基于模糊推理的机动目标自适应多模型跟踪算法

针对当前空中来袭目标的主要特点,运用模糊理论和多模型理论探索空中机动目标跟踪问题,并设计了一种模糊自适应多模型(FAMM)目标跟踪算法,该算法采用五个基本模型,以加速度估值作为模糊推理系统的输入,经模糊推理融合得到系统状态和方差的估计值以及下一时刻的滤波模型(最多三个).经Monte Carlo仿真研究,与IMM算法相比较,该算法不仅在目标弱机动或不机动条件下,而且在复杂机动时能更稳定、精确地跟踪目标,较好地满足了海上对空防御作战中跟踪机动目标的需求.

Adaptive Multiple Model Tracking Algorithm for Maneuvering Target Based on Fuzzy Inference

Focused on the characteristics of air-target, fuzzy theory and multiple models theory are applied in the field of tracking target, and a new tracking algorithm, called fuzzy adaptive multiple models (FAMM), is presented. On the basis of five models, the acceleration, resulted by running Kalman filtering, is considered as the input to the fuzzy inference system, by which, the state and covariance and a switch for models are assigned as the output of system. The Monte Carlo simulation results indicate that it is more stable and accurate to track maneuvering target by using the FAMM algorithm rather than IMM. It will be more efficient for warship air-defense.