Nonlinear predictive control of unicycle robot (4): simulations

Antonio Sala, UPV

Difficulty: ** ,       Relevance: ,      Duration: 21:15

Summary:

This video simulates the trajectory optimization via nonlinear model predictive control (Casadi+Ipopt) whose detailed analysis of the code was carried out in videos [mpcunic2EN] and [mpcunic3EN].

In here, we carry out several optimizations and simulate their results:

• Minimum-time trajectory $L=1$

• Weighting of control actions: $v^2$, $v^2+w^2$, and also $v^2{w}^2$.

• Reducing the number of waypoints

• Obtaining different trajectories by adding $2\pi$ to the start or goal configurations.

• Close ‘parking-like’ maneuvers.

These simulations help understanding the strengths and weaknesses of the idea. As discussed in earlier videos, the trajectory optimization step is just one of the many ones needed in a realistic mobile robotics application (high-level planning, safe corridor computation, trajectory optimization, implementation decisions, reactive navigation/replanning options).

*Link to my [ whole collection] of videos in English. Link to larger [ Colección completa] in Spanish.