非模型化误差理论及其GNSS应用

Due to the complicated spatiotemporal characteristics of GNSS disturbing atmosphere and multipath effect as well as the knowledge lack for their physical background, it is rather difficult to establish the suitable parameterization models, capturing their effects. As a result, the unmodeling errors are objectively existing as spatiotemporal signals, limiting the high-precision GNSS applications. From the compensation aspects of function model and stochastic model, the project aims to synthetically study the theories and methods of processing unmodeling errors in different application modes, such as static positioning, (Network) RTK, PPP etc, with varying observation scenarios. The research contents are as follows. The significance testing theory of unmodeling errors is addressed including the construction of significance testing statistics and its most powerful testing method. For unmodeling errors compensated by stochastic model, we study the efficient method for processing multi-epoch color error correlated model; while for unmodeling errors compensated by function model, we constructing the spatiotemporally correlated stochastic model of unmodeling errors, with which we study the methods for estimation, prediction and application of unmodeling errors. Furthermore, with compensated unmodeling errors, we will comprehensively investigate the integer (partial) ambiguity estimation method and integer validation theory with both geometry and measurement domains taking into account. In addition, the reliability theory and method will also be studied for the reformulated GNSS models with compensated unmodeling errors. The objective of this project is to build up a set of completely new theories and methods for GNSS data processing with effectively capturing the unmodeling errors, realizing the adaptive processing flowchart to the specific application mode and observation scenario.

由于GNSS大气扰动和多路径效应等误差的时空特性复杂以及对其物理背景认知有限，难以建立合理的参数化模型有效补偿，导致非模型化误差以时空信号方式客观存在，制约了高精度GNSS应用。项目从函数模型和随机模型补偿角度出发，系统地研究不同应用模式（静态、网络-RTK、PPP）和观测环境下非模型化误差的处理理论与方法，研究内容包括：非模型化误差显著性检验统计量的构造与最大功率的检验方法；研究随机模型补偿时多历元相关的有色噪声处理方法，函数模型补偿时非模型化误差的时空相关随机模型构建、以及时空域非模型误差的估计、预报与使用方法；研究非模型化误差补偿的（部分）模糊度固定方法和同时顾及几何域和观测值域的模糊度可靠性检验理论；研究非模型化误差补偿的新观测模型的可靠性评估理论与方法等。项目旨在建立全新的含非模型化误差的GNSS数据处理理论和方法，实现随观测环境和应用模式自适应调整的非模型化误差处理流程。

项目系统并研究了GNSS非模型化误差的处理理论与方法，超额完成了项目既定的全部研究内容，并达成了预期研究目标。取得的主要成果有：①研究了不同应用模式下非模型化误差的显著性检验理论，包括非模型化误差的显著性检验，多路径的实时探测，周跳和数据中断修复等。②研究了非模型化误差的随机模型和函数模型补偿理论与方法，提出了非模型化误差的多历元部分参数化方法，以及顾及测站相关非模型化误差的实时自适应加权模型等。③研究了不同高精度应用模式下（静态基线、RTK、网络RTK和PPP）非模型化误差补偿的整数模糊度固定和可靠性评估理论与方法。④研究了非模型化误差的补偿对观测模型的影响，包括三频观测值信号的随机模型估计问题，GNSS观测值及接收机信号质量等。项目研究成果发表论文39篇，其中SCI/EI检索25/13篇，获省部级和行业协会科技进步奖2项，申请发明专利1项、软件著作权1项；参加学术会议和交流报告37人次，其中国际会议10人次；培养博士和硕士研究生12名。