集群系统下多弹道实时测量数据处理的UKF算法  

UKF algorithm for real-time measurement of multi trajectory data processing in cluster system

作　　者：潘建平[1] 王盛玺[1] 刘宗伟[1] 

机构地区：[1]太原卫星发射中心技术部,山西太原030027

出　　处：《电子设计工程》2016年第13期61-63,共3页Electronic Design Engineering

摘　　要：随着外弹道测量数据采样频率的提高和多发导弹齐射技术的成熟,普通计算机受限于运算能力和内存已无法满足计算要求。文中采用无迹卡尔曼滤波(UKF)算法,结合质点弹道模型,基于曙光TC4000A集群系统设计了针对多弹道实时测量数据处理的并行计算软件,完成了对高频率采样数据的多弹道并行计算。该算法提高了弹道计算精度和计算效率,对集群系统下的并行计算在现代军事中的应用有较强的借鉴意义。With the rapid increase of sampling frequency of the measured data in multi trajectory and the development of multiple missile technology,Ordinary computer can not satisfy the demand of calculating due to its poor operational capability and？internal？memory. To overcome the problem talked above,the cluster system based on Dawn TC4000 A is introduced to our research. We design a parallel computing software using the Unscented Kalman Filter algorithm（UKF）, which improves the accuracy and efficiency of ballistic calculation has a strong reference value in the application of the modern military.

关 键 词：集群系统 弹道计算 UKF滤波 并行计算 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]