基于改进粒子滤波的光电目标跟踪算法研究  

Studies on Tracking of Optoelectronic Target Algorithm Based on AEPF

作　　者：熊茂涛[1,2] 张勇[3] 吴钦章[1] 

机构地区：[1]中国科学院光电技术研究所,成都610209 [2]中国科学院研究生院,北京100039 [3]重庆工商大学,重庆400067

出　　处：《系统仿真学报》2010年第4期886-889,共4页Journal of System Simulation

基　　金：中国科学院"西部之光"人才培养计划基金(2007414)

摘　　要：在光电目标跟踪与定位中,结合扩展卡尔曼滤波和粒子滤波的优点和目标跟踪的非线性特征,提出了一种非线性系统的基于"当前"统计模型的自适应扩展卡尔曼粒子滤波算法,根据光电目标的测量信息修正加速度方差,消除随机误差和噪声的干扰,提高预测的精度。通过MonteCarlo对比仿真实验表明该算法正确有效,定位精度较高,滤波效果得到改善,同时增强了稳定性,优于一般的EKF、PF和EPF算法,为光电目标的精确跟踪与定位的实现提供一种新的方法。In the tracking and orientation of optoelectronic targets,a new nonlinear Adaptive Extended Kalman Particle filtering （AEPF） algorithm based on Current Statistical model was proposed,due to nonlinear character of the system connect with the excellence of the Extended Kalman filtering and the Particle filtering. It expressed variation of acceleration with the information of position and angle to carry out self-adaption,and eliminated random error and noise variance to elevate the accuracy of tracking. Analytic results of Monte Carlo simulation prove the AEPF algorithm is right and feasible,and the accuracy and the stability are both improved. It has better performance than the traditional EKF,PF,EPF algorithms. The method has afforded a new application to the tracking and orientation of optoelectronic targets.

关 键 词：光电目标跟踪 “当前”统计模型 粒子滤波 AEPF算法 

分 类 号：V557[航空宇航科学与技术—人机与环境工程]