Model Predictive Control of Wheeled Mobile Robots

Different control techniques can be used to perform trajectory tracking of nonholonomic Wheeled Mobile Robots (WMRs). The application of Model Predictive Control (MPC) for trajectory tracking of WMRs allows for an easy incorporation of input and output constraints that may arise in practical situations. Also, the control effort required to solve the trajectory tracking problem can be minimized in a straightforward manner using the built-in optimization methods in MPC. Computation time is the biggest hurdle in adapting MPC strategies for trajectory tracking. This research applies a non-feasible active set MPC algorithm that is much faster than the conventional active set methods. Traditionally, non-feasible active set method is used in plant control where the system is time-invariant. The novelty of this work lies in the application of non-feasible active set method to time-varying models of Wheeled Mobile Robots. A comparative study is conducted on feasible and non-feasible active set methods using various reference trajectories.