ON THE DESIGN OF SELF-TUNED REGULATORS USING STATE-SPACE APPROACH

The paper introduces a new approach to the design of a self-tuning regulator by making use of the state space model of the concerned system coupled with the minimization of a linear combination of the state error variance, the average controller energy and an additional cross-weighting term between the system states and the system inputs. It is shown that the proposed formalism permits a relatively simple solution of the adaptive regulator design problem due to the choice of a single stage performance index and a direct estimation of the state estimator gain matrix through the use of the innovations model of the concerned system. The method requires the solution of a linear matrix algebraic equation having coefficients estimated using a sequential parameter estimation algorithm. Besides the computational simplifications, the new approach has the additional advantages of being equally applicable to minimum and non-minimum phase, single-input single-output and multi-input multi-output systems. It is also pointed out that the resulting regulator yields a stable closed loop system under simple conditions. The results are illustrated numerically using simulation studies of a simple system.