摘要：AbstractModel-free adaptive control has been used to design feedback in cases when accurate mathematical modeling of the system can not be realized. By using only input-output information, the control approach is attractive with respect to minimal design efforts. In this contribution an improved model-free control program is proposed and applied to flexible systems. The main control idea is based on the compact-form dynamic linearization technique. A linearized controller structure which contains a matrix of unknown time-varying parameters namely pseudo-partial derivative can be designed. First, an equivalent linearized data-based model of the original system is established and corrected locally by using the recursive least-squares estimation algorithm. Then a modified objective function with respect to the controller parameter matrix considering minimization of the current tracking errors and its variations is proposed for control performance improvement. Finally, the required control inputs are calculated to fulfill control requirements. For illustration and as example, the newly introduced method is applied to reduce vibrations of an elastic crane representing a multivariable system. The control effectiveness is verified nummerically and compared with conventional model-free and PI controllers.
关键词：KeywordsModel-free adaptive controlDynamic linearizationRecursive least-squaresEstimation parametersVibration control
