Abstract
Electroosmotic effect is usually utilized to generate the flow field in microfluidic systems. In many of these microdevices, an accurate control over the output flow rate of the microfluidic part is necessary for the successful operation of the whole system. In this study, a combined feedback/feedforward strategy is proposed to control the output flow rate in a micro-T-junction. First, finite element model of the electroosmotic flow in the T-junction is generated; second, using the adaptive neural fuzzy inference system, the finite element model forms a basis for generating training data for building an inverse model of the flow in the micro-T-junction. This inverse model serves as a controller in the feedforward part of the system. Then, in order to make the controller robust against disturbances and uncertainties such as dimensional tolerances, a Mamdani-type fuzzy logic controller is incorporated in the feedback part of the controller. Finally, simulation results are presented in order to proof the performance of the designed controller.
Get full access to this article
View all access options for this article.