Correction of various environmental influences on Doppler wind lidar based on multiple linear regression model
Correction of various environmental influences on Doppler wind lidar based on multiple linear regression model
复制标题
基于多元线性回归模型修正各种环境对多普勒测风激光雷达的影响
DOI：
10.1016/j.renene.2021.12.018
复制
发表时间：
2022-01-01
期刊：
影响因子：
8.7
通讯作者：
Qu, Hongya
中科院分区：
文献类型：
--
作者：
Tang, Shengming;Li, Tiantian;Qu, Hongya
多普勒测风激光雷达（DWL）在风能、气象、航空等领域的应用日益广泛，人们对其精度和可靠性进行了大量的研究，并与安装在气象塔上的风速计进行了比较。然而，由于气象塔的高度有限，以前的检查主要集中在100米的范围内。为了进一步验证DWL的性能，特别是在100 m以上的高度，在中国的两个国家气象观测站（深圳和锡林浩特）进行了实验测试。深圳天文台的气象塔高356 m，可以验证100 m以上的DWL。不同的环境变量，包括湿度，降水，风的特性，和表面粗糙度长度，进行了研究，以量化其对DWLs的测量误差的影响。提出了一种基于多元线性回归模型的测量误差修正方法，消除了环境因素引起的测量误差。校正后的DWL数据可以改善高达9.6%的DWL和塔数据之间的线性回归的斜率，相关的均方根误差可以减少高达37%。(c)2021爱思唯尔有限公司保留所有权利。
Doppler wind lidar (DWL) is being increasingly employed in various areas, such as wind energy, meteorology, aviation, and so on. Extensive studies have been conducted to validate its accuracy and reliability compared with anemometers mounted on meteorological towers. However, previous examinations mainly focused on a range up to 100 m because of the limited heights of meteorological towers. To further validate the DWL performance, especially above a height of 100 m, experimental tests were carried out at two national meteorological observatories in China (Shenzhen and Xilinhaote). The meteorological tower at Shenzhen Observatory is 356 m high, which enables validation of DWLs above 100 m. Different environmental variables, including humidity, precipitation, wind characteristics, and surface roughness length, were investigated to quantify their effects on the measurement errors of DWLs. Moreover, a correction methodology based on multiple linear regression model was proposed to eliminate the measurement error induced by environmental conditions. The corrected DWL data can be improved by up to 9.6% regarding the slope of the linear regression between the DWL and tower data, and the associated root mean square errors can be reduced by up to 37%.(c) 2021 Elsevier Ltd. All rights reserved.