Underactuated Robotics

Calendar

SES #	TOPICS	KEY DATES
1	Fully- vs. under-actuated systems Preliminaries
2	Nonlinear dynamics of the simple pendulum	Problem set 1 out
3	Introduction to optimal control Double-integrator examples
4	Double integrator (cont.) Quadratic regulator (Hamilton-Jacobi-Bellman (HJB) sufficiency), min-time control (Pontryagin)
5	Dynamic programming and value interation: grid world, double integrator, and pendulum examples	Problem set 1 due Problem set 2 out
6	Acrobot and cart-pole: controllability, partial feedback linearization (PFL), and energy shaping
7	Acrobot and cart-pole (cont.)
8	Policy search: open-loop optimal control, direct methods, and indirect methods	Problem set 2 due
9	Policy search (cont.): trajectory stabilization, iterative linear quadratic regulator (iLQR), differential dynamic programming (DDP)	Problem set 3 out
10	Simple walking models: rimless wheel, compass gait, kneed compass gait
11	Feedback control for simple walking models
12	Simple running models: spring-loaded inverted pendulum (SLIP), Raibert hoppers
	Midterm
13	Motion planning: Dijkstra’s, A-star	Problem set 3 due
14	Randomized motion planning: rapidly-exploring randomized trees and probabilistic road maps	Problem set 4 out
15	Feedback motion planning: planning with funnels, linear quadratic regulator (LQR) trees
16	Function approximation and system identification	Final project proposal due
17	Model systems with uncertainty: state distribution dynamics and state estimation	Problem set 4 due
18	Stochastic optimal control	Problem set 5 out
19	Aircraft
20	Swimming and flapping flight	Problem set 5 due
21	Randomized policy gradient
22	Randomized policy gradient (cont.)
23	Model-free value methods: temporal difference learning and Q-learning
24	Actor-critic methods Final project presentations
25	Final project presentations	Final projects due