DifferentialDynamicProgramming.jl: A package for solving Differential Dynamic Programming and trajectory optimization problems.

Research output: Non-textual formSoftware

Abstract

Solves the same problem as iLQG, with an added constraint on the KL-divergence between the new trajectory distribution and the distribution induced by a previous controller. This feature can be used in an outer loop with repeated experiments between the iterations if the model used for optimization is uncertain.
Original languageEnglish
Place of Publicationhttps://github.com/baggepinnen/DifferentialDynamicProgramming.jl
Publishergithub
Publication statusPublished - 2016

Subject classification (UKÄ)

  • Control Engineering
  • Robotics

Fingerprint

Dive into the research topics of 'DifferentialDynamicProgramming.jl: A package for solving Differential Dynamic Programming and trajectory optimization problems.'. Together they form a unique fingerprint.

Cite this