Automated estimator parameter selection for an IBM head/disk assembly

ISA Trans. 2005 Jul;44(3):413-22. doi: 10.1016/s0019-0578(07)60213-4.

Abstract

This paper presents the application of a discrete adaptive observer (DAO) to an IBM head/disk assembly system. Because of the difficulties in tuning, a genetic algorithm is implemented off-line to obtain optimal observer parameters for the DAO. Simulations show that the genetic algorithm is successful in choosing appropriate observer gains. Furthermore, as a result of these optimal gains, the observer state and parameter estimates converge accurately and quickly.