基于先验约束信息的航姿参考系统滤波算法研究

Accuracy is the core problem of inertial measurement and attitude heading reference system. Errors accumulate over time is the main technical difficulties of inertial system. To explore new sensor information fusion algorithm, has been the focus of research in the field of inertial technology. Based on prior constraint information to establish a simplified navigation position reference system error model and the filtering algorithm, improve the system accuracy, namely the basic research ideas for this application. This research through the establishment of a dynamic model of the information with a priori constraints, the abnormal noise on the influence of dynamic measurement and how to use the observed value and forecast the time series to eliminate its impact. Robust analysis method was used with navigation position reference system state of time-varying nonlinear model is transformed into the error of the approximate linear time-invariant model, which is based on the transfer function of frequency domain analysis, combined with prior constraint information carrier, will be complex identification problem in the time domain analysis, in the frequency domain analysis model simplified system parameter identification based on spectrum analysis of the related problems. Forecast tracking performance and parameter estimation algorithm through a binding navigation position reference system to evaluate ground vehicle simulation and practical test. According to the simulation conclusion, guide to complete navigation position reference system based on micro inertial device design and production. Research results have high theory value and application prospect of promotion.

精度是惯性测量与航姿参考系统的核心问题。误差随时间积累是惯性系统的主要技术难点。探索新的传感器信息融合算法，一直是惯性技术领域研究的焦点。基于先验约束信息建立简化的航姿参考系统误差模型和滤波算法，提高系统精度，即为本申请的基本研究思路。本研究通过建立带有先验约束信息的动态模型，研究异常噪声对动态测量的影响以及如何利用观察值和状态预报值的时间序列来消除其影响。采用鲁棒分析方法，将航姿参考系统的时变非线性模型转化为误差状态的近似线性定常模型，从而使基于传递函数的频域分析成为可能，结合载体先验约束信息，将时域分析中复杂的辨识问题，在频域分析模型中简化成基于相关频谱分析的系统参数辨识问题。预测跟踪性能和参数估计算法通过一个绑定航姿参考系统的地面车辆仿真和实际测验进行评估。并根据仿真结论，指导完成基于微惯性器件的航姿参考系统的设计与制作。研究成果具有较高的理论价值和应用推广前景。

通过建立带有先验约束信息的动态模型，系统研究了动态条件下航姿参考系统（AHRS）性能特性和误差特征，搭建了测试平台。对AHRS理论模型与实际系统的差异进行了分析，分析了模型误差对AHRS测量精度的影响，系统研究了异常噪声对动态测量的影响以及如何来消除其影响；提出可利用观测值和状态预报值的时间序列来消除其影响；研究了不同约束的卡尔曼滤波算法，研究在卡尔曼滤波的数学模型中增加约束条件方程，推导出约束状态下的卡尔曼滤波递推方程；根据具体应用背景和先验约束信息对系统进行了有效合理的线性定常简化，推导出相应的载体运动频域约束条件；通过系统仿真验证了算法的可行性和算法精度；设计并制作出航姿参考系统，通过建立可行准确的误差模型、滤波算法和温度补偿，来提高AHRS精度；对关键技术指标进行计算、仿真和实验验证；开展了AHRS 应用研究，根据应用实际情况，优化改进系统结构、电路和算法；最后制作出满足车载导航要求的 AHRS 样机，静态精度优于 0.01 度,动态精度优于 0.05 度，并成功应用于多项国防军事装备。研究成果具有较高的理论价值和市场应用前景。