Terry, Pang
Strategy
Tackle known smooth / contact dynamics.
ICRA / RA-L (September)
L4DC (November)
Sell it in the case of Neural Network dynamics
L4DC (November)
Further directions (maybe next year)
Naming Advice
Went through several variants, but would be good to decide on a name.
Candidates (in increasing order):
Update 1: Addressing high dimensionality
Trajectory optimization for Drake's quadrotor (12 states, 4 inputs):
Uses NO gradients (completely zero-order), only 1000 samples required for each knot point.
Takes less than 10 seconds (200 timesteps, but could be much faster if we can sample in batch)
Why is it the case that only 1000 samples can work so well for a 16 dimensional system?
Short answer: we're simultaneously sampling in all directions.
Long answer: Bellman's curse of dimensionality is not the same as Bakhvalov's (MC integration). But what is it about the dynamics that kills the curse? Can we also use Quasi Monte Carlo / Sparse Grids?
Update 2: Trajopt through Quasi-dynamic Contacts
Initial Guess
Exact Gradients
(QP's KKT gradients)
Randomized Smoothing (Gradient Sampling)