基于模糊推理的机动目标自适应多模型跟踪算法  

Adaptive Multiple Model Tracking Algorithm for Maneuvering TargetBased on Fuzzy Inference

作　　者：黄泽汉[1] 邢昌凤[1] 

机构地区：[1]海军工程大学兵器工程系,湖北武汉430033

出　　处：《情报指挥控制系统与仿真技术》2004年第5期42-46,共5页Information Command Control System and Simulation Technology

摘　　要：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MonteCarlo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。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.

关 键 词：目标跟踪 多模型 模糊推理 仿真 

分 类 号：O231[理学—运筹学与控制论]