交互多模型扩展卡尔曼滤波算法的FPGA实现

Implementation of IMMEKF algorithm on FPGA

对用于纯方位角度跟踪问题的交互多模型扩展卡尔曼滤波(IMMEKF)算法进行了研究,将其转换为适于硬件实现的形式。在此基础上,给出了基于现场可编程门阵列(FPGA)的实现方案,并讨论了硬件实现的资源优化和时间优化问题。软硬件仿真结果表明:IMMEKF硬件算法能够实现对单目标的纯方位角度跟踪,并且在保证与软件仿真具有相当精度的前提下能大幅减少运算时间。

Interactive multiple model extended Kalman filtering（IMMEKF） algorithm for bearings-only tracking is studied and transfer to the form suitable for hardware implementation. On this basis, an implementation scheme based on FPGA is proposed. Resource optimization and time optimization issues are discussed. Simulation results show that the hardware-based algorithm of＂ IMMEKF can realize bearings-only tracking on single target, and operating time is sharply reduced while guarantee the precision is comparative to software simulation.