Adaptive Control for Partially Known Systems

Adaptive control has been considered as an alternative in designing high-performance control systems, from the beginning of the 1950s. Since then, most of the adaptive control schemes have been formulated either in the continuous-time or in the discrete-time framework. Both approaches commonly use ``black-box'' models for describing the process to be controlled; models with known structure but unknown parameters. These models have the advantage that they are general but also the disadvantage that many parameters have to be estimated. There are in practice, however, many adaptive problems where the system can be described as partially known in the sense that part of the system dynamics is known and another part unknown. This is the kind of system considered in this book. Most of the adaptive algorithms that are reliable - in the sense that they guarantee closed-loop stability and some performance behaviour - require to a certain extent some system knowledge and a checking procedure for the caution update of the parameter estimates.
 
Adaptive control has been considered as an alternative in designing high-performance control systems, from the beginning of the 1950s. Since then, most of the adaptive control schemes have been formulated either in the continuous-time or in the discrete-time framework. Both approaches commonly use ``black-box'' models for describing the process to be controlled; models with known structure but unknown parameters. These models have the advantage that they are general but also the disadvantage that many parameters have to be estimated. There are in practice, however, many adaptive problems where the system can be described as partially known in the sense that part of the system dynamics is known and another part unknown. This is the kind of system considered in this book. Most of the adaptive algorithms that are reliable - in the sense that they guarantee closed-loop stability and some performance behaviour - require to a certain extent some system knowledge and a checking procedure for the caution update of the parameter estimates.