Compensator Design for Dynamic Optimisation of Multivariable Systems with Prescribed Pole Locations

Abstract

The paper presents a method for the design of optimal compensators with prescribed pole locations for linear multivariable systems. From the state equations of the multivariable system, an augmented set of state equations are obtained following the procedure of Pearson and Ding [6]. For the augmented state equations, a quadratic index is assigned for the calculation of optimal control gains. The state and control weighting matrices of the quadratic index are selected from the specification of the closed-loop poles of the resulting system using frequency domain optimality relation. From this quadratic index, a state feedback matrix is calculated for the augmented system by solving the relevant Riccati equation. Once this matrix is known, the compensator structure is decided based on method by Pearson and Ding [6].

Indexing terms:

• Compensator design
• Optimal control
• Pole placement