摘要
针对机动目标跟踪过程中建立的目标模型和目标的实际运动模式出现失配的问题,提出了从一组离散模型集中选出最优模型,并自适应调整模型参数,使模型逼近目标实际运动模式的交互式多模型算法。蒙特卡罗仿真表明,该算法与传统的常速模型与自适应协同转弯模型交互算法(IMM-CV/ACT)相比,在目标发生强机动时,能及时有效的把跟踪误差峰值控制在测量标准差之下,适合于强机动目标跟踪。
In maneuvering target tracking, the established dynamic model may mismatch the real system mode. This paper presents an adaptive interacting multiple model （ AIMM ） algorithm in order to match the real system mode. It selects the optimal model from the candidate model set and adjusts it adaptively if necessary. The Monte - Carlo simulation illustrates that this algorithm is superior to the conventional IMM - CV/ACT algorithm, which uses the CT （Constant Velocity） model and the ACT （Adaptive Coordinated Turn） model for interacting. The results also show that it can be applied to the high level maneuvering target tracking for its ability to control the tracking error under the deviation of measurement quickly and effectively.
出处
《计算机仿真》
CSCD
2008年第4期326-329,共4页
Computer Simulation
基金
国家自然科学基金项目资助(60434020
60602049)
