论文部分内容阅读
研究把多普勒雷达测量数据引入Kalman滤波的新方法.根据测量噪声协方差矩阵的分解导出一种理想的线性测量方程的等价形式,在方向余弦估计和误差补偿的基础上,给出可实现的测量方程及其对应的序贯处理的滤波方程.这种序贯处理结构有助于方向余弦继承位置测量更新带来的性能改善,从而减小其估计误差.蒙特卡罗仿真表明,这种序贯滤波算法,不但可以提高状态估计精度,而且其性能优于传统的推广Kalman滤波.
A new method of introducing Doppler radar measurement data into Kalman filter is studied.An equivalent form of ideal linear measurement equation is derived according to the decomposition of covariance matrix of measurement noise.On the basis of direction cosine estimation and error compensation, And the corresponding filtering equation of sequential processing.The sequential structure can help to improve the performance of the direction cosine inherited position measurement update and reduce its estimation error.The Monte Carlo simulation shows that Sequential filtering algorithm, not only can improve the state estimation accuracy, but also its performance is better than the traditional generalized Kalman filter.