【摘要】 该文提出一种新型联合约束的级联交互式多模型卡尔曼滤波器,该滤波器由两个滤波器前后两级串联而成;第1级为标准交互式多模型滤波器;第2级为联合约束滤波器。联合约束滤波器的约束条件对第1级滤波器中的多模型集合中各子模型均有效。联合约束滤波器采用平滑约束卡尔曼滤波算法对第1级滤波结果进一步优化。以机动目标实时跟踪为实际工程应用背景,数值仿真和飞行实验结果证明新的联合约束性级联交互式多模型滤波器比标准交互式多模型滤波器具有更小的估计误差和方差,所增计算量合理可行。该文为交互式多模型滤波器和机动目标跟踪两个方向的进一步改进提供了有益借鉴。更多还原

【Abstract】 A novel unified cascade constrained interactive multi-model Kalman filter is put forward. The filter is composed of two cascade connected filters, a standard interactive-multiple-model and a unified constrained filter. The latter is effective for everyone in model set of controlled plant and refines the estimation of the former using smoothly constraint Kalman algorithm. Numerical simulation and flying experiments are made for maneuvering target tracking and lower estimated error and covariance are achieved by the unified cascade constrained interactive multi-model Kalman filter compared with conventional interactive multi-model filter. The added computation cost is reasonable and acceptable. The paper is valuable reference for maneuvering target tracking and interactive multi-model filter.更多还原