40(2):367-367.
[5] 王丁. 基于改进粒子滤波的检测前跟踪算法[D]. 成都:电子科技大学电子工程学院, 2012.
[6] 朱磊, 樊继壮, 赵杰, 等. 未知环境下的移动机器人SLAM方法[J]. 华中科技大学学报(自然科学版), 2011, 39(7):9-13.ZHU L, FAN J Z, ZHAO J. SLAM method for mobile robot in unknown environment[J]. Huazhong University of Technology (Natural Science Edition), 2011, 39(7):9-13.
[7] DIETER FOX. Adapting the sample size in particle filters through KLD-sampling[J]. TheInternational Journal of Robotics Research, 2003, 22:985-1003.
[8] 卢风顺, 宋君强, 银福康, 等. CPU/GPU协同并行计算研究综述[J]. 计算机科学, 2011, 28(3):5-9.LU F S, SONH J Q, YIN F K, et al. Survey of CPU/GPU synergetic parallel computing[J]. Computer Science, 2011, 28(3):5-9.
[9] 徐李超, 张祺. 一种仿人机器人步态优化的新方法研究[J]. 广东工业大学学报, 2012, 29(1):50-54.XU L C, ZHANG Q. Gait Optimizing of humanoid robots using a new method[J]. Journal of Guangdong University of Technology, 2012, 29(1):50-54.
[10] 王法胜, 鲁明羽, 赵清杰, 等. 粒子滤波算法[J]. 计算机学报, 2014, 37(8):1679-1694.WANG F S, LU M Y, ZHAO Q J, et al. Particle filtering algorithm[J]. Chinese Journal of Computer, 2014, 37(8):1679-1694.
[11] 于金霞, 汤永利, 许景. 一种改进的自适应优化粒子滤波算法研究[J]. 小型微型计算机系统, 2013, 34(6):1446-1450.YU J X, TANG Y L, XU J. Research on an improved particle filter algorithm based on adaptive optimization mechanism[J]. Journal of Chinese Computer Systems, 2013, 34(6):1446-1450.
[12] 左军毅, 张怡哲, 梁彦. 自适应不完全重采样粒子滤波器[J]. 自动化学报, 2012, 38(4):647-652.ZUO J D, ZHANG Y Z, LIANG Y, Particle filter based on adaptive part resampling[J]. Acta Automatica Sinica, 2012, 38(4):647-652.
[13] 冯驰, 王萌. 粒子滤波器重采样算法的分析与比较[J]. 系统仿真学报, 2009, 21(4):1101-1110.FENG C, WANG M. Analysis and comparison of resampling algorithms in particle filter[J]. Journal of Systems Simulation, 2009, 21(4):1101-1110.
[14] 刘洞波. 移动机器人粒子滤波定位与地图创建方法研究[D]. 长沙:湖南大学电气与信息工程学院, 2013.
[15] 郭睿冉, 宋建新. 图像压缩感知的基于GPU的并行重构算法研究[J]. 电视技术, 2014, 38(11):15-19.GUO R R, SONG J X. Research and implementation of parallel signal reconstruction about image compressed sensing based on GPU[J]. Digital Video, 2014, 38(11):15-19.
[16] 刘燕龙, 原玲, 姜文超, 基于Calculix的船舶疲劳强度并行计算方法研究与应用[J]. 广东工业大学学报, 2015, 32(4):77-82.LIU Y L, YUAN L, JIANG W C. Parallel mechanism research and application of calculix in ship fatigue analysis environment[J]. Journal of Guangdong University of Technology, 2015, 32(4):77-82.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

基于粒子滤波的SLAM算法并行优化与实现

Parallel Optimization and Implementation of SLAM Algorithm Based on Particle Filter

HTML

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 2

Metrics

本文评价

推荐阅读 0

[1]	熊武, 刘义. 粒子滤波算法在BDS高铁铁轨静态形变监测中的应用研究[J]. 广东工业大学学报, 2022, 39(04): 66-72.
[2]	吴用, 万频, 王永华, 梁颋, 卢强. 认知无线网络小型移动主用户的定位算法研究[J]. 广东工业大学学报, 2017, 34(01): 60-64.