REDUCED-DIMENSIONAL FUZZY LOGIC CONTROLLER BASED ON LQR FUSION FUNCTION FOR LINEAR INVERTED PENDULUM SYSTEM

Abstract

There are mainly two concerns while one starts thinking about the designing of a controller, first one is response of the overall system and other one is control signal. Response of the overall closed loop system comprises of transient and steady state behavior from which the transient response can be seen through the transient parameters such as peak overshoots, undershoots, rise time and oscillations presents in the response of the plant and steady state behavior by steady state error. It is desirable that for better performance of the controller, the control action must be well under the operating range and should be smooth in nature. While simulating controller, first we get mathematical description of system. Mathematical description involves mathematical approximations, system parameter uncertainties, disturbances such as noise and plant nonlinearities. We can use conventional PID but it requires highly accurate mathematical description due to its linear nature. For a nonlinear system, response of conventional PID is unsatisfactory. So, fuzzy controller is appropriate for these nonlinear conditions as fuzzy controller - generates better responses for nonlinear conditions due to its ability to handle inherent nonlinearities and plant uncertainties. Inherent non-linarites and uncertainties of a plant can be easily handled by the fuzzy logic controller. Complexities may arise in FLC while designing due to increase in the numbers of rules as number of input increases. To overcome this problem LQR fusion technique has been used. The reduced of inputs hence the numbers of rules are reduced. The performance of LQR fuzzy controller is evaluated using MATLAB and SIMULINK on IP and MagLev system and the results are discussed in this thesis.