Parametric identification by using Model Reference Adaptive System (MRAS)
The research of Model Reference Adaptive System (MRAS) is highly interesting. It is important in investigating a system suffering both from perturbations and noises in the environment under which it operates and from the aging problem. These problems are not easy to deal especially when the system is poorly known. The uncertainties of both system parameters and system behavior can be soothed by using MRAS’s which can be used as adaptive model-following systems, parameter identifiers, or state observers. The research in this work finishes a theoretical approach which leads to a design method for parameter identifiers from which the parameters of a working system can be extracted on-line. The theoretical basis for this approach is the Popov’s theorem from which the hyperstability and positivity concepts are induced and quoted. It follows that design methods for both state/output systems and input/output systems can be developed.Then based on these results, a parameter identifier for speed control of DC motors, which are second- order systems, is introduced as an example when only input and output of the systems are available. The system performance has been evaluated by software simulation. Possible perturbation on parameters is considered during the system simulation. So far some similar approaches are roughly given as can be referred from the books listed at the end of this work, but these approaches are left incomplete. It is hoped that this work finishes the approaches as completely as possible.