Notebook Entry

Today's task list:

  • [x] Work on parsing the walking data
  • [] Make generic settings on the lab website
  • [] Work on the website theme
  • [] Setup a data backup for the website
  • [] Review the TODO items on the Yeadon paper
  • [] Run variations in guesses for structural id
  • [] Work on BMD papers
  • [] Book flight for BMD
  • [] Book hotel for BMD
  • [] Post update about BMD copyright

Walking System ID

I finally had a good solid day of work on the walking system identification problem and got some results.

I have a new SimpleControlSolver class that handles finding the gains for a walking cycle in my DynamicistToolKit, check out the walk branch:

https://github.com/moorepants/DynamicistToolKit/blob/walk/dtk/walk.py#L12

It is almost fully tested but there is some polishing needed. It can handle any arbitrary number of sensors and controls. We could theoretically find the gains to control every joint torque in the body given all of the things a person can sense.

The example notebook but that runs the solver on some real data is here:

http://nbviewer.ipython.org/6289985

The final plot gives the gains as a function of time in the step for two sensors and three controls in the right leg. Now I need to see if this controller can actually drive Ton's human body model. I'd like to wrap his C code with Python for ease of use.

comments powered by Disqus