Evolving single input FLC for double inverted pendulum

Abstract

In case of multivariable and nonlinear system, double inverted pendulum is absolute unstable which demands designing of an efficient controller for this kind of system. In this research, an efficient controller for double inverted pendulum (DIP) is demonstrated using LQR theory, information integration technology and signed distance method of fuzzy logic controller (FLC). The optimal linear control theory is applied to design of LQR for DIP, and then information fusion technique is used for reduction of six output variables to two output variables: error and change of error. After getting feedback signal in two variables, a powerful method signed distance technique is proposed to cut the given inputs of the FLC to a single input FLC. This single input fuzzy logic controller produce a 2-D linear control surface using only seven rules. MATLAB (7.8.0) software is used for simulation of the controller in Simulink and the simulation results show feasibility of the Single Input Fuzzy Logic Controller (SIFLC). The performance comparison shows that the proposed controller can give better transient and steady state performance over the conventional fuzzy logic controller.