MIT 6.832 Underactuated Robotics, Spring 2009

About this film

Instructor: Russell Tedrake Robots today move far too conservatively, using control systems that attempt to maintain full control authority at all times. Humans and animals move much more aggressively by routinely executing motions which involve a loss of instantaneous control authority. Controlling nonlinear systems without complete control authority requires methods that can reason about and exploit the natural dynamics of our machines. This course discusses nonlinear dynamics and control of underactuated mechanical systems, with an emphasis on machine learning methods. Topics include nonlinear dynamics of passive robots (walkers, swimmers, flyers), motion planning, partial feedback linearization, energy-shaping control, analytical optimal control, reinforcement learning/approximate optimal control, and the influence of mechanical design on control. Discussions include examples from biology and applications to legged locomotion, compliant manipulation, underwater robots, and flying machines. See the complete course at: http://ocw.mit.edu/6-832s09 Album art photographs are courtesy of Jason Dorfman.

Directors & creators

MIT OpenCourseWare

Subjects

underactuated robotics;actuated systems;nonlinear dynamics;simple pendulum;optimal control;double integrator;quadratic regulator;Hamilton-Jacobi-Bellman sufficiency;minimum time control;acrobot;cart-pole;partial feedback linearization;energy shaping;policy search;open-loop optimal control;trajectory stabilization;iterative linear quadratic regulator;differential dynamic programming;walking models;rimless wheel;compass gait;kneed compass gait;feedback control;running models;spring-loaded inverted pendulum;Raibert hoppers;motion planning;randomized motion planning;rapidly-exploring randomized trees;probabilistic road maps;feedback motion planning;planning with funnels;linear quadratic regulator;function approximation;state distribution dynamics;state estimation;stochastic optimal control;aircraft;swimming;flapping flight;randomized policy gradient;model-free value methods;temporarl difference learning;Q-learning;actor-critic methods

View on archive.org ↗