算法
对数
探测器
计算机科学
稳健性(进化)
偏移量(计算机科学)
观测误差
光学
数学
物理
电信
生物化学
基因
统计
数学分析
化学
程序设计语言
作者
Jun Zhang,Weixian Qian,Guohua Gu,Mao Chen,Kan Ren,Chunhong Wu,Xiaoyu Peng,Qian Chen
出处
期刊:Applied Optics
[The Optical Society]
日期:2019-09-20
卷期号:58 (28): 7741-7741
被引量:9
摘要
A four-quadrant detector is a kind of photoelectric detector that can quickly and accurately measure the incident angle of light. However, its ability to measure in a large field of view (FOV) is limited by its hardware structure and its calculation principle. To solve these problems, this paper proposes an improved algorithm that can extend the measurement linear range without reducing its measurement accuracy. After that, through simulation and experiment, we compare it with many other location algorithms, including the most widely used classical algorithm and the logarithmic algorithm suitable for large FOVs. Finally, the following conclusions can be drawn from both theoretical data and experimental data: the improved algorithm can significantly improve the measurement accuracy over 50% in the same FOV condition, and the measurable range can be expanded over 25% in the same accuracy requirement. At the same time, the robustness of noise does not decrease; when the root mean square error of the classical algorithm fluctuates at 0.1° in different SNR conditions, the improved algorithm is also 0.1°, while the logarithmic algorithm can reach 1.7°, and other algorithms are around 0.25°. In addition, the improved algorithm is more stable in measuring a certain direction and can effectively avoid the influence from the offset of incident light in another axis.
科研通智能强力驱动
Strongly Powered by AbleSci AI