A New Robust Model Predictive Control Strategy For Balancing Of An Inverted Pendulum

Abstract

In this thesis, we presents a methodology to design a controller for the balancing of an inverted pendulum. The inverted pendulum is one of the most difficult control problem in which the pendulum needs to be balanced against the cart, which moves only in two direction to the left or to the right. Different controller’s are designed to balance the pendulum and produces results which are more effective and fast. The objective is to find out the best control strategy that gives performance which is efficient and better regarding the angle of the pendulum and cart’s position. The mathematical model of an inverted pendulum is implemented in MATLAB/SIMULINK. Linear Quadratic Regulator (LQR), Proportional Integral Derivative (PID), H-infinity and Robust Model Predictive Control (RMPC) are designed to balance the unstable inverted pendulum system. A comparative study and analysis of these controllers implemented in MATLAB is presented in this paper to explore and investigate the controller, which is more robust and efficient in balancing the inverted pendulum system. A Linear Quadratic Regulator (LQR), Model Predictive Control (MPC), H-infinity and PID control strategies are applied for controlling the system equations of the inverted pendulum model and are analyzed and compared. The results of controllers implemented in MATLAB shows that all the controllers implemented are able to stabilize the inverted pendulum system. The result reveals that Robust Model Predictive Control (RMPC) strategy gives quick response as compared to other conventional control strategies. The overall performance is also more desirable than LQR, H-infinity and PID controllers.