API  4.2
For MATLAB, Python, Java, and C++ users
MocoInverse: solving muscle and actuator redundancy

A common goal in musculoskeletal biomechanics is to estimate muscle and actuator behavior that drove an observed motion; this is often called the muscle redundancy problem.

We can solve this problem by tracking the observed motion, as with Computed Muscle Control (using the "slow target") or MocoTrack. Alternatively, we can prescribe the motion exactly, as with Static Optimization, EMG-driven simulation [1, 2], and the Muscle Redundancy Solver [3]. The advantage of prescribing the motion is the problem is more robust and solves faster, as the nonlinear multibody dynamics are no longer part of the optimization problem. The disadvantage is that we cannot predict deviations from the observed motion.

In Moco, you can use the MocoInverse tool to solve optimization problems in which kinematics are known, including the muscle redundancy problem.

The MocoInverse tool adds a PositionMotion component to your model, with splines created from a kinematics data file that you provide.

Note
The kinematics you provide should already obey any kinematic constraints in your model. MocoInverse attempts to alter your provided kinematics to satisfy the kinematic constraints (using Model::assemble()), so the tool will not actually track the kinematics as provided if they violate constraints.

Sometimes it is not possible to achieve the desired motion using muscles alone. There are multiple possible causes for this:

  • the muscles are not strong enough to achieve the required net joint moments,
  • the net joint moments change more rapidly than activation and deactivation time constants allow,
  • the filtering of the data causes unrealistic desired net joint moments. For these reasons, you may want to add "reserve" actuators to your model. See ModelFactory::createReserveActuators() and ModOpAddReserves.

See Prescribed kinematics, Prescribed kinematics with kinematic constraints, and Prescribed kinematics for more information.

[1] Arnold, E., Hamner, S., Seth, A., Millard, M., Delp, S. (2013). How muscle fiber lengths and velocities affect muscle force generation as humans walk and run at different speeds Journal of Experimental Biology 216(11), 2150-2160. https://dx.doi.org/10.1242/jeb.075697
[2] Jackson, R., Dembia, C., Delp, S., Collins, S. (2017). Muscle-tendon mechanics explain unexpected effects of exoskeleton assistance on metabolic rate during walking Journal of Experimental Biology 220(11), jeb.150011. https://dx.doi.org/10.1242/jeb.150011
[3] Groote, F., Kinney, A., Rao, A., Fregly, B. (2016). Evaluation of Direct Collocation Optimal Control Problem Formulations for Solving the Muscle Redundancy Problem Annals of Biomedical Engineering 44(10), 2922-2936. https://dx.doi.org/10.1007/s10439-016-1591-9