Fuzzy Adaptive Controller Design Methodology for Robotics Systems

In this paper, an adaptive fuzzy control approach for complex task involving robot/environment interaction is presented. The approach implementation is based on fuzzy logic controller (FLC) design and optimization methodology which operates in two stages. In the first stage, the FLC parameters are trained and optimized offline using a method based on Solis´ and Wetts´ algorithm so that the constraints in terms of interpretability and semantics are respected. For convenience of analysis, the structure of the FLC is divided into multi-input-single-output (MISO) controllers. In the second stage, an online learning of the optimized FLC is then implemented into the proposed control structure, so that the closed-loop system maintains good performances. To show the performances of the proposed approach, simulations are carried out on a three-link robot in three dimensional space, performing contour following under force constraints