基于污染分布的动态导航RAIM处理与质量控制

RAIM algorithms have been apllied in the most GNSS receivers, but the performance of the traditional algorithms are not good enough. For example, the probability of the miss detection and the fault alert is not controlled very well, and the requirement of many users can not be satisfied. The most improtant uncertainty is the probability of the distribution of the observation, and the precondition of traditional RAIM algorithm is based on the Normal distributi of observation error. When the assumed distribution of observation error is contaminated, the results baesed on the LS esimator may be departured from the fact, and the probability distribution of the detecting statistic is also not accorded from the model, so that the calculating and ditecting results of the RAIM algorithm will not ensue the users' safety very well, so the new theory and methods are needed to improve the performance of RAIM algorithm. This item focuss on the uncertainty and make some researching on it, and the most researching contents consist of modeling the distribution of the observation error, the distribution of the position error and the distribution of the detect statistic, biulding the user interity influence function, making the estimator optimization, making the detecting statistic optimization so as to minimize the integrity risk of user.

RAIM算法广泛应用于动态导航接收机，对动态导航解有实时完好性增强作用。一般情况下，该算法是建立在一个理想化的假设前提之上：观测误差服从正态分布，采用最小二乘估计方法，基于正态分布对检验统计量进行统计推断。倘若观测误差是一复杂污染分布，将会导致以下结果：采用最小二乘估计解是有偏的、检验统计量也有悖于模型化的概率分布，基于此假设进行统计推断必然会出现偏差，造成算法应用性能不高，满足不了高水平完好性应用需求，因此需要新理论对其进行完善与改进。本项目拟将稳健估计理论、抗差自适应滤波理论、可靠性分析理论引入到RAIM算法中，构建用户完好性风险影响函数，针对复杂污染分布条件下的误差概率分布进行建模与分析，以用户完好性风险最优控制为原则，选用最优的参数估计器，构建最优的检验器，目标是提高统计推断的准确性，降低动态导航完好性风险，增强RAIM处理的稳健性，从理论上完善现行RAIM处理方法。

课题聚集基于污染分布的动态导航RAIM处理与质量控制方法并展开研究，主要研究内容为：将抗差自适应滤波、可靠性分析等理论引入到RAIM算法中，提出目标函数构建原则及用户完好性性能优化策略，利用长期观测数据，对伪距观测误差进行分析并建立模型；以最优估计和假设检验理论为基础，以用户完好性风险、连续性风险最优控制为原则，选用最优的参数估计器，构建最优的检验器，优化动态导航RAIM处理性能。.课题形成的创新性成果包括：1）提出了基于峰度、偏度、周期性相关系统、QQ-Plot图等误差模型分析方法，并给出了基于蒙特卡洛分布估计的模型求解方法；2）提出基于完好性风险最优控制的目标函数构造方法，给出了三种完好性性能优化策略；3）提出了基于检验阈值扩展的RAIM优化算法、最优加权最小二乘法的RAIM优化算法、基于设计向量的非最小二乘估计RAIM优化算法等；4）提出了一种基于抗差扩展卡尔曼滤波的动态序贯RAIM优化算法。课题通过大量数据计算分析，比较、分析、验证了上述方法的性能。