非陀螺找北仪的时序分析与数字滤波

Time series analysis and digital filtering for a gyroless north seeker

为进一步提高非陀螺找北仪的精度 ,分析了系统的多维自回归时间序列模型 ,并对影响模型参数估计精度的因素进行了讨论 ,指出了最小二乘参数估计方法的误差主要是由于 3倍频以内分次谐波和同频干扰引起。根据模型与信号功率谱分析 ,提出了降低系统噪声影响 ,提高参数估计精度的数字滤波方法 ,并通过试验对滤波方法进行了验证。

The principle of an accelerometer based north seeker is introduced in this paper. ARV (auto regressive vector) model of multivariable time series and its parameters estimation error of the least squares method are also analyzed. Simulation shows that the error is mainly brought about by fraction harmonics and common frequency disturbances. A digital filtering method is proposed to decrease attenuate harmonics effectively according to system model and power spectrum analysis. The principle that accelerometer is used instead of gyro to measure the horizontal component of the earth rate is feasible. By properly modeling and digital filtering, parameters estimation accuracy and north seeking accuracy can be effectively improved.