摘要
在动态导航定位中,目前绝大多数数据处理理论和软件都假设系统状态误差和观测模型误差为高斯白噪声。但在实际应用中,由于卫星轨道误差、大气环境等因素的干扰,使得观测误差和动力学模型误差往往不属于白噪声序列,而是具有一定时间相关或空间相关性的有色噪声。本文将有色噪声归为随机模型进行研究,采用多项式长除法将有色噪声模型展开成级数形式,再根据误差理论求取有色噪声的方差,由该方差修正有色噪声的随机模型,利用现代时间序列分析理论求出状态参数的最优估计值。为了说明该方法的正确性和有效性,用一组动态GPS实测数据进行验证,计算结果表明该方法能有效地抑制有色噪声对动态系统参数估值的影响。
In kinematic positioning and navigation,the state noises and model noises in most data processing are considered to be zero-mean Gaussian white noises.In fact,the observation error and dynamic model error are not the white noise sequence but colored noises which are either temporal correlation or spatial correlation.There are two methods to deal with colored noises; one method is considering them to be stochastic noises,while as model noises in the other method.In this paper,some researches are given on the first method.A new approach is presented by polynomial-quotient which translates colored noises into infinite series,and chooses the finite series,calculates the variance of colored noises.The random model can be corrected with this method.The optimal filter is designed by using the modern time series analysis method.At last,a test using GPS vehicle date provides direct evidence that the proposed method can control the influences of the colored state noises effectively on the parameter estimation.
出处
《测绘科学技术学报》
CSCD
北大核心
2013年第5期443-447,共5页
Journal of Geomatics Science and Technology
基金
国家自然科学基金项目(41274016
40974010
41174006)
