一种基于干涉测角衰减记忆Kalman算法的仿真应用  

A Simulation Application of Fading-memory Kalman Algorithm Based on Interferometric Goniometry

作　　者：张建军 吕琳 韩明 吕自鹏 ZHANG Jianjun;LV Lin;HAN Ming;LV Zipeng(Tianjin 712 Communications Broadcasting Limited by Share Ltd.,Tianjin 300462)

机构地区：[1]天津七一二通信广播股份有限公司,天津300462

出　　处：《软件》2021年第10期57-59,共3页Software

摘　　要：本文结合干涉仪测角Kalman滤波算法、衰减记忆Kalman滤波算法,实际应用时,由于模型与实际过程可能不符,此时新息对估计值的修正作用下降,导致卡尔曼滤波发散。增大新息对估计值的权重,是扼制滤波发散的一个可行途径。本文对比Kalman滤波算法、衰减记忆Kalman滤波算法仿真分析,并给出了仿真结果,由仿真结果可知,在一定通道相位噪声的条件下,对于正弦运动,采用衰减记忆以后,卡尔曼滤波结果得到了明显的改善,角度角速度误差均明显减小。增大新信的比重,同时监控算法是否发散,可以明显改善滤波效果。This paper combines interferometer angle measurement and Kalman filtering algorithm,fadingmemory Kalman filtering algorithm.In practical application,due to the possible inconsistency between the model and the actual process,the correction effect of innovation on the estimated value decreases,resulting in the divergence of Kalman filter.Increasing the weight of innovation to the estimated value is a feasible way to curb the filter divergence.This paper compares the simulation analysis of Kalman filtering algorithm and fading-memory Kalman filtering algorithm,and gives the simulation results.The simulation results show that under the condition of certain channel phase noise,the Kalman filter results are significantly improved for sinusoidal motionafter using fading-memory,the angular and angular velocity errors are significantly reduced.Increasing the proportion of new signals and monitoring whether the algorithm is divergent can significantly improve the filtering effect.

关 键 词：卡尔曼滤波 衰减记忆 解模糊 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]