Data-Driven Active Disturbance Rejection Control of Pendulum Cart Systems

This paper proposes the discrete-time data-driven Active Disturbance Rejection Control (ADRC) algorithm for nonlinear Single Input-Multi Output pendulum cart systems (PCSs). The PCS is controlled using a Single Input-Single Output control system structure in cart position control loop. The performance of the control structure with data-driven ADRC controllers is compared with that of a control structure with a discrete-time data-driven Model-Free Control-intelligent Proportional Integral Derivative (MFC-iPID) controller and with linear controller having the same number of tunable parameters as ADRC and MFC-iPID through real-time experimental results. In order to have a fair comparison, the controller parameters are tuned in model-based term using a metaheuristic Gravitational Search Algorithm (GSA) optimizer.