Extreme-ANFIS: A novel learning approach for inverse model control of Nonlinear Dynamical Systems

摘要

The paper proposes a novel, simple and faster learning approach named `Extreme-ANFIS' to tune premise and consequent parameters of Takagi-Sugeno Fuzzy Inference System (TS-FIS). Further the Extreme-ANFIS is used to design inverse model of nonlinear dynamical system. In this paper, the product concentration of non-isothermal Continuous Stirred Tank Reactor (CSTR) is controlled effectively by controlling inlet reactant temperature by using the Extreme-ANFIS based inverse model control technique. The effectiveness of proposed controller is verified by simulating it in MATLAB and comparing with conventional PID control.