Note: simple_simulator could be replaced by any other simulation environment
Reference trajectories for the quadruped were optimized using the Kinodynamic Opimization Library, the references are added to this repository but if you wish to generate your own feel free to modify the yaml files in python/risc/demos/solo12/planner/config
A good starting point might be at risk_sensitive_control/python/risc/demos/cliff
. In the cliff demo, the optimal control problem is that of a point mass crossing a cliff. This problem is adopted from Farshidian for its simplicity. More complicated demos involving locomotion can be found in risk_sensitive_control/python/risc/demos/solo
. We setup optimal control problems accroding to crocoddyl
convention while adding noise models. The main solver code is in risk_sensitive_control/python/risc/solvers/risc.py
.
- Bilal Hammoud (bah436@nyu.edu)