Fuzzy-PI Controller Tuned With ICA: applied to 2 DOF robot control trajectory

Abstract

The main aim of this study is to introduce Fuzzy-PI based imperialist competitive algorithm (ICA) for the control of trajectory of 2 DOF robot planar. The dynamics that describes the robot is complex, non-linear and coupled model, thus it is a good benchmark model to test the efficiently and the robustness of the prosed schema. Fuzzy-PI is based on rather simple mathematical tools, which adequately investigate the ambiguities and reflect uncertainties. ICA is devoted to find the near optimal parameters of fuzzy logic (scaling factor) and PI parameters and its performance is compared to grey wolf optimizer (GWO), Whale optimizer algorithm (WOA) and teaching learning based optimization (TLBO). In order to test the robustness (the presence of external perturbation) and the efficiently of the proposed approach the PID controller is also tuned with ICA. The involved comparative study of our work shows that ICA was the outstanding one and the proposed controller outperform PID-ICA before and after the robustness test.