模型组转换算法在机动目标跟踪中的应用

为改进变结构多模(VSMM)估计中模型组转换(MGS)算法对运动规律较复杂目标的跟踪,提出了用于模型组激活的核心模型转移法。该法为每个模型组确定一核心模型,并利用当前时刻滤波结果中的模型概率和似然等先验与后验信息,判断模型组的激活。仿真结果表明,核心模型转移法的滤波结果平稳,为MGS算法提供了一种简单易行且全面利用系统信息的模型组激活方法。

Application of Maneuvering Target Tracking with Model Group Switching Algorithm

To improve model group switching(MGS) algorithm in a variable structure multimodel tracking the target with complex moving rule,a core model diverting method was put forward in this paper.The new method was used to determine a core model for each set of model,and judge whether some model was active according to the model possibility and likelihood that were priori and posterior information.The simulation results showed that the filter results was relative smooth by using the core model diverting method,which provided a simple method and using integrated information of the system to activate the set of model for MGS algorithm.