基于MPF的惯性导航系统快速对准技术

针对目前应用于惯性导航系统初始对准中的扩展卡尔曼滤波存在对准精度低、时间长的缺点,提出一种基于模型预测滤波的快速对准技术。该方法将惯性器件测量误差视为模型误差使用MPF进行实时预测,并以此来修正惯性导航系统的平台误差角。这样,MPF不仅有效提高了平台误差角的估计精度,而且降低了系统状态变量的维数,大大缩短了初始对准时间。仿真结果表明,MPF在平台对准精度和快速性方面均明显优于EKF,初时对准时间仅是EKF的10%,而对准精度却高于EKF。

A fast alignment method for inertial navigation system based on MPF

A fast alignment method based on model predictive filter （MPF） is put forward to solve the problems of low accuracy and long time, consisting in extended Kalman filter （EKF） applied to initial alignment of inertial navigation system （INS）. In this method, the measurement error resulted from the inertial unit is considered as the model error to be predicted in time through MPF, then it is fed back to INS to correct the platform angles error. Consequently, MPF not only improves the estimation precision of the platform angle, but also decreases the state dimension of INS, further shorten the initial alignment time. The simulation results show that MPF is superior to EKF in precision and rapidity of the platform alignment, and especially MPF needs less alignment time than EKF, only 10% of EKF, while the alignment precision is better than EKF.