Kalman弹道滤波器的快速收敛性研究

在弹道修正过程中，需根据一段实测弹道参数，准确地估算常规弹箭的飞行弹道。工程中，常应用Kalman滤波方法，结合质点弹道模型建立六状态滤波弹道模型。速度初值估计的准确性，是对滤波过程造成影响的主要因素。为改善这一状况，采用多项式拟合用以对速度状态进行估计并以此作为滤波初始值。在此种估计下滤波，速度滤波过程可快速收敛，位置滤波过程收敛速度大幅度提高，可使最优弹道诸元参数获取时间缩短30%。

Study of the Fast Convergence for Kalman Ballistic Filter

During the trajectory correction, requiring some measured trajectory parameter to estimate the flight of conventional projectile accurately. In engineering, the Kalman filtering method is used to establish a six-state filtering ballistic model based on the particle trajectory model. Accuracy of initial estimated value of velocity that affects the filtering process is the main factor. To improve this situation, polynomial fitting is used to estimate the velocity state and as the initial value of the filter. In this estimation, the velocity filtering process can be rapidly convergent, and the speed of convergence of the position filtering process can be greatly improved, which can shorten the time of the optimal ballistic parameters by 30%.