Projeto e controle de pêndulo invertido sobre rodas baseado no modelo fuzzy Takagi-Sugeno.

Abstract

Real-time implementation of control systems often involves uncertainties associated with the system’s complex dynamics and nonlinear nature, which makes its analytical analysis challenging for controller design. In this context, this work addresses the development of a two-wheeled inverted pendulum, used as an experimental platform for the application of control techniques. The system modeling is carried out using the Takagi-Sugeno methodology, which allows representing its nonlinear dynamics through the combination of local linear submodels. For system stabilization, a controller based on the Parallel Distributed Compensation (PDC) technique is employed. The development includes both the construction of the physical plant and the digital implementation of the controller on an embedded system, responsible for signal acquisition, control law processing, and actuator driving. The results indicate that the controller based on the PDC fuzzy technique was able to stabilize the two-wheeled inverted pendulum in the upright position, as well as perform reference tracking with satisfactory performance under practical conditions. The system responded adequately to small disturbances, although some limitations related to actuator capacity were observed in situations involving larger displacement demands. Overall, the work demonstrates the feasibility of applying controllers based on Takagi-Sugeno models to real systems, considering the challenges inherent to practical implementation.