针对传统基于视速度双矢量粗对准中,由于传感器随机噪声的影响,存在对准精度差,收敛速度慢的缺点,提出了一种新型自适应Kalman滤波的参数识别粗对准方法.该方法通过对视速度运动进行建模,设计采用自适应Kalman滤波对模型参数进行参数识别,从而有效地消除视运动中的随机噪声,提高粗对准的精度和收敛速度.由于自适应滤波的特点,新方法不需要对传感器误差进行统计,使其在实际系统中具有更加广泛的应用价值.针对双矢量粗对准的计算特点,设计了一种矢量重构算法,从而尽可能地规避双矢量共线性问题,加快了粗对准的收敛过程.仿真与转台实验表明,与传统方法对比,新方法在相同的对准时间内具有更高的对准精度,在相同的对准精度下,具有更高的收敛速度.转台实验的最终对准精度为-0.1391°,标准差为0.012°.%Traditional dual-vector coarse alignment with apparent velocity has the problems of poor alignment precision and slow convergence rate due to the influence of random noises on inertial sensors.To solve this problem,a new dual-vector coarse alignment method is designed,which uses a new adaptive Kalman filter to estimate the parameters in an apparent velocity model without using the accurate covariance of the measurement noises.Meanwhile,a reconstructed algorithm with recognized parameters is adopted for the dual-vector,which can avoid the collinearity of the dual-vector.Analysis and simulation indicate that,by using this method,the random noises in the measured apparent velocity can be effectively eliminated compared with the traditional dual-vector coarse alignment.Simulation and turntable experiments show that,compared with traditional methods,the proposed method for the coarse alignment can acquire more accurate alignment results with the same alignment time,and can improve the convergence rate with the same alignment accuracy.The turntable tests by the new method show that the yaw error is-0.1391 ° and the standard deviation is 0.012°.
展开▼
