针对变结构多模型算法中模型集自适应较复杂,且模型扩展受模型结构限制等问题,提出了一种最小模型组最优模型扩展的机动目标跟踪算法。该算法以最小模型组作为基础有效模型集,采用模型组切换方法进行模型组自适应;并根据Kullback-Leibler距离准则在连续的模型空间中对基础模型组进行最优模型扩展。因此,该算法具有模型集自适应简单、模型激活不受模型结构的限制等优点。多组实验仿真结果表明:该算法既可以对相同结构的模型进行激活,也可以对不同结构的模型进行激活;在没有明显增加计算量的同时,提高了目标的跟踪精度,具有较好的跟踪效果。

A new approach to maneuver target tracking, named best-model augmentation on minimal model-group algorithm, is proposed in this paper. In the algorithm, the minimal model-group is adopted as the basic model group and model-set adaptation applies model-group switching algorithm. Besides, the basic model group is augmented by a variable model, called best-model, which is obtained depending on Kullback-Leibler distance criterion. Thus, the proposed algorithm has some advantages such as simple model-set adaptation and the activation of the models which has different structure. The simulation results show that the model with similar structure can be activated and the model with different structures can be activated as well in the proposed algorithm. At the same time, it has a small amount of computation and higher tracking accuracy.

国家自然科学基金资助项目;高等学校博士学科点专项科研基金资助项目;教育部新世纪人才项目