Comparison of State Feedback Controller Design Methods for MIMO Systems
Prachi Barsaiyan and Shubhi Purwar
are considered. Pole Placement design allows all closed loop poles to be placed in desirable locations but does not provide robust controller. LQR is an optimal control design method that results in a system with guaranteed robustness. The state and control input weighting matrices, for the LQR cost function, must be selected by the designer in an attempt to achieve the desired system performance. Third method considered, is a Eigenstructure Assignment method of designing a state feedback controller to regulate as desired the output response of minimum phase linear multivariable square plants with CB having full rank (B, input matrix and C, output matrix). This method provides the advantage of allowing great flexibility in shaping closed-loop system responses by allowing specification of closed-loop eigenvalues and corresponding eigenvectors, but has the disadvantage that stability-robustness is not guaranteed. On the other hand, LQR assures stability-robustness with full state feedback but does not provide the flexibility of assigning eigenvalues and eigenvectors in placing closed-loop structure. II. CONTROLLER DESIGN METHODS In this section, four state feedback controller design methods Pole Placement method, Linear Quadratic Regulator (LQR), an Eigenvalue Assignment method for shaping the output response and a method based on Moores Eigenstructure Assignment, are discussed. A. Pole Placement or Pole Assignment Method The Pole Placement design allows all closed loop poles to be placed in desirable locations. The necessary and sufficient condition for the arbitrary placement of closed loop eigenvalues in the complex plane (with the restriction that complex eigenvalues occur in complex conjugate pairs) is that the system