In the longer term we may want to support lifting the acceleration values, that is, doing the inversion "numerically",
This would mean adding support for additional algebraic states a calculated via the constraint:
M @ model.a - model.f_v - J_n @ self.lambda_normal = 0
and making the dynamics
self.f_x = ca.vertcat(model.v, model.a)
This is just a comment, for the future, not this PR. We may want to open an issue on this.
Originally posted by @apozharski in #75 (comment)
In the longer term we may want to support lifting the acceleration values, that is, doing the inversion "numerically",
This would mean adding support for additional algebraic states
acalculated via the constraint:and making the dynamics
This is just a comment, for the future, not this PR. We may want to open an issue on this.
Originally posted by @apozharski in #75 (comment)