A modified fuzzy PI controller for a flexible-joint robot arm with uncertainties

Abstract Fuzzy control strategy offers an alternative approach for many conventional control systems, which has certain advantages over the other techniques. For example, the improved PI fuzzy controller we designed in this paper, which is based on an existing fuzzy PI controller, can control an uncertain flexible-joint robot arm to produce satisfactory tracking results. The improved fuzzy PI controller not only can control (stable and unstable) conventional linear systems, performing as well as the conventional PI controller, but also is capable of controlling many nonlinear systems such as the flexible-joint robot arm under investigation which contains uncertainties within 10% tolerance of all nominal system-parameter values. In this improved fuzzy PI controller, we used only three simple membership functions plus six simple fuzzy logic control rules. In this paper, we describe the design principle of this controller and demonstrate its tracking performance in handling nonlinearity, flexibility, and uncertainty within the flexible-joint robot arm system.