小波中值滤波在晃动基座初始对准中的应用

针对捷联惯性导航系统(SINS)在晃动基座下进行初始对准时,陀螺和加速度计受外界干扰噪声较大,而单独采用低通滤波或小波滤波法不能有效实时地实现噪声抑制问题,采用小波阈值消噪和中值滤波相结合的混合滤波方法来实现信号的消噪。将输出信号首先经中值滤波消除信号中的突变噪声,再利用小波阈值消噪滤除振动噪声和白噪声,从而有效地消除基座晃动带来的噪声。测试实验表明,提出的混合滤波法对车载惯导系统输出信号噪声滤波效果明显,在发动机启动的情况下,可基本达到静态信号采集的效果。对准仿真实验表明,提出的滤波法能辅助捷联惯导系统实现在晃动基座下的精对准。

Application of Wavelet Median Filter to Initial Alignment of Shaking Base

The gyroscope and accelerometer is influenced greatly by the environmental interference noises when the initial alignment of SINS on the shaking base is carried out.Furthermore, the noise is difficult to be supressed effectively and in real time by using the low pass filtering (LPF) or wavelet filtering independently.A mixed filtering method combined the wavelet threshold denoising and mediam filtering has been proposed to eliminate the signal noise in this work.The burst noise in the output signal is removed firstly by the median filter,then the vibration noise and white noise are filtered by wavelet threshold method.Therefore, the noise which is brought in by the shaking base can be lowered down effectively without reducing the precision of alignment.The experiment results show that the pre filtering method presented in this paper has obvious effects in filtering the noise of signal output from vehicle navigation system,and the signal acquisition results while the engine is running are matching well with the static signals. And the simulation results show that this filtering method works well in the precise alignment of shaking base as an assist to SINS.