基于FPGA的高精度惯导算法实现  

Implementation of high precision inertial navigation algorithm based on FPGA

作　　者：池乃权 武元新[1] CHI Nai-quan;WU Yuan-xin(Shanghai Jiaotong University,Shanghai Key Laboratory of Navigation and Location-based Services,Shanghai 200240,China)

机构地区：[1]上海交通大学,上海市北斗导航与位置服务重点实验室,上海200240

出　　处：《信息技术》2021年第6期33-39,共7页Information Technology

基　　金：教育部联合基金项目(6141A02022309)。

摘　　要：通过将函数迭代积分和切比雪夫多项式结合,iNavFIter算法首次将运动不可交换性误差降低到了机器截断误差级别,为新一代高精度或高动态惯性导航系统奠定了算法基础。文中利用了FPGA并行度高、灵活性强的特点,在硬件上对iNavFIter算法进行了实现,并对算法的时序和资源利用率进行了优化。该设计使用Xilinx的Kintex-7型号FPGA开发板,将算法打包成IP核部署到FPGA上,并通过以太网与上位机进行通信,接收上位机发送的陀螺加速度计数据,并将解算结果发送给上位机进行结果展示。惯导仿真数据表明,设计的FPGA IP核具备了高精度、实时计算的能力。By combing functional iterative integration and Chebyshev polynomial,the inertial navigation computation of iNavFIter for the first time reduces the motion non-commutativity error to the level of machine truncation error,and thus,lays a solid algorithmic base for new-generation precision/dynamic inertial navigation systems.This article implements the iNavFIter algorithm by way of FPGA,and take advantages of its high parallelism and flexibility,in which the timing and resource utilization are specially optimized.This design uses an Xilinx s Kintex-7 FPGA development board,and communicates with the host computer via Ethernet for sensor data input and computation result output.Tests with simulated inertial sensor data show that this hardware system is capable of real-time computation at precision accuracy.

关 键 词：iNavFIter FPGA 高精度 实时性 

分 类 号：U666.12[交通运输工程—船舶及航道工程]