ANFIS modelling of a twin rotor system using particle swarm optimisation and RLS

Abstract

Artificial intelligence techniques, such as neural networks and fuzzy logic have shown promising results for modelling of nonlinear systems whilst traditional approaches are rather insufficient due to difficulty in modelling of highly nonlinear components in the system. A laboratory set-up that resembles the behaviour of a helicopter, namely twin rotor multi-input multi-output system (TRMS) is used as an experimental rig in this research. An adaptive neuro-fuzzy inference system (ANFIS) tuned by particle swarm optimization (PSO) algorithm is developed in search for non-parametric model for the TRMS. The antecedent parameters of the ANFIS are optimized by a PSO algorithm and the consequent parameters are updated using recursive least squares (RLS). The results show that the proposed technique has better convergence and better performance in modeling of a nonlinear process. The identified model is justified and validated in both time domain and frequency domain