Flower pollination algorithm optimization applied for Lyapunov-based fuzzy logic speed controller of a complex drive system

Abstract

With the current trends in manipulator design, plentiful examples of machines with flexible links and joints can be instantiated. Lighter construction that allows bending and torsion offers multitude of advantages, such as lower energy consumption and better operation safety when collision is possible in the working environment. However, precise control of applications with such mechanical constructions is very challenging. The state variables might be affected by torsional vibrations, and identification of controller parameters is more difficult, which makes the controller tuning complicated. Therefore, this work focuses on tackling the issues related to speed control of electric drives with sophisticated, and elastic couplings. The robustness against parameter uncertainty is provided through the use of a fuzzy logic system. The speed controller design process incorporates the selection of the rule base, designation of membership functions, as well as controller gain optimization using a nature-inspired technique – the flower pollination algorithm (FPA). Increased damping of torsional vibrations, as well as decreased sensitivity to inertia changes is expected compared to other conventional control solutions, such as PI control. In this study both numerical and experimental studies are conducted.