摘要
由于模型参数不能自适应调整和卡尔曼滤波器固有的特点,传统的当前统计模型算法跟踪突发强机动目标时性能显著下降。本文通过采用机动检测方法并借鉴强跟踪滤波器的思想,提出了一种改进的自适应强跟踪算法。利用量测残差的统计距离将目标机动划分为不同的状态,相应调整模型参数和滤波器增益,提高机动模型和系统模式的匹配程度,增强了系统对强机动目标的跟踪能力并保持对一般机动目标良好的跟踪性能。
The performance of the original current statistical model（CSM） is very poor in the case of tracking a sudden maneuvering target,because its parameters cannot be adaptively adjusted and have inherent characteristics existed in Kalman filter.By adopting the detection method of motion and the idea of the strong tracking filter,an adaptive modified strong tracking algorithm is proposed based on CSM.By using the statistical distance of observation residuals to sort the maneuver states of targets,the parameters and the filter gain are adjusted to improve the match between the motion model and the system mode.Thus the approach enhances tracking performance for sudden maneuvering targets and maintains good performance for general motion.
出处
《数据采集与处理》
CSCD
北大核心
2008年第2期191-195,共5页
Journal of Data Acquisition and Processing
基金
国家"八六三"高技术研究发展计划资助项目
关键词
机动目标
机动模型
当前统计模型
强跟踪滤波器
maneuvering target
motion model
current statistical model
strong tracking filter
