Trajectory Optimization

MIT 6.832: Underactuated Robotics

Spring 2021, (mini-)Lecture 9

Follow live at https://slides.com/russtedrake/spring21-lec09/live

(or later at https://slides.com/russtedrake/spring21-lec09)

Lecture 9 Recap

• Big Idea: \( x(0) \) instead of \( \forall x \)
• General form: \( \min_{u(\cdot)} \int \ell(x(t), u(t)) dt, \) subject to ...
• Many "transcriptions"
  • Direct transcription, shooting, collocation, pseudo-spectral
• Many solvers
  • convex, sequential-quadratic programming (SQP), augmented Lagrangian, ...