Author:

Abstract:

This thesis concerns the optimal design of mechatronic systems, which re quires that, among the structural and control parameters, an optimal cho ice has to be made with respect to design specifications in an integrate d way. To accomplish that, this work provides tools to computer-aid the integrated design of mechatronic systems. Two main challenges have been identified when dealing with integrated design: modeling due to their mu ltidisciplinary nature and optimization due to their non-convex nature. To overcome the former issue, guidelines for modeling mechatronic system s considering the system flexibilities, the system motion and the contro l action have been proposed. These parametric models may easily contain several degrees-of-freedom and be unsuitable for control design purposes . To obtain a concise description of the system, model reduction procedu res have been employed. To tackle the later issue, two strategies have been investigated: the nested and the d irect design strategies. The nested design strategy combines nonlinear o ptimization methods and the model-based control design techniques. The d irect design strategy considers, simultaneously, control and structural parameters using non-linear optimization or genetic algorithms. Three case studies have been exploited: a vibro-acoustic system and two pick-and-place machines, containing serial and parallel kinematic chains . The pick-and-place machines present configuration-dependent dynamics, which inevitably affect the performance and the stability of the control system. Important issues have been identified when the direct strategy is employed for designing a mechatronic system with configuration-depend ent dynamics, suggesting that the nested design strategy suits better this task. It can be concluded that the optimal design of mechatronic sy stems can only be accomplish when control and structural parameter are c onsidered in an integrated way.