ISSN 1000-1239 CN 11-1777/TP

计算机研究与发展 ›› 2017, Vol. 54 ›› Issue (11): 2434-2444.doi: 10.7544/issn1000-1239.2017.20170309

所属专题: 2017车联网关键技术与应用研究专题

• 网络技术 • 上一篇    下一篇

不确定环境下移动对象自适应轨迹预测方法

夏卓群1,2,3,胡珍珍1,2,罗君鹏1,2,陈月月3   

  1. 1(综合交通运输大数据智能处理湖南省重点实验室(长沙理工大学) 长沙 410114); 2(长沙理工大学计算机与通信工程学院 长沙 410114); 3(国防科学技术大学计算机学院 长沙 410114) (xiazhuoqun@sina.com)
  • 出版日期: 2017-11-01
  • 基金资助: 
    国家自然科学基金项目(61572514);湖南省自然科学基金项目(14JJ7043);湖南省交通厅科技进步与创新项目(201405)

Adaptive Trajectory Prediction for Moving Objects in Uncertain Environment

Xia Zhuoqun1,2,3, Hu Zhenzhen1,2, Luo Junpeng1,2, Chen Yueyue3   

  1. 1(Hunan Provincial Key Laboratory of Intelligent Processing of Big Data on Transportation (Changsha University of Science and Technology), Changsha 410114); 2(School of Computer and Communication Engineering, Changsha University of Science and Technology, Changsha 410114); 3(College of Computer, National University of Defense Technology, Changsha 410114)
  • Online: 2017-11-01

摘要: 已有的轨迹预测方法难以对移动对象运动轨迹进行准确地描述,尤其在复杂且不确定的车载自组织网络(vehicular ad hoc network)(也称车联网)环境中.为了解决这一问题,提出基于变分高斯混合模型(variational Gaussian mixture model, VGMM)的环境自适应轨迹预测方法ESATP(environment self-adaptive prediction method based on VGMM).首先,在传统高斯混合模型的基础上使用变分贝叶斯推理近似方法处理混合高斯分布;其次设计变分贝叶斯期望最大化算法学习计算高斯混合模型参数,有效运用参数先验信息得到更高精度预测模型;最后,针对输入轨迹数据特征,使用参数自适应选择算法自动调节参数组合,灵活调整混合高斯分量的个数和轨迹段大小.实验结果表明:所提方法在实验中表现出较高的预测准确性,可应用于车辆移动定位产品中.

关键词: 环境自适应, 变分高斯混合模型, 参数自适应选择算法, 轨迹预测

Abstract: The existing methods for trajectory prediction are difficult to describe the trajectory of moving objects in complex and uncertain environment accurately. In order to solve this problem, this paper proposes an self-adaptive trajectory prediction method for moving objects based on variation Gaussian mixture model (VGMM) in dynamic environment (ESATP). Firstly, based on the traditional mixture Gaussian model, we use the approximate variational Bayesian inference method to process the mixture Gaussian distribution in model training procedure. Secondly, variational Bayesian expectation maximization iterative is used to learn the model parameters and prior information is used to get a more precise prediction model. This algorithm can take a priory information. Finally, for the input trajectories, parameter adaptive selection algorithm is used automatically to adjust the combination of parameters, including the number of Gaussian mixture components and the length of segment. Experimental results perform that the ESATP method in the experiment shows high predictive accuracy, and maintains a high time efficiency. This model can be used in products of mobile vehicle positioning.

Key words: environment adaptive, variational Gaussian mixture model (VGMM), parameter adaptive selection algorithm, trajectory prediction

中图分类号: