A new implementation for simplified adaptive control

The command generator tracker (CGT) approach to model reference adaptive control of linear time-invariant multi-input, multi-output plants is considered. An algorithm is presented for controllable and observable non-ASPR (almost strictly positive real) plants. This algorithm is characterized by the presence of supplementary dynamics in a feedback path inside the adaptive mechanism. These additional dynamics process the output error to form a signal which is multiplied by a new adaptive gain. Therefore, the algorithm differs from other CGT adaptive control algorithms for non-ASPR plants, which require supplementary dynamics to be inserted in parallel with the plant. A generalized metastate representation of the closed-loop adaptive system is proposed. This representation is utilized together with Lyapunov´s second method to prove the boundedness of the true error between the plant and model outputs in the presence of bounded plant input and output disturbances