It is unclear what data you may or may not have, so here are a few thoughts:
1. It would be best to use measured external forces to the model with a tracking algorithm, such as Static Optimization or Computed Muscle Control. The external forces would include pedal forces, seat forces, and handlebar forces. You could then follow some of the examples (such as gait1018 or gait2392 included with the installation).
2. If you don't have measured external forces, you might have to use some constraints to generate forces at the pedals and seat, and then validate that the forces generated are good enough for your work. You could see this preliminary work as a starting example (this model is too simple to use for research):
https://simtk-confluence.stanford.edu/d ... ling+Model
It sounds like you may have an idea of what forces you want to apply at the pedals, so you might be able to create ExternalLoads files (as seen in the gait1018 and gait2392 examples), and then add a constraint at the seat to apply forces there.
