基于随机模型近似的再入目标自适应跟踪算法

Adaptive tracking algorithm for reentry vehicle based on stochastic model approximation

针对再入目标跟踪问题,基于加速度动力学模型和随机模型近似思想,提出了分段匀Jerk自适应模型及跟踪算法.该算法引入Jerk动力学模型和Jerk分段均匀假设,给出了机动加速度的递推模型;根据随机模型近似思想提出了新的过程噪声定义方法并给出了分段匀Jerk模型和过程噪声的自适应方法;结合状态扩展方法和分离差分滤波算法实现了再入目标的实时自适应跟踪.仿真实验表明,相比基于分段匀加速模型的跟踪算法,该算法在保证了再入目标稳态跟踪精度的同时,对目标突变状态具有较强的跟踪能力.

For the problem of reentry vehicle（ RV） tracking,adaptive piecewise constant Jerk model and tracking algorithm were proposed based on kinetics acceleration model and stochastic model approximation. Recursive model of target acceleration was induced by introducing kinetics Jerk model and assumption of piecewise constant Jerk. A new definition and adaption method of process noise was proposed according to the idea of stochastic model approximation. The real-time RV tracking was achieved by divided difference filter based on the augmented state model. Simulation results show that the proposed algorithm has the similar tracking accuracy on stable state as the tracking algorithm based on the piecewise constant acceleration model,but it has better performance on tracking state mutation than the latter.