FAST QR decomposition algorithm of block Hankel matrix in subspace system identification method

摘要：

Block Hankel matrix projection in subspace system identification method can be computed via the QR decomposition. Without consider the special structure of the block Hankel matrix, the QR decomposition of it has the computation complexity of 2(nq)2(mp-nq/3) for a mp×nq matrix. A fast QR decomposition algorithm is presented by using the hyperbolic Householder transformation with considering the special structure of the block Hankel matrix. The computation complexity is reduced to nq2(16mp+9np+5nq)/2. The numerical test of a (1000×2)×(40×5) block Hankel matrix show that the time cost of the fast algorithm is 1/68 times of the ordinary algorithm.

关键词： subspace system identification QR decomposition hyperbolic Householder transformation

作者: Chun Yang Jinping Ou

作者单位: School of Civil Engineering,Harbin Institute of Technology,Harbin,150090,China School of Civil Engineering,Harbin Institute of Technology,Harbin,150090,China;School of Civil & Hyd

会议类型: 国际会议

会议名称: 第六届国际振动工程会议(The 6th International Conference on Vibration Engineering)(ICVE’ 2008)

会议地点: 大连

会议语种:英文

在线出版日期: 2008-06-04（万方平台首次上网日期，不代表论文的发表时间）