Título: Neuro-Adaptive Robust Control Configurations Based on Variable Structure Control
Autores: Alves, Marco Antonio; Nóbrega, Eurípedes
Resumo: Among the several control methods developed during the last two decades, robustness has been the most emphasized characteristic in general. Recently, the compromise between robustness and performance has motivated new studies, considering the natural opposition between then. Structure variable control is a distinguished method because of its remarkable simplicity. Despite it has been developed long ago, it has attracted the attention recently due to its inherently robustness, and also because it is equally applied to linear and nonlinear systems. The basic idea is to restrict the state space of a given plant through a so called sliding surface, whose dynamics is simpler than the original plant dynamics. Enforcing a state space trajectory from the initial conditions to reach the surface, once there the plant dynamics is substituted by the surface dynamics. For adequately designed surfaces, they present the invariance property, guaranteeing an intrinsic robustness, because the new dynamics does not depend on the plant parameters. Associating this method to artificial neural networks, some of the proposed configurations may present simultaneously performance and robustness. In this work, a new configuration is proposed, implementing a neuro-adaptive control method using the variable structure approach to adjust the neural network weights, and consequently presenting also robustness. The central idea is to add a second control signal to a regular controller, generated through a neural network, whose online learning is also robust. It is expected that the controller performance will be maintained, independently of perturbations caused by structural or parametric variations. The configuration is explored through three different cases. These cases and also the robust learning technique are presented. Numerical simulations presenting good results justify the expectations for the configuration.
Código DOI: 10.21528/CBRN2003-036
Artigo em PDF: 6CBRN_036.PDF
Arquivo BibTex: 6CBRN_036.bib