State-space adaptive control through a modified bootstrap algorithm for parameter and state estimation

摘要

A modified bootstrap algorithm for combined parameter and state estimation is used for self-tuning control of MIMO systems. The suboptimal control law through state feedback and based on an infinite-time quadratic cost function is applied at each step. The superiority of the modified bootstrap method is established by comparison with other methods such as recursive prediction error and another bootstrap scheme. The comparison is performed in terms of parameter estimation, state estimation and regulation.