摘要
"当前"统计模型及其自适应卡尔曼滤波算法虽能对强机动目标进行较好跟踪,但存在对弱机动目标跟踪误差较大的缺陷。针对这一问题,在推导传统"当前"统计模型适用范围的基础上,对"当前"加速度的概率密度函数进行改进,得到一种修正的"当前"统计模型算法。为克服算法对加速度极限值的依赖,进一步提高跟踪精度,利用神经网络将2种参数信息融合,通过其输出对系统方差作加权调整。仿真结果表明,不论是对弱机动目标还是强机动目标,新算法较传统的算法都有较高的跟踪精度。
The current statistical model and adaptive Kalman filtering algorithm operates well in tracking strong maneuvering targets well but makes bigger error in tracking weak maneuvering targets. To solve this issue, an applicable bound of conventional current statistical model is derived. Based on the derivation, the current acceleration＇ s probability density function is improved and an improved current statistical model tracking algorithm is proposed. To overcome its limitation on acceleration and further im- prove its tracking precision, two sources of information are fused by using the neural network. Then the output of the network is used to adjust the system variance. Simulation results show that the proposed algorithm is better in tracking not only the weak maneuvering targets, but also the strong maneuvering targets.
出处
《现代防御技术》
北大核心
2011年第6期147-151,共5页
Modern Defence Technology
关键词
当前统计模型
机动目标跟踪
神经网络
信息融合
current statistical model
maneuvering targets tracking
neural network
information fusion
