| 摘要: | 本研究透過全波方法中的矩量法針對合成孔徑雷達回波訊號進行模擬，並透過改進後的成像算法進行聚焦處理。所得的聚焦圖像相 較於傳統方法不同之處在於可完整描述觀察目標物的散射電磁特性、 評估多徑效應、驗證系統性能以及研究極化信息等優點。利用麥克斯 韋方程和邊界條件，所推導出的波動方程式，可用於描述金屬和介電 質物體的電磁行為。透過三角形面所定義的基礎函數及測試函數，可 將其用於將積分方程離散為矩陣方程。對於矩量法矩陣方程求解，採 用雙共軛梯度法並結合預條件加速基函數 MLFMA 迭代方式求解。在計 算出等效電流密度之後，散射場可以在天線照射區域內進行積分。通 過在不同位置所得的散射場進行合成孔徑過程，可以生成回波數據。 同時，可以通過改變發射和接收天線的極化來獲取全極化的回波數據。 然後通過改進後的 Omega-K 算法對全極化的原始數據進行濾波處理 可獲得全極化合成孔徑聚焦圖像。
本方法成功地演示了在不同材料的散射特性、多物體之間的相互 作用、單站和雙站系統的仿真結果比較以及在不同帶寬下的影響討論。 利用不同的介電常數材料，如聚四氟乙烯、陶瓷和砷化鎵，在聚焦圖 像中可表現出不同的散射特性。通過改變兩個目標之間的間距，呈現 出不同程度的相互作用。並且探討了在不同帶寬下散射特性和對於成
像特性的影響。在幾種情況下，討論了單站和雙站之間的散射行為。 通過所提出的方法中發射和接收天線的極化變化，可以生成全極化 SAR 圖像。這些數據可用於極化分解方法比較和誤差分析。
為了驗證所提方法的有效性和性能，在無回波室內進行了一組測 量。透過個人計算機對於雷達收發系統和運動控制器進行同步以獲取 回波數據和位置向量。將前述資料透過成像算法，可以獲得最終的聚 焦圖像。結果顯示所提出的方法與實驗結果十分吻合，成功地證明了 所提方法的可靠性。
透過數值電磁學分析方法中矩量法具有高理論精度的特點，本方 法更完整的描述電磁波與目標物的散射訊息，並成功連結了基於電磁 理論的回波訊號及成像算法的完整過程。此方法可用於未來新型合成 孔徑雷達系統設計、不同觀測目標物散射分析討論。;In this dissertation, based on electromagnetics theory, a complete link between the echo signal of synthetic aperture radar (SAR) and imaging processing is established. Inspired by the method of moment (MoM), which has high theoretical accuracy in computational electromagnetic analysis methods, a SAR raw data simulator based on the MoM is different from the conventional point-target model, which is simply a pure mathematical model. Using Maxwell’s equations and boundary conditions, the electromagnetic wave integral equations are derived to formulate electromagnetic characteristics involving the interaction between metallic and dielectric objects. The simultaneous discretization of the integral equations and surfaces using a set of basis and testing functions defined with triangles leads to dense matrix equations. To accelerate the calculation of large dense matrix equations, they are solved iteratively via the multilevel fast multipole algorithm (MLFMA). After computing the equivalent current density, the scattering field can be integrated over the antenna-illuminated area. By utilizing the synthetic aperture process within different positions, echo data can be generated. In addition, fully polarized echo data can be acquired by changing the polarization of the transmitting and receiving antennas. The fully polarized focus SAR image can be obtained by the refined Omega-K algorithm. The results can be utilized to discuss the scattering mechanism, evaluate the effects of multipath, validate the system performance, and investigate the polarization information.
The proposed method successfully demonstrates comparisons of the simulation results for different materials, interactions between objects, and between monostatic and bistatic systems, and the results at different bandwidths. Different dielectric constant materials, such as Teflon, ceramics glaze (CG) and gallium arsenide (GaAs) are utilized to present the abilities of different scattering characteristics in the focus images. By varying the spacing between two targets, the different degrees of interaction are presented. The effects of scattering characteristics and imaging properties under different bandwidths are investigated. The scattering behaviors between monostatic and bistatic systems are discussed in several cases. Through polarization changes in transmission and receiving antennas with the proposed method, fully polarimetric SAR images can be generated. These data can be used for a comparison with the polarization decomposition method and for error analysis. To validate the effectiveness and performance of the proposed method, a set of measurements was obtained in an anechoic chamber. A network analyzer and a motion controller synchronized by a personal computer were utilized to acquire the echo data and the position vector simultaneously. Then, applying the imaging algorithm to the raw data, the final focus images were obtained. The results from the full-wave method compare well with the experimental results, which successfully demonstrates the reliability of the proposed method.
In conclusion, the simulation of SAR imags based on the MoM and the refined Omega-K algorithm offers more complete recovery of scattering information. The proposed method is useful in designing future novel SAR systems and the analysis of the electromagnetic characteristics of different observation targets. |
