| 摘要: | 過去十年之間，Mesosphere-Stratosphere-Troposphere(MST)雷達廣泛應用觀測大氣結構，例如大氣亂流(Turbulence)、大氣降水(Precipitation)、層狀結構以及大氣三維風場(3-Dimensional Wind Field)等，由於中壢特高頻(Very High Frequency, VHF)雷達站的觀測範圍位於飛機航道上，因此在觀測期間飛機飛越雷達波束時，則使得觀測資料內有飛機雜波(Clutter)在其中。
本文為利用小波轉換(Wavelet Transform, WT)以及希爾伯特-黃轉換(Hilbert-Huang Transform, HHT)作為基礎設計濾除飛機訊號之步驟，文中利用2017年5月15日至16日(UT)之雷達資料，比較利用兩種轉換所設計的步驟之濾除飛機訊號命中率與誤報率，而在結果中以利用希爾伯特-黃轉換作為基礎者命中率在十公里上下皆比以小波轉換作為基礎者高，並且在十公里以下命中率高達93.28%，誤報率僅有5.09%。
文末，我們將雷達資料通過希爾伯特-黃轉換所設計濾除飛機訊號步驟後，比較濾除飛機訊號前後之飛機回波功率，結果顯示飛機回波功率可濾除將近40dB左右，在飛機訊雜比部分各個高度飛機訊雜比在濾除飛機訊號後皆比濾除前皆明顯降低10-15dB左右，此外，我們將飛機回波功率進行高度之校正，可以發現飛機回波功率經由校正後趨勢變化較為一致，並且我們將實部與虛部本質模態函數C1至C4分量檢測到的飛機回波功率，與利用頻譜圖手動擬合之飛機回波功率進行比較，結果顯示其兩者趨勢相似，然而亦比較大氣平均都卜勒平移、頻譜寬以及大氣回波功率經由濾除飛機訊號步驟前後是否受到影響，在大氣平均都卜勒平移、頻譜寬部份，結果顯示濾除飛機訊號前後大氣平均都卜勒平移、頻譜寬差異皆小於0.05 Hz，然而我們的頻率解析度為0.0763 Hz，因此其值為可以接受之範圍，而大氣回波功率部分濾除飛機訊號前後改變並不大，然而大氣訊雜比的部分濾除飛機訊號後皆比濾除飛機訊號前大3dB左右。
;In the past decades, Mesosphere-Stratosphere-Troposphere radars have been widely applied to observe atmospheric structure and dynamics, such as turbulence, precipitation, layer structure, three-dimensional wind field and so on. Unwanted clutter were detected when observing atmospheric structure by Chung-Li VHF radar located near Taiwan Taoyuan International Airport.
The purpose of the following research is to investigate the removal of clutters based on wavelet transform (WT) and Hilbert-Huang Transform (HHT). Radar data from May 15th to 16th (UT) in 2017 are used to compare the success rate and error rate for effectiveness of both proposed methods in the removal of clutter. In the results, a 93.28% selectivity was observed below ten kilometers when HHT was employed, much higher than WT. In addition, a 5.09% false positive rate was only recorded for HHT below ten kilometers.
Besides that, airplane’s echo power from HHT-assisted clutter removal radar data was compared, and the result shows that about 40dB can be filtered out. In the airplane’s signal-to-noise ratio section, the result shows that about 10-15dB can be filtered out. Furthermore, A more consistent trend was obtained after the calibration on range for airplane echo power were made. Airplane echo power from C1 to C4 of Intrinsic Mode Function(IMF) and spectrum were also took into comparison, where it shows that the two airplanes’ echo power had similar trend . However, the effects of HHT-assisted method on other atmospheric parameters such as mean Doppler shift, spectrum width and air echo power were also discussed. The result shows that mean Doppler shift and spectrum width difference before and after filtering are less than 0.05Hz which is acceptable, the frequency resolution is 0.0763Hz. No significant changes in the air echo power were recoded. Finally, Atmospheric signal-to-noise ratio can be increased by about 3dB after filtered out airplane clutter. |
