变尺度最小斜度UPF的Jerk模型机动跟踪研究  

Jerk Model Tracking of Scaled Minimal Skew Simplex UPF

作　　者：李黎[1,2] 刘忠[1] 张建强[1] 贺静波[1] LI Li;LIU Zhong;ZHANG Jianqiang;HE Jingbo(Naval University of Engineering,Wuhan 430033,China;Unit 92941,Huludao 125000,China)

机构地区：[1]海军工程大学,湖北武汉430033 [2]92941部队,辽宁葫芦岛125000

出　　处：《信息工程大学学报》2019年第3期286-290,共5页Journal of Information Engineering University

基　　金：国家自然科学基金资助项目(61401493)

摘　　要：为提高Jerk模型在复杂环境下对机动目标的跟踪能力,减小计算较精确机动目标状态的高阶模型的负担,提出变尺度最小斜度无味粒子滤波(Scaled Minimal Skew Simplex Unscented Particle Filter,SMSS-UPF)算法。SMSS-UPF在轻量级计算中解决非线性空间高维数滤波精度低的问题,同时满足重要性分布与后验概率密度的高重合性,可改善对弱Jerk模型机动的跟踪能力。仿真结果表明,SMSS-UPF能自适应逼近不同强度的Jerk机动进行跟踪,减小系统噪声方差和测量噪声方差带来的估计误差。与传统UPF相比,计算复杂度显著减小。A SMSS-UPF algorithm is proposed which employs a scaled minimal skew simplex unscented particle.It improves the Jerk model to track maneuvering target in a complex environment,and reduce the calculation burden about the more accurate state model of the high-order maneuvering target.Thanks to SMSS-UPF,high dimensional filtering accuracy in nonlinear space is improved.The introduced method can satisfy the high overlap of important distribution and the posterior probability density because of a particle filter.So the mobile tracking capability in Jerk model is stronger.For the adaptive approach of SMSS-UPF,simulation results confirm that the method can decrease the system noise variance and measurement noise variance estimation error against the traditional UPF,which also reduces the time complexity.

关 键 词：JERK模型 机动目标 变尺度最小斜度无味变换 粒子滤波 

分 类 号：TN953[电子电信—信号与信息处理]