Dynamic Modeling

Planning:

ADAMS (leg effects)
Similar testing to that done on the robot for comparison

ADAMS (slope/friction limitations)
Increase slope and determine point of 63% (or some other value => 50%, 70%, 90%) velocity decrease (-3dB?) relative to flat, normal running. State why the cutoff point was selected. Change the friction properties and repeat. Use these results, and try to verify with robot on different slopes and ground types.

ADAMS (other-active servoing, periodic force)
Periodic force can be added using the disturbance code that was used for constant radius circular turns ("make_cockroach4_CMP_6dc_PERT.cmd").
Active servoing may have to include servo dynamics (may be tedious), or a leg angle path that is preset to be periodic (tedious, but probably not as bad as other option).

Older Dynamic Modeling page notes

-- MarkCutkosky - 01 Aug 2003 Moved from AMcClung 's project area to the main RobotSimulation page on the grounds that this is relevant to general simulation & modeling tasks.

Over the next month or so, I plan to investigate the dynamic modeling of a CBot on various surfaces. This will start with a brief exploration in the limitations of our current design with additional friction variables, reflecting different feet/surfaces. Investigations may also include new designs with varied actuation techniques (eg. linkages, active leg angle stance changes).

-- AMcClung? - 18 Jun 2003

I would like to see us converge on a primitive "Theory of turning" -- essentially to be able to say, "If you want to turn without slowing down too much then you should do this, and here is why. Conversely, if you want to turn on a dime do this. And this is why this is the right thing to do. And indeed, see how the experimental results and simulations bear us out. This relates to the discussion started with NoahCowan on SprawlTurning-- MarkCutkosky - 24 Jun 2003

Dynamic modeling goals

==================

• Investigate limitations of current design, as motivation for new climbing strategies
  • Slope vs Performance(velocity, COM periodicity?, turn rate(nonlinear)?)
  • Stride period & Slope vs Performance
  • May be able to add force (constant or phased) to model, similar to having additional normal force (ie, fan/wings)
• Investigate how friction effects limitations
  • repeat Slope vs Perfomance for several friction values (relating to different foot interactions - ie. surfaces&feet)
• Time permitting:
  • Expansion of model for active servoing
  • Linkage based model (4bar, cam, etc.)

• Need definite test plan
• Good for those interested in modelling, may be difficult to get up to speed w/ADAMS