论文部分内容阅读
准确的弹道系数辨识和精确的目标状态估计是再入目标高精度跟踪与高可靠识别的关键。一方面,状态估计的误差会造成模型参数(弹道系数)的辨识风险;另一方面,模型参数的辨识偏差又会导致模型失配从而降低目标状态的估计精度。因此,需要实现再入目标的状态估计和参数辨识的联合优化。针对再入目标弹道系数未知情形,提出了一种基于期望最大化(EM)框架并采用粒子滤波(PF)平滑器实现的PF-EM联合优化算法。在E步基于粒子平滑器得到目标状态的后验平滑估计,M步采用数值优化算法更新上一次迭代的弹道系数,通过E步和M步的不断迭代,以保证状态估计和弹道系数辨识的一致性。算法仿真对比表明:所提算法的状态估计和参数辨识精度均优于传统的状态增广算法。
Accurate ballistic coefficient identification and accurate target state estimation are the key points to re-enter the target with high-precision tracking and high-reliability identification. On the one hand, the error of state estimation can lead to the identification risk of model parameters (ballistic coefficients); on the other hand, the identification error of model parameters can lead to model mismatch and thus reduce the estimation accuracy of target states. Therefore, joint optimization of state estimation and parameter identification of reentry target is needed. A new PF-EM joint optimization algorithm based on expectation maximization (EM) framework and particle filter (PF) smoother is proposed for unknown target ballistic coefficient of reentry target. In step E, we obtain a posteriori smoothness estimate of the target state based on the particle smoother. In step M, the numerical optimization algorithm is used to update the ballistic coefficient of the last iteration, and the iterative steps E and M are used to ensure consistent state estimation and ballistic coefficient identification Sex. Simulation results show that the proposed algorithm is better than the traditional state-based augmentation algorithm in state estimation and parameter identification.