分布式不确定性系统状态估计

Distributed Estimation Algorithm for Uncertain System

研究了分布式不确定性系统的状态估计。以两部雷达构成的分布式系统为对象，考虑杂波环境、量测噪声和目标不确定性机动下的多目标跟踪。针对其中的两类不确定性问题，提出了机动目标自回归统计模型，并将它与联合概率数据关联相结合，给出了一种新的分布式多目标跟踪算法。仿真结果证明了其快速性和自适应性。

To the best of the authors' knowledge, there is just one paper by Chang on tracking algorithm for multisensor network system that has taken into account all the following threefActors: multi-targets, uncertainty of target maneuvering, and uncertainty of sensor measurements. But the IMM (interaction multiple models) employed by Ref. [4]required highcomputation. We present a new algorithm that overcomes the shortcoming of Ref. [4]. Ournew model uses the auto-regressive value aS the mean value of target's acceleration, regards the acceleration as a time-related process, and assumes modified Rayleigh distributiondense function. This new auto - regressive statistical model is highly effective even whenstrong uncertainty of target maneuvering exists. In dealing with uncertainty of sensor measurements, we employ JPDA (joint probabilistic data association) method. Thus very highlysuccessful association probability is attainable and JPDA method is suitable even for environment that has two un favorable features: low detective probability and multi-clutters. Computer simulation results show that this new algorithm has high computaion efficiency, highaccuracy and strong adaptability.