A new decoupling precompensator for indirect adaptive control of multivariable linear systems

This note presents an indirect adaptive control algorithm for a class of multiple-input multiple-output (MIMO) linear systems. The controller consists of four parts. A precompensator and a feedback filter diagonalize the transfer function between a set of auxiliary control variables v
i
and a set of auxiliary output variables x
i
. Each SISO system with input v
i
and output x
i
is then easily controlled by standard SISO pole placement techniques. A model of the system is identified on-line, and is used to update the parameters of the precompensator, the feedback filter, and the SISO controllers. The main advantage of such an algorithm lies in the decoupling of the closed-loop system. This quality is fundamental for tracking performances. The a priori knowledge required for implementation of the algorithm consists of the system observability indexes.