摘要
针对多模型算法在机动目标跟踪中存在的问题,运用交互多模型算法(IMM)和自适应滤波理论,设计了一种自适应交互多模型算法(AIMM),结合目标运动模型对目标当前加速度和其方差进行估计,并在此基础上给出了AIMM中模型集和模型转移概率的设计方法,进行了计算机仿真。蒙特卡罗仿真结果表明,与标准IMM算法相比,该算法比IMM算法的跟踪性能有很大提高,跟踪复杂机动目标比IMM有更快的收敛速度,跟踪滞后问题得到较好的解决,跟踪目标的稳定性和精确性均优于IMM算法,有利于机动目标的实时跟踪。
A new adaptive interacting multiple model(AIMM) algorithm is designed using the standard IMM algorithm and adaptive theories, according to the problems in maneuvering target tracking using IMM. Combined with the target maneuvering models, the current acceleration of the target and its covariance are estimated, by which the model set and the model transition probability are designed. The Monte Carlo simulation results indicate that the tracking performance is proved when tracking complex maneuvering target using AIMM. The AIMM has faster constringency and more stable and accurate tracking performance than the standard IMM which is effective in real-time target tracking.
出处
《情报指挥控制系统与仿真技术》
2005年第4期24-27,42,共5页
Information Command Control System and Simulation Technology
关键词
目标跟踪
交互多模型算法
自适应
蒙特卡罗仿真
target tracking
interacting multiple model algorithm
adaptive
Monte Carlo simulation