Design of NMPC-Based Framework for Mobile Robot Motion in Unstructured Environments

The paper treats a design of the mobile robot motion framework based on Nonlinear Model Predictive Control (NMPC). This approach relies on laser range-finder measurements and safety regions described around the detected obstacles. The controller optimization involves both actuator and environment constraints excluding safety regions. The safe motion of the mobile robot using the proposed framework is provided in unknown static and dynamic environments. Stability issue of the used closed-loop motion control system is guaranteed employing direct Lyapunov method. The simulation setup was conducted using a 2D simulator (Stage) in the Robot Operating System (ROS) environment. The obtained results through different scenarios demonstrates that proposed NMPC-based framework approach ensures smooth trajectories from start to the goal point.