会议专题

Sparsity-Enhanced Linear Time-Invariant MIMO System Identification

This paper addresses the problem of linear timeinvariant multi-input-multi-output (MIMO) system identi.cation. Specifically, we focus on identifying the finite impulse responses (FIRs) of a MIMO system. Observing that the FIRs are often approximately sparse, namely containing many nearzero elements, this paper proposes to use the l1 regularized least squares (l1-LS) method as the estimator. Comparing to the traditional identi.cation methods, such as least squares, the l1-LS method exploits the sparse nature of the FIRs, hence brings three advantages: (1) better estimation of the timedelays, (2) better estimation of the effective lengths of the FIRs, and (3) lower requirement of input-output data. Simulation results validate the ef.cacy of the proposed sparsity-enhanced identification approach.

multi-input-multi-output (MIMO) system identification finite impulse response (FIR) sparsity l1 regularized least squares (l1-LS)

Wei Shi Qing Ling Gang Wu

Department of Automation, University of Science and Technology of China, Hefei, Anhui 230027, P. R. China

国际会议

2011 China Control and Decision Conference(2011中国控制与决策会议 CCDC)

四川绵阳

英文

2026-2029

2011-05-23(万方平台首次上网日期,不代表论文的发表时间)