应用反卷积实现脉冲计量中的信号重构与系统辨识  

Signal Reconstruction and System Identification Using the Deconvolution in Pulse Metrology

作　　者：王园[1] 朱江淼[1] 

机构地区：[1]北京工业大学电子信息与控制工程学院,北京100124

出　　处：《信号处理》2013年第4期532-535,共4页Journal of Signal Processing

摘　　要：本文研究了在脉冲计量领域中的基于反卷积算法的信号重构与系统辨识。我国目前的脉冲参数国家基准过渡时间为7.25±0.50ps,带宽为56.3±5.8%GHz。如果采用传统的计量方法,该基准无法实现对标称带宽为70GHz的Agilent86100C高速取样示波器以及标称过渡时间为5ps的Picosecond4016负阶跃脉冲源进行校准。本文将反卷积算法引入到测量数据处理中,并结合具体问题对传统的反卷积算法进行了改进,克服了传统的反卷积算法对阶跃信号计算误差过大的问题,并进行了实验验证,实现了量值传递。This paper studied the signal reconstruction and system identification based on deconvolution algorithm in pulse metrology. The transition duration of pulse parameter national standard of China is 7.25＋0.50ps, and bandwidth is 56.3＋5.8% GHz. High speed sampling oscilloscope Agilent86100C has the bandwidth of 70GHz and minus step pulse source Picoseeond4016 has the transition duration of 5ps, according to traditional calibration method, these two instrue- ments could not be calibrated by the national standard. The deconvolution algorithm was used in the processing of measure- ment data in this paper, and conventional deconvolution method was improved to overcome the large errors while step signal was deconvoluted. Experiment verified the new algorithm, realized the value transfer.

关 键 词：反卷积 校准 信号重构 系统辨识 

分 类 号：TM935.38[电气工程—电力电子与电力传动]