基于改进粒子滤波的井下跟踪算法研究与实现

发布时间：2018-03-24 02:30

本文选题：井下跟踪　切入点：无线传感器网络　出处：《计算机应用研究》2017年05期

【摘要】：井下环境复杂多变,射频信号易受到阴影效应、多径衰落等因素的影响。采用传统的粒子滤波跟踪方法误差较大,研究了一种基于改进粒子滤波的井下跟踪算法。初始化阶段利用第一次指纹匹配算法的定位结果来设计初始化概率分布函数;采用核函数法与指纹匹配技术相结合的算法,在采样数据中搜索与目标节点指纹特征相匹配的位置并加权得到位置坐标作为跟踪中的观测值;最后利用粒子滤波将观测值与目标运动状态相融合以跟踪目标运动轨迹。实验结果表明,粒子滤波算法较优化卡尔曼滤波算法更适用于井下跟踪;改进的算法有效增强了跟踪系统的可靠性,提高了跟踪精度,满足了井下的跟踪要求。
[Abstract]:The underground environment is complex and changeable, and the radio frequency signal is easily affected by the shadow effect and multipath fading. In this paper, an improved particle filter based downhole tracking algorithm is studied. In the initialization stage, the initial probability distribution function is designed by using the location result of the first fingerprint matching algorithm, and the kernel function method is combined with the fingerprint matching technique. The position matching the fingerprint feature of the target node is searched in the sampled data and the position coordinate is obtained as the observation value in the tracking. Finally, the particle filter is used to track the moving trajectory of the target by combining the observed values with the moving state of the target. The experimental results show that the particle filter algorithm is more suitable for underground tracking than the optimized Kalman filter algorithm. The improved algorithm can effectively enhance the reliability of the tracking system, improve the tracking accuracy and meet the requirements of underground tracking.
【作者单位】：内蒙古科技大学信息工程学院;
【基金】：内蒙古自治区科技计划资助项目(201502013-1) 内蒙古自治区自然基金资助项目(2015MS0623)
【分类号】：TD76;TN713

【相似文献】