Robotics

CSE P 590

Spring 2014: Monday 6:30-9:20, EEB 037

Instructor: Emo Todorov, todorov@cs.washington.edu
TA: Vikash Kumar, vikash@cs.washington.edu
TA: Igor Mordatch, igor.mordatch@gmail.com


Course Description

This course will provide a technical overview of modern robotics. Topics will include mechanism design, estimation, control, learning. The theory will be illustrated with example applications to flying, walking, object manipulation with both physical and simulated robots.


Catalyst Forum

Online course evaluation, closes June 12

Homework (see Forum for files)

Homework 1, due May 15

Homework 2, due May 29

Homework 3, due June 13 (no extensions)

Lecture slides

Introduction

Synthesis of complex movements with optimal control (research talk)

Linear algebra

Numerical optimization

Kinematics

Jacobian methods

Human-Robot Interaction (research talk)

Markov decision processes and Bellman equations

Controlled diffusions and Hamilton-Jacobi-Bellman equations

Linear-Quadratic-Gaussian (LQG) control

Approximate optimization

Trajectory optimization

Bayes filter

Kalman filter

Extended/Unscented Kalman filter

Particle filter

Multi-joint dynamics

Recommended Readings

Extended Kalman Filter Tutorial

World Robotics 2013: Executive summary

Optimal Control Theory (book chapter)

Rotating objects using quaternions (online tutorial)

R. Stengel, Optimal Control and Estimation (book)

L. Siavicco and B. Siciliano, Modelling and Control of Robot Manipulators (book)

Code

plotmat.m : visualization of 2-by-2 matrices