• rraCmcPage

Performing RRA in OpenSim: Key Elements

1. Actuators: specify the actuators (torques, for CMC they are muscles)

• a. Forces and moments acting on the pelvis from ground are residual actuators
  b. Internal joints are actuated by torques
  c. Optimal forces are the maximum output of ideal actuators (torques, linear forces). Torque (force) applied is optimal_force x control_value
  

2. ControlConstraints: specify the bounds on the controls

• a. Joint torques (and muscles) have a maximum magnitude of 1.
  b. Residuals have bounds exceeding their anticipated force requirement. Weightings are implicit in this description. A high optimal_force means that large output force (torque) does not require a large control value (i.e. low cost). Conversely, residuals with low optimal force require high control values that incur higher costs.
  

3. Tasks: specify kp, kv, and for each kinematic coordinate being tracked.

• a. Selection of kp and kv are not arbitrary. They define the behavior of the error dynamics for each q as a second order linear system. We can write the kp and kv for the desired system behavior in terms of system poles. For a (stable) critically damped system (real negative poles) kp = lambda^2 and kv = -2*lambda.
  b. enables kinematics of joints (coordinates) for which we have high confidence (e.g. knee flexion, hip flexion) to be weighted more heavily compared to those of less confidence (e.g. hip internal rotation and ankle inversion).
  

4. RRA Setup File:

• a. Model (lock subtalar and mtp joints, why?)
  b. Actuators (NOTE: residual at pelvis should be applied at scaled location of COM)
  c. Tasks (how strongly/loosely to track individual joint kinematics)
  d. Adjust COM flags
  e. Desired Kinematics
  f. Control Constraints (bounds on the controls)
  g. External loads: Specify bodies that external loads are applied to.
  

5. Run RRA/CMC with slow walking IK and GRF data, and previously scaled model.

6. Trouble-shooting:

• a. Four important things to do in any case
  

  • i. Check the pelvis COM location in Actuator files
    ii. “Lock” the subtalar and mtp joints in *.osim file
    iii. Make sure “use_fast_optimization_target” is
    

    • • false (unchecked) for RRA

      • true (checked) for CMC
      

    iv. Make sure “cmc_time_window” is
    

    • • <cmc_time_window> = T
      • 0.001 s for RRA
      • ~0.01 s for CMC
      

  b. Are the residual actuators (FX, FY, FZ, MX, MY, MZ) saturating?
  

  • i. Try increasing their min/max ranges in the ControlConstraints file.
    ii. Check that the GRF are being applied.
    

  c. Is a dof tracking poorly?
  

  • i. Consider increasing its weight in the Task file
    

  d. Does an actuator seem too weak?
  

  • i. Consider increasing its optimal force in the Actuator file (torque actuators) or *.osim file (muscle actuators), but provide good justification for this. How does the optimal force influence the objective function?