摘要
针对基于传统粒子滤波的GPS(Global positioning system)定位数据处理方法存在粒子退化的问题,研究了基于马尔可夫链蒙特卡罗(Markov chain Monte Carol,MCMC)粒子滤波的GPS定位数据处理算法,引入典型的MCMC方法——Metropolis Hastings(M-H)抽样算法。利用观测伪距非高斯误差分布,建立重要密度函数,将MCMC粒子滤波与建立的GPS系统非线性状态空间模型结合。实测数据实验结果表明,MCMC粒子滤波可有效抑制粒子退化,解决了GPS定位数据滤波这一非线性非高斯问题,避免了噪声的高斯假设和非线性部分的线性化误差,与基于传统粒子滤波的GPS定位数据处理方法相比,该方法降低了定位数据经纬度和速度估计误差,获得了更高的定位精度,并能够在GPS信号质量较差情况下,对GPS定位数据有效滤波,保证载体在此期间内保持较高的位置精度。
Global positioning system （GPS） positioning data processing algorithm based on the Markov chain and Monte Carlo （MCMC） particle filter is discussed to solve the problem of GPS positioning data processing based on the standard particle filter （PF） suffers from severe sample degeneracy. The standard MCMC method, Metropolis Hastings （M-H） sampling, is incorporated into the particle filter algorithm framework, and applied to GPS positioning data processing. It combines the particle filter with the GPS system nonlinear dynamic state-space model. The MCMC method is adapted to solve the degeneracy phenomenon of particle filter. It is effective to nonlinear and non-Gaussian state estimation problems in GPS positioning data processing. Experimental results based on real GPS data show that the MCMC particle filter can increase the sample variety and reduce sample degeneracy. GPS positioning data processing based on the MCMC particle filter reduces the error of position and velocity, compared with GPS positioning data processing based on the standard particle filter. Moreover, the MCMC particle filter can provide a highly accurate positioning value as an aided method when the GPS signal quality is poor.
出处
《数据采集与处理》
CSCD
北大核心
2013年第2期213-218,共6页
Journal of Data Acquisition and Processing
基金
国家自然科学基金(61101161)资助项目
航空科学基金(2011ZC54010)资助项目
辽宁省博士启动基金(20101081)资助项目
