| A new Adaptive Pole-Placement (APP) approach for plants with overparameterization and/or incomplete excitation is presented in this dissertation. Unlike most of the indirect approaches the one developed here requires that only an upper bound on the plant order be known a priori and does not require any specific excitation order. The ill-conditioning of the on-line controller design problem, which has been a stumbling block in the development of adaptive control for nonminimum phase plants, is avoided by removing the common factor between the numerator and the denominator of the estimated plant model. The global stability of this APP algorithm is proved. A numerically stable algorithm for finding the common factor of two polynomials is also developed. Some efficient sequential and parallel algorithms for implementing this APP approach are presented. It is shown that only O(N... |
