两步快速可重构无迹卡尔曼滤波算法测量导弹滚转角

Two-step fast reconfigurable unscented kalman filter algorithm for measurement of missile rolling angle

快速准确地测量旋转导弹的滚转角是非常重要的,利用地磁场测量滚转角是常用的一种方法。由于磁场测量手段中不可避免的误差,使得滚转角不能及时准确的被测量。提出了两步快速可重构无迹卡尔曼滤波（ARUKF）算法,可以快速的校准磁强计及测量电路,从而准确的估计导弹的滚转角。与现有算法不同,ARUKF的状态向量根据滚转角参数与测量电路的校正参数而提出,并且在不同阶段可以配置不同的被估计参数,这样可以较大程度上节约弹载计算机资源。为了滤波算法快而准确的收敛,该算法在滤波初始阶段,根据量测信息与测量电路特征参数进行滤波初值的两步快速调整。通过仿真与半实物仿真结果表明,两步ARUKF算法在收敛特性和计算成本方面优于现有的UKF算法。

It is very important to quickly and accurately measure the roll angle of the rotating missile. The geomagnetic field is often utilized to measure the roll angle. However,due to the unavoidable error of the magnetic field measurement,roll angle cannot be measured accurately on time. In this paper,one kind of two-step reconfigurable Kalman filter（ARUKF） algorithm is proposed,which aims to quickly calibrate the magnetometer. The corresponding measurement circuit is also realized. Compared with the existed algorithms,the state vectors of the ARUKF are obtained based on the roll angle parameters and the calibration parameters of the measurement circuit. Different estimated parameters can be configured at different steps,which can save the computer resources significantly. Moreover,to filter the algorithm quickly and accurately,the two-step adaptive processing of the initial value is performed at the initial step of filtering according to the measurement information and characteristic parameters of the measurement circuit. Simulation and hardware-in-loop experimental results show that the proposed ARUKF algorithm is superior to the existed UKF algorithms in terms of convergence characteristics and computational cost.