自适应转弯模型的交互多模型算法研究  

Research on Adaptive Turning Model in Grid Multiple Model Algorithm

作　　者：王华楠[1] 刘高峰[1] 顾雪峰[1] 

机构地区：[1]海军工程大学电子工程学院,武汉430033

出　　处：《弹箭与制导学报》2008年第5期241-244,248,共5页Journal of Projectiles,Rockets,Missiles and Guidance

摘　　要：以强机动目标跟踪为背景,在分析常速转弯模型和IMM算法基础上,基于"当前"统计模型的设计思路,提出了自适应转弯模型。通过机动频率系数和角加速度方差修正常速转弯模型的系统状态噪声方差矩阵,设计了不同角速率的转弯模型集,给出了相应模型的状态转移概率,建立了一种自适应转弯模型的交互多模型算法。Monte Carlo仿真实验表明:该算法性能优于其它IMM算法,且计算量小,便于使用。The adaptive-turning model was proposed with the background of strong maneuvering target tracking, which based on constant turning model ＆ IMM algorithm analysis and the thought of ＂current＂ statistic model. By correcting noise variance matrix of system state of constant turn model with maneuvering frequency coefficient and angular acceleration variance, .the turning model set with different angular was designed, the model probability matrix was given. An interactive muhiple model algorithms with adaptive-turning model were built. The Monte Carlo simulation experiment indicates that this algorithm excels CV-IMM Mgorithm in capability and with low computation, especially convenient for application.

关 键 词：机动目标跟踪 常速转弯模型 自适应转弯模型 交互多模型 

分 类 号：TP319.9[自动化与计算机技术—计算机软件与理论]