基于精细抗干扰滤波的惯性导航系统初始对准方法研究

Inertial Navigation System (INS) has particular advantages, such as high autonomy, high accuracy in short time, concealment and can successively supply position, velocity, attitude angle, and so on. As a dead reckoning system, in the process of calculating navigation parameters, the attitude errors accumulate with time extendibility and induce other navigation errors. Therefore, one of the most important problems for INS is initial alignment, which affects the navigation accuracy directly. The inertial navigation systems unavoidably exist multiple unknown inputs, including system structural variation, measurement noise, environmental disturbances, system faults, and so on. Accordingly, it results in the reduced accuracy and reliability of the initial alignment. .In this project, firstly, a design framework for a class of elegant anti-disturbance filter is proposed. Then, the characters of multiple disturbances in an INS are systematically analyzed and can be divided into different types. A new class of INS error model with multiple disturbances is established. Secondly, an anti-disturbance initial alignment approach for INS is presented based on disturbance attenuation and rejection filtering scheme against multiple disturbances. In addition, a class of anti-disturbance fault tolerance initial alignment method is addressed for INS with multiple disturbances and system faults. Finally, the validation method is discussed based on measured data in an experiment or hardware-in-the-loop simulation technique to provide a theoretical basis for high-accuracy and high reliability inertial navigation system or inertial integrated navigation system technologies.

惯性导航系统（INS）具有短时间内精度高、完全自主、能连续提供位置、速度、姿态信息等优点。作为一种航迹推算系统，在计算导航参数过程中，姿态误差随时间积累并引起其它的导航误差。因此，初始对准是INS关键技术之一，并直接影响到导航精度。INS往往受到系统结构变化、量测噪声、环境干扰以及系统故障等多种因素的影响，从而导致初始对准的精度和可靠性降低。.本项目首先提出一类精细抗干扰滤波器设计框架；其次，系统分析INS多干扰源特性，对惯性导航系统中的干扰进行分类建模，构造一类多源干扰INS误差新模型；然后，针对惯性导航系统中多源干扰，设计一类具有干扰抵消与抑制性能的抗干扰高精度初始对准方法；针对惯性导航系统存在多源干扰和系统故障，设计一类抗干扰高可靠性容错对准方法；最后，研究基于实测数据或半物理仿真技术有效性验证方法，为高精度和高可靠性惯性导航系统或惯性组合导航系统技术研究提供理论基础。

惯性导航系统普遍受到测量随机噪声、惯性器件漂移、传感器故障、建模误差以及外部干扰等多种类型干扰的影响。大部分初始对准方法都是假设这些干扰为高斯噪声或等价的范数有界干扰，然后利用卡尔曼型滤波器或鲁棒滤波方法进行初始对准研究。为了提高惯导系统的精度和可靠性，因此需要研究抗干扰滤波方法及其在惯导中的应用。. 本课题重点研究了多源干扰系统的干扰分类建模，抗干扰滤波方法及其在惯性导航系统初始对准和容错控制中的应用。主要研究内容包括：. （1）针对同时含有多源干扰和故障的非线性系统，研究抗干扰容错控制方法。对于系统中多源干扰分类建模，建立多源干扰系统抗干扰容错控制的研究框架。. （2）针对含有多源干扰的惯性导航系统以及惯性组合导航系统，系统地分析INS多干扰源特性，对干扰进行分类建模，构造一类含多源干扰和故障的INS误差新模型；针对惯性导航系统同时存在多源干扰和系统故障，设计一类抗干扰高可靠性容错对准方法，为高精度和高可靠性惯性导航系统或惯性组合导航系统技术研究提供理论基础。. （3）针对具有未知干扰的复杂非高斯、非线性随机分布系统，建立了随机分布系统抗干扰控制及滤波研究框架。. 研究结果已在国内外学术期刊及会议上发表论文20篇，分别有4篇、15篇论文被SCI、EI收录。授权/申请国家发明专利3项。出版学术专著1部。已培养5名硕士研究生。本项目的研究具有重要的理论意义和实用价值。