| 基于高斯拟合的相干激光雷达风速估计算法 |
| |
| 引用本文： | 王平春,陈廷娣,周安然,韩飞,王元祖,孙东松,王国成.基于高斯拟合的相干激光雷达风速估计算法[J].红外与激光工程,2018,47(12):1230006-1230006(6). |
| |
| 作者姓名： | 王平春 陈廷娣 周安然 韩飞 王元祖 孙东松 王国成 |
| |
| 作者单位： | 1.中国科学技术大学 地球和空间科学学院,安徽 合肥 230026; |
| |
| 基金项目： | 国家自然科学基金（41574180，41374156，41204108） |
| |
| 摘 要： | 分别利用高斯拟合估计算法（Gaussian fitting estimation algorithm，以下简称Gauss估计算法）和最大似然（Maximum Likelihood，ML）离散谱峰值（Discrete Spectral Peak，DSP）估计算法（ML DSP）处理实测回波信号，计算得到风速扰动的功率谱密度（Power Spectral Density，PSD）。根据Kolmogorov湍流理论中PSD与频率的-5/3关系，比较不同距离门下的PSD，采用高频区域的风速误差作为风速估计性能参数，分析比较不同距离情况下风速误差，并利用自相关系数分析风速时间变化的相关性。结果表明:在距离较低的探测区域Gauss估计算法的风速误差微弱小于对应的ML DSP估计算法，二者之间的风速误差差值最多不超过0.05 m/s。而在距离较高的区域，两种算法的风速误差差值从820 m处的0.06 m/s增加至1 200 m的0.16 m/s。在风速的时间相关性分析上，Gauss估计算法的风速时间自相关系数明显大于对应的ML DSP估计算法，说明Gauss估计算法处理的风速数据更具有稳定性。
|
| 关 键 词： | 相干激光雷达 风速估计 最大似然估计 高斯拟合 |
| 收稿时间： | 2018-07-05 |
Wind velocity estimation algorithm based on Gaussian fitting in coherent lidar |
| |
| Affiliation: | 1.School of Earth and Space Science,University of Science and Technology of China,Hefei 230026,China;2.New Star Institute of Applied Technology,Hefei 230031,China |
| |
| Abstract: | The power spectral density(PSD) of the wind velocity disturbance was calculated by processing the measured echo signals by using Gaussian fitting estimation algorithm and maximum likelihood (ML)discrete spectral peak(DSP) estimation algorithm respectively. According to Kolmogorov turbulence theory, PSD has the relationship of -5/3 slope of frequency. It could be compared by different PSD under different distance gates. Wind velocity error in the high frequency region was used as the parameter of wind velocity estimation for comparing performance, and the error under different distances was analyzed and compared. The correlation of the relationship between wind velocity and time series was analyzed by using the autocorrelation coefficient. The results show that the wind velocity error of Gaussian fitting estimation algorithm is less than that of the corresponding ML DSP estimation algorithm in the low detection area, and the difference between the two wind speed errors does not exceed 0.05 m/s. In the area with higher distance, the difference of wind velocity error between the two algorithms increases from 0.06 m/s at 820 m to 0.16 m/s at 1 200 m. In the time-dependent analysis of the wind velocity, the autocorrelation coefficient of Gaussian fitting estimation algorithm between wind velocity and time is significantly larger than that of the corresponding ML DSP estimation algorithm, which shows that the wind velocity data processed by Gaussian fitting estimation algorithm is more stable. |
| |
| Keywords: | |
|
| 点击此处可从《红外与激光工程》浏览原始摘要信息 |
|
点击此处可从《红外与激光工程》下载全文 |
|
