基于多模型估计提高捷联惯导系统初始对准的精度

采用卡尔曼滤波技术进行捷联惯导系统静基座初始对准时，不能准确获得实际系统中噪声的方差，存在一定的不确定性；同时，由于一些状态不可估计或估计效果很差，采用简化状态方程进行对准，存在着模型参数不确定性。本文采用多模型估计方法处理这些不确定性问题，大大提高了对准的精度，仿真结果表明这一方法是有效性的。

Improvement on Initial Alignment Accuracy of SINS Using a Multiple Model Estimator

When the Kalman filtering method is adopted in initial alignment for a strapdown inertial navigation system (SINS) on a stationary base, the variance of system noise can not be obtained with accuracy, namely there exists some uncertainty. Meanwhile, as some states can not be estimated or have poor estimation, the simplified state equation employed in the alignment process will cause model parameter uncertainty. A multiple-model estimator was used to solve the problem of uncertainty, and the alignment accuracy was improved remarkably. The simulation results show that the multiple-model estimator is effective in initial alignment for SINS on a stationary base.