磁强计
粒子群优化
校准
电磁线圈
偏移量(计算机科学)
Levenberg-Marquardt算法
算法
趋同(经济学)
物理
控制理论(社会学)
数学
计算机科学
磁场
统计
人工智能
量子力学
人工神经网络
经济增长
经济
程序设计语言
控制(管理)
作者
Siran Li,Danyue Ma,Jixi Lu,Bozheng Xing,Kun Wang,Yanan Gao,Bangcheng Han
出处
期刊:Measurement
[Elsevier]
日期:2022-10-01
卷期号:202: 111878-111878
被引量:16
标识
DOI:10.1016/j.measurement.2022.111878
摘要
A novel in situ method has been demonstrated for calibrating parameters (coil constants and nonorthogonal angles) of triaxial coils of a three-axis optically pumped magnetometer (OPM). In total, 12 error sources including nonorthogonal errors, proportional coefficient errors and zero-offset errors have been considered into the calibration model. We utilize a particle swarm optimization (PSO) algorithm to estimate the errors, and compare it with a regular Levenberg-Marquardt (LM) algorithm. The results demonstrate that the PSO algorithm has higher accuracy, better convergence stability, faster convergence rate and less calibration uncertainty. The calibration realizes the simultaneous measurement of the coil parameters and the correction of the 12 error sources. The coil constants are (88.93 ± 0.22) nT/mA, (88.78 ± 0.12) nT/mA, and (197.64 ± 0.29) nT/mA which deviate 1.2–1.4 % from the theoretical values, respectively. The nonorthogonal angles are (89.50 ± 0.06) °, (89.46 ± 0.04) ° and (89.40 ± 0.02) ° which correct the errors to 19.6–30.7 % of the original values, respectively. This study realizes in situ coil calibration without auxiliary devices and can be applied to most types of three-axis OPMs including the miniature magnetometers with higher accuracy.
科研通智能强力驱动
Strongly Powered by AbleSci AI