Implementation of Successive Convexification: A Superlinearly Convergent Algorithm for Non-convex Optimal Control Problems by Yuanqi Mao, Michael Szmuk and Behcet Acikmese
Reqires matplotlib, numpy, scipy and cvxpy.
This framework provides an easy way to implement your own models.
Fixed- and free-final-time optimization is possible.
The following models are currently implemented:
- Rocket trajectory model with free-final-time based on Successive Convexification for 6-DoF Mars Rocket Powered Landing with Free-Final-Time by Michael Szmuk and Behçet Açıkmeşe.
Video example of generated trajectories
-
Differential drive robot path planning with free-final-time and non-convex obstacle constraints:
-
2D rocket landing problem Feed-forward input tested in a box2d physics simulation