Discrete time model predictive control and simulation of electric powered wheelchair

Electric powered wheelchairs (EPW) are among the assistive technology products that contribute to the independent mobility of individuals with disabilities. ETSs have to provide the desired mobility even in different ground structures and slope conditions. For this, it is vital to develop appropriate control methods. In this study, an ETS guided by a joystick and driven by direct current (DC) motors on its rear wheels has been considered. This ETS sets an example a multiple input-multiple output (MIMO) system with coupling effects. In the mathematical modelling of the system, Lagrange Formulation with dissipative energy has been used. Both state space model representation and model representation based on transfer function approach have been obtained in continuous and discrete time. For this purpose, by using MIMO ETS model based on state space representation, an optimal multivariable PID control design has been made with Model Predictive Control (MPC) method. In addition, independent each other PID controllers have been designed with parametric equations method for the ETS model, which consists of two independent single input-single output (SISO) subsystems based on transfer function representation and accepts coupling effects as disturbances. The performance of these control methods has been compared in a simulation environment. In addition, a two-dimensional animation of the movement has been recorded by following a generated position trajectory. Keywords: Electric Powered Wheelchair (EPW), multi input-multi output (MIMO) system, Modelling, Speed Control, Model Predictive Control (MPC), PID, Controller Matching, Trajectory Tracking, Simulation