为了实现大气水平能见度的快速测量,设计了基于电荷耦合器件(CCD)的大气水 平能见度测量系统,并对该系统采用的局部加权回归算法进行了研究。通过实验对比了局部 加权回归和普通线性回归算法的拟合直线,并与商用NQ-1能见度仪进行了同步对比测量。 结果表明:使用局部加权回归替换普通线性回归算法的CCD测量系统可以提高与能见度仪测 量数据的相关性,相关系数从85%提高到96%;与能见度仪测量数据的实时对比表明,系统测 量的误差值从3 km减小到0.8 km左右。可见采用局部加权回归算法的CCD测量系统基本满足 大气水平能见度测量的要求。

In order to realize the rapid measurement of atmospheric horizontal visibility, the atmospheric horizontal visibility measurement system based on charge-coupled device (CCD) is designed, and the local weighted regression algorithm adopted in the system is investigated. The fitting lines of the local weighted regression and ordinary linear regression algorithms are compared by experiments, and the synchronous contrast measurement with the commercial NQ-1 visibility meter is carried out. Results show that the CCD measurement system with local weighted regression algorithm instead of ordinary linear regression algorithm can improve its correlation with the measurement data of visibility meter, and the correlation coefficient can increase from 85% to 96%. The real time comparison with the measurement data of the visibility meter shows that the error value of measurement system is reduced from 3 km to about 0.8 km. It can be found that the CCD measurement system using the local weighted regression algorithm basically meets the requirement of atmospheric horizontal visibility measurement.