9.4 Using Moco to solve optimization problems

Owned by Former user (Deleted)
Oct 27, 2023
1 min read

OpenSim Moco is a software toolkit built on top of OpenSim for solving musculoskeletal optimal control problems. Moco can solve "inverse" problems (given experimental motion data, estimate quantities that were not measured during an experiment) and predictive problems (predicting a walking motion, without experimental motion data), as well as problems in between. Moco solves these problems using direct collocation, which is a generic method for solving optimal control problems. Learn more from Moco's website and bioRxiv preprint.

This example will show you how to use Moco to predict a squat-to-stand motion in Matlab. More specifically, the goals of this lab are as follows:

  1. Predict a new motion.
  2. Track motion data using a torque-driven model.
  3. Estimate muscle activity from motion data.
  4. Estimate the effect of a passive assistive device on a motion and muscle activity.  

Get started at the link below.

https://simtk-confluence.stanford.edu:8443/display/OpenSim/Moco%3A+Predict+a+Squat-to-stand