基于FLANN的磁通门总场梯度计误差校正  

Correction of fluxgate total field gradiometer based on FLANN

作　　者：黄玉[1,2] 郝燕玲 

机构地区：[1]哈尔滨工程大学理学院,黑龙江哈尔滨150001 [2]哈尔滨工程大学自动化学院,黑龙江哈尔滨150001

出　　处：《华中科技大学学报（自然科学版）》2011年第5期83-87,共5页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(60775001;61004130;60834005)

摘　　要：针对磁通门磁力计轴间非正交、对应轴指向偏差、灵敏度不一致及零点漂移所引起的误差会降低总场梯度计的测量精度问题,建立了磁通门总场梯度计的测量误差模型,提出了一种基于函数链接型神经网络(FLANN)的总场梯度计误差校正方法,将2个磁场力计实际输出分别通过校正模型以构成理想的2个三轴磁力计,使输出与待测量一致.数值仿真及实验测试均证明了校正方法具有良好的收敛性,能显著地抑制总场梯度测量误差.The errors caused by nonorthogonality, different sensitivities, and zero-shifts among three axes and inconsistent direction of corresponding axis for fluxgate magnetometers can reduce the accuracy of total field gradiometer, so it is necessary to correct these errors and to compensate total field gradient measured by imperfect gradiometer. A measurement error model for fluxgate total field gra- diometer was established, and its error correction method based on functional link artificial neural net- work （FLANN） was proposed. The numerical simulations and experimental tests proved a good convergence of the algorithm, which can depress remarkably the measurement error of total field gradient. The research can provide a feasible way to improve the performance of total field magnetic gradiometer.

关 键 词：总场梯度计 误差校正 三轴磁力计 函数链接型神经网络 辨识 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]