针对单模型跟踪算法无法实现对弹道导弹连续精确跟踪的问题,提出了一种地心地固坐标系下弹道导弹全阶段连续跟踪的交互式多模型(IMM)算法。根据导弹不同飞行阶段的受力情况建立IMM算法的两个滤波模型集:助推段采用恒轴向力模型,中段、再入段采用被动段跟踪模型,两个模型都利用滤波精度高、数值稳定性较好的求容积卡尔曼滤波算法进行滤波。针对该IMM算法在混合估计过程中引入偏差的问题,采用无偏混合的方法,以两种典型的跟踪场景进行仿真校验,实验结果验证了该文算法的优越性。

Aiming at the problem that single model tracking algorithm can not track the ballistic missile accurately, an interacting multiple model (IMM) algorithm is proposed in Earth-centered Earth-fixed coordinate for continuously tracking the missile for all the flight stage. According to the force condition in different stages of trajectory, the corresponding tracking model is established. The axial force of target almost keep constant in boost phase, the constant axis force model is adopted; coast and reentry phase use the passive phase tracking model to track the missile. Both the models use the cubature Kalman filter algorithm to filter. In view of the problem that a bias is caused in the mixing stage of this IMM algorithm, an unbiased mixing method is adopted. The simulation results verify the superiority of the algorithm.

TN957.51

国家自然科学基金资助项目;安徽省自然科学基金资助项目