Position Control of a Mobile Inverted Pendulum System Using Model-Free Intelligent Controllers

摘要

Model-Free Control (MFC) is a Data Driven Control (DDC) method which is designed for complex or nonlinear systems. Such a system is the Inverted Pendulum System which is known as unstable in open loop and has a high degree of nonlinearity and limitations. This complexity of the system makes it difficult to be controlled using the classical control methods. Advanced methods as Model Predictive Control (MPC), Linear Quadratic Regulator (LQR) or Sliding Mode Control (SMC) have been proposed over time to control this type of plant. Since it has a single input and two outputs which have a strong correlation, the inverted pendulum system represents a challenge for every control algorithm. There are applications for this type of plant to control the position of the cart or/and the angle of the pendulum. In this paper, a mobile inverted pendulum is considered which is a particular case of the classical inverted pendulum. For this plant, two MFC algorithms are proposed to control the position of the cart and the results are used in a comparative analysis. The MFC method uses intelligent PID (iPID) controllers which represent an improvement of the classical PID algorithm, based on the ultralocal model concept. The first algorithm is represented by an iPD controller while the second one is a conjunction between MFC and SMC which is materialized in a MF-SMC controller that compensates the estimation errors from the iPD design using an augmented control signal. The simulation results present the efficiency of MFC strategies of a positioning application for a mobile inverted pendulum.