基于二阶差分滤波器的水下目标纯方位角跟踪  

Research on second order divided difference filter algorithm for underwater target bearing-only tracking

作　　者：王宏健[1] 徐金龙[1] 么洪飞[1,2] 张爱华 

机构地区：[1]哈尔滨工程大学自动化学院,黑龙江哈尔滨150001 [2]齐齐哈尔大学计算机与控制工程学院,黑龙江齐齐哈尔161006

出　　处：《哈尔滨工程大学学报》2014年第1期87-92,共6页Journal of Harbin Engineering University

基　　金：国家自然科学基金资助项目(E091002/50979017);教育部高等学校博士学科点专项科研基金资助项目(20092304110008);哈尔滨市科技创新人才(优秀学科带头人)研究专项基金资助项目(2012RFXXG083);教育部新世纪优秀人才支持计划基金资助项目(NCET-10-0053)

摘　　要：针对被动单声呐平台对水下目标的纯方位角跟踪问题,提出基于二阶差分滤波器的水下目标纯方位角跟踪方法。采用二阶Stirling插值公式对系统模型中的非线性项进行线性化,使系统误差协方差矩阵正定,保证了滤波算法的稳定性,避免了传统扩展卡尔曼滤波算法由于需要计算Jacobian矩阵而导致计算复杂、难以应用的问题。建立了被动声呐平台测量模型和水下目标的运动学模型,并应用Monte Carlo方法完成仿真实验。仿真结果表明,基于二阶差分滤波器的水下目标纯方位角跟踪算法具有较快速准确的跟踪响应,通过RMSE概率统计方法进一步验证了这种方法具有较高的估计精度。In order to deal with the problems concerning the single passive sonar based underwater target bearingonly tracking, the underwater target bearing-only tracking algorithm based on the second order divided difference filter has been proposed. This algorithm uses the second order Stirling interpolation formula to linearize the nonlinear terms of the system models, which makes the system error covariance matrix positively definite, ensures the stability of the filtering algorithm, and makes it easy to implement a comparison with the traditional extended Kalman filtering algorithm due to the avoidance of calculating the Jacobian matrix. The underwater target kinematics model and the passive sonar measurement model were established, and the Monte Carlo method was applied to complete the simulation experiment. The simulation results show that the second order divided difference filter based on the underwater target bearing-only tracking algorithm has a fast and accurate tracking response. The RMSE probability statistical method was used to further validate that this method has a higher estimation precision.

关 键 词：水下目标 纯方位角跟踪 二阶差分滤波器 扩展卡尔曼滤波器 MONTE Carlo仿真 

分 类 号：U631.1[交通运输工程—船舶与海洋工程]