本文關鍵詞：星上SAR實時成像處理關鍵技術研究　出處：《北京理工大學》2016年博士論文　論文類型：學位論文

  更多相關文章： 星載SAR 星上實時處理 FPGA CS因子 矩陣轉置 CORDIC算法

【摘要】：隨著星載SAR成像技術的不斷進步和應用需求的增加,在星上完成SAR實時成像處理已成為一種發(fā)展趨勢。星載SAR成像算法復雜、處理量大,在星上環(huán)境對系統(tǒng)體積、重量、功耗的嚴格約束下,星上SAR處理系統(tǒng)的實時性難以保證。為了在單片F(xiàn)PGA上實現(xiàn)星載SAR成像所有處理流程,構建星上SAR實時成像處理系統(tǒng),本文重點研究了星載SAR成像算法的實時性優(yōu)化、SAR成像處理矩陣轉置存儲方法以及星載SAR成像算法中的多種超越函數(shù)的實現(xiàn)。論文主要工作及創(chuàng)新成果如下:1.對星載SAR成像算法、星上SAR實時成像核心處理器和系統(tǒng)架構進行研究。選擇了以FPGA為核心處理器實現(xiàn)CS成像算法作為構建星上SAR實時成像處理系統(tǒng)的技術途徑,提出了標準化、模塊化、可擴展的星上SAR實時成像處理系統(tǒng)架構。2.針對星載SAR成像算法在FPGA實現(xiàn)中遇到的運算量大、實時性不足等問題,對多普勒參數(shù)估計方法進行選擇和優(yōu)化,研究了相位補償因子對成像精度和實時性的影響,提出了一種補償因子區(qū)域不變的CS成像算法。該方法在保證星載SAR算法成像質量的前提下,能夠有效減小補償因子的計算量,提高SAR成像算法處理效率。3.針對SAR成像處理過程中轉置存儲器的數(shù)據(jù)訪問效率過低的問題,對SAR成像處理矩陣轉置存儲方法進行研究,提出了矩陣分塊三維映射法和矩陣分塊交叉映射法。這兩種轉置存儲方法充分利用了轉置存儲器的數(shù)據(jù)訪問時序特性,能夠有效提高SAR成像處理過程中轉置存儲器的數(shù)據(jù)訪問效率,提高SAR成像處理實時性。4.針對星載SAR成像算法中多種超越運算的FPGA實現(xiàn)占用資源過大的問題,對實現(xiàn)正余弦、反正切函數(shù)的傳統(tǒng)CORDIC算法進行了角度覆蓋范圍、硬件資源、運算精度等方面的優(yōu)化和改進,同時對實現(xiàn)反正余弦函數(shù)的雙迭代結構的CORDIC算法進行了改進,解決了原算法計算結果超出反正余弦函數(shù)值域的問題,進而實現(xiàn)了一種多模式CORDIC算法,能夠實現(xiàn)正余弦函數(shù)和反正余弦函數(shù)運算的實時切換,在正余弦、反正余弦函數(shù)都需要的算法實現(xiàn)中能夠有效減少硬件資源占用。5.本文最后結合各章節(jié)研究內容,在單片F(xiàn)PGA上實現(xiàn)了星載SAR成像算法所有處理流程,進而構建了星上SAR實時成像處理系統(tǒng)原理樣機,為實現(xiàn)星上SAR實時成像處理奠定了技術基礎。
[Abstract]:With the continuous progress of satellite borne SAR imaging technology and the increase of application demand, it has become a development trend to complete SAR real-time imaging processing on the star. The spaceborne SAR imaging algorithm is complex and has a large amount of processing. Under the strict constraint of the volume, weight and power consumption of the satellite environment, the real-time performance of the SAR processing system on the satellite is difficult to guarantee. In order to realize on a single FPGA for Spaceborne SAR imaging of all processes, construction of spaceborne SAR real-time image processing system, this paper focuses on the realization of a variety of spaceborne SAR imaging algorithm, the real-time optimization of SAR imaging matrix transpose storage method and spaceborne SAR imaging algorithm of transcendental function. The main work and innovative results of this paper are as follows: 1. the research on the satellite borne SAR imaging algorithm, the core SAR real-time imaging core processor and the system architecture are carried out. The CS imaging algorithm based on FPGA as the core processor is chosen as the technical way to build real-time SAR imaging processing system. A standardized, modular and scalable SAR real-time imaging processing system architecture is proposed. Computation of 2. for Spaceborne SAR imaging algorithms encountered in the implementation of FPGA, the problem of real-time, the estimation of Doppler parameters selection and optimization method, studied the effect of phase compensation factor to the imaging accuracy and real-time, proposes a compensation factor invariant CS imaging algorithm. This method can effectively reduce the computation of compensation factor and improve the processing efficiency of SAR imaging algorithm on the premise of guaranteeing the imaging quality of the spaceborne SAR algorithm. 3., in view of the low efficiency of data access in transposed memory in the process of SAR image processing, we studied the matrix transposition storage method of SAR imaging processing, and proposed the matrix block three dimensional mapping method and the matrix block cross mapping method. The two transposition storage methods make full use of the data access timing characteristics of the transposed memory, which can effectively improve the data access efficiency of the transposed memory in the process of SAR imaging, and improve the real-time performance of SAR imaging. 4. for a variety of operations beyond the spaceborne FPGA SAR imaging algorithm to achieve resource is too large, the traditional CORDIC algorithm to realize the sine cosine and arctangent function is optimized and improved angle of coverage, hardware resources, operation accuracy, and implementation of the CORDIC double iterative algorithm of structure of sine and cosine functions the improved algorithm solves the original results beyond the range anyway cosine function, and implements a multi mode CORDIC algorithm can realize sine cosine function and cosine function anyway real-time switch operation, in the realization of sine and cosine, anyway, cosine function needs algorithm to reduce the hardware resource occupied. 5., finally, combined with the research contents of each chapter, we implemented all processing flow of spaceborne SAR imaging algorithm on single chip FPGA, and constructed the prototype of SAR real-time imaging processing system, which laid the technical foundation for real-time SAR imaging processing.
【學位授予單位】：北京理工大學
【學位級別】：博士
【學位授予年份】：2016
【分類號】：TN957.52

【參考文獻】

相關期刊論文 前10條

1 張召才;;國外商用現(xiàn)貨技術在空間任務中的發(fā)展與應用[J];衛(wèi)星應用;2015年01期

2 閆雯;陳禾;謝宜壯;劉小寧;;一種補償因子區(qū)域不變的CS成像算法[J];北京理工大學學報;2014年03期

3 邊明明;李盛林;岳榮剛;;星載SAR在軌成像實時處理算法研究[J];航天器工程;2013年06期

4 王峰;郭金生;李暉;;商用現(xiàn)貨器件在衛(wèi)星中的應用[J];航天器工程;2013年04期

5 胡海鳳;趙建平;吳冬梅;路娜;秦永益;;基于FPGA的反余弦函數(shù)的實現(xiàn)[J];電子技術;2013年06期

6 鮑勝榮;周海斌;;SAR實時成像高效矩陣轉置研究和實現(xiàn)[J];現(xiàn)代雷達;2013年03期

7 肖靖;胡學成;章宏;陳知明;;基于最小熵的斜視SAR多普勒調頻率估計[J];現(xiàn)代雷達;2013年01期

8 邊明明;畢福昆;汪精華;;實時SAR成像系統(tǒng)矩陣轉置方法研究與實現(xiàn)[J];計算機工程與應用;2011年22期

9 高許崗;蘇衛(wèi)民;顧紅;;一種基于最小熵準則的新的自聚焦優(yōu)化算法[J];兵工學報;2010年12期

10 謝宜壯;朱柏承;汪精華;;星載SAR實時成像處理系統(tǒng)結構設計[J];兵工學報;2010年S2期

，

本文編號：1341244