基于变形球面分布假设的全姿态高精度磁电子罗盘误差补偿研究

项目名称： 基于变形球面分布假设的全姿态高精度磁电子罗盘误差补偿研究

项目编号： No.61273082

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 自动化技术、计算机技术

项目作者： 李希胜

作者单位： 北京科技大学

项目金额： 81万元

中文摘要： 本项目针对引起磁电子罗盘测量误差的硬磁干扰、软磁干扰等因素展开深入研究。研究在载体绕X、Y、Z三轴之一发生转动情况下磁场测量信号所受影响，分析磁场测量信号与地磁场矢量、载体硬磁干扰矢量之间的关系，找出其中引起航向测量误差的非线性关系，建立此种情况下的磁场测量方程。以在载体绕X、Y、Z三轴之一发生转动情况下所得到的磁场测量方程为基础，通过转动过程分解，建立全姿态范围内磁电子罗盘磁场测量方程。最终在磁场测量方程基础上提出基于变形球面分布假设的全姿态高精度磁电子罗盘两步式误差补偿算法以实现航向精确测量。第一步基于椭球面拟合对磁场测量值进行粗修正；第二步基于变形球面分布假设对磁场测量值进行细修正，解决基于椭球面分布假设的校准算法的剩余误差问题。在地磁场高精度测量基础上实现航向高精度测量，研制在全姿态范围内精度达到±0.3 (最大值)的高精度磁电子罗盘。

中文关键词： 磁电子罗盘；椭球拟合；磁场畸变；非线性误差；超限学习机

英文摘要： The research focuses on the factors which cause errors to magnetic electronic compass measurement, such as hard iron magnetism (induced magnetism) effect and soft iron magnetism (permanet magnetism) effect. Influence on measured value of magnetic field caused by these factors when the carrier rotating around one of its three axes(X, Y and Z)will be studied firstly. The relationship of the measured magnetic field value to geomagnetic field vector and carrier's hard iron magnetic field vector will be analyzed. As a result, nolinear relationship causing heading measurement error will be found and magnetic field measurement equations will be established. And then on the basis of the measurement equation when carrier rotating around one of the X, Y and Z axis of carrier, the all attitude magnetic field measurement equation of the electronic compass will be established by disintegrating the rotation process.Finally, a two-step all attitude high-precision error compensation algorithm based on the deformed spherical distribution assumption will be presented. The first step of the algorithm is roughly revising the measured magnetic field value by ellipsoid fitting; the second step is to correct the measured value finely based on the deformed spherical distribution assumption.This research aims at realizing high-precisi

英文关键词： magnetic electronic compass；ellipsoid fitting；magetic field distortion；non-linear error；extreme learning machine algorithm