一种稳健的多传感器目标跟踪算法  

A Robust Tracking Algorithm for Multiple-Sensor

作　　者：王燊燊[1] 冯金富[1] 王方年[2] 黄峰[1] 张佳强[1] 

机构地区：[1]空军工程大学工程学院,陕西西安710038 [2]桂林空军学院,广西桂林541010

出　　处：《现代防御技术》2011年第6期157-162,共6页Modern Defence Technology

基　　金：国家自然科学基金(60772162);国家"863"高技术研究发展计划(2010A×××4C)

摘　　要：研究了一种基于双波段红外和雷达传感器的自适应目标跟踪算法。利用目标多普勒信息和红外辐射信息,建立具有树形结构的红外雷达跟踪系统状态估计模型,采用平方根无迹卡尔曼滤波方法实现双波段红外传感器的初步融合,并将其作为融合节点与雷达传感器进行基于自适应交互式多模型算法的深层融合。在交互式多模型算法的基础上,研究了寻求快速准确的机动目标模型匹配方法,通过局部优化来达到最终融合结果的优化。仿真结果表明,该算法能有效提高机动目标跟踪精度,具有一定的稳健性和自适应性。An adaptive tracking algorithm is presented based on radar and double wave band infrared seeker. The state estimation model with tree-structure radar and infrared seeker fusion algorithm is con- structed using the infrared irradiation and the Doppler information of target. A new adaptive fusion algo- rithm for two different infrared seekers based on square root unscented kalman filters is presented, the re- sult of which is fused with the measurement of an active radar sensor sequentially to estimate the final tar- get state using adaptive interacting multiple model method. The exact model matching method is studied and optimization of the final fusion is obtained by local optimization. Simulation results show that the fusion algorithm can improve tracking precision effectively so it is robust and adaptive.

关 键 词：双波段 多传感器融合 平方根无迹卡尔曼滤波 自适应交互式多模型 

分 类 号：TN953[电子电信—信号与信息处理] TP212.9[电子电信—信息与通信工程]