System Identification Using Embedded Dynamic Signal Analyzer

Abstract

The effect of model uncertainty is increased in optical disc drives (ODDs) because the mechanical and electrical components are manufactured with a wide error tolerance to reduce production costs. In this paper, we introduce a new technique for system identification whereby an embedded dynamic signal analyzer (EDSA) is used for the estimation of the system frequency characteristics. The estimated transfer function can be utilized to adapt a controller and improve the performance and stability of the system in terms of the model uncertainty. The EDSA comprises a sine generator, system identifier, and band-pass filter. A design methodology is proposed for the production of an accurate and cost-effective sine signal, which minimizes error and prevents overflow. Moreover, the proposed algorithms are implemented in a system-on-chip for an ODD. The experimental results show that the EDSA exhibits good performance and reliability. (C) 2012 The Japan Society of Applied Physics