Abstract—Human motion prediction through computational simulation can serve as a tool to anticipate the result of surgery or to help in the design of prosthetic/orthotic devices. The latter is the motivation in a project being run by the authors, devoted to the design of an active stance-control knee-ankle-foot orthosis (SCKAFO) as an assistive device for the gait of incomplete spinal cord injured (SCI) subjects. Optimization is a well-suited technique to tackle the human motion prediction problem, and several approaches have been proposed in the literature. However, no matter which is the used approach, the implementation of these methods represents a great challenge in terms of both convergence and efficiency. Therefore, the authors intend to firstly address the analysis of a certain measured motion through forward dynamics, which can be considered as an intermediate step towards the prediction problem, since it requires dynamical consistency too, but does not suffer from the same high amount of uncertainty. Consequently, a systematic study of the different alternatives to obtain, through forward dynamics, the drive efforts at joint level that produce a certain known motion is started in this paper. Three model-based control methods have been implemented for the gait of a healthy subject, and their performances have been compared.