The effect of kinematic constraints in the inverse dynamics problem in biomechanics

The inverse dynamics is a classic problem to be solved in gait analysis. The solution is the vector of joint reactions and driving actions which is essential to perform a kinetic analysis of the human gait. Multibody system techniques are widely used to solve this problem. However, it is not clear in the literature if mathematical constraints of the joints should be fulfilled nor what is the effect of these constraints on the kinetic results. The effect of using or not kinematic data consistent with the joint constraints on the kinetic results is not clear either. Different methods to solve the inverse kinematics and inverse kinetics problems were implemented in this work to satisfy the joint constraints, the procedure to fulfil them or the type of joint constraints imposed. The goodness of these methods has been evaluated by means of dynamic residuals, which are required due to all the errors arising from gait measurement that may unbalance the dynamic equation. Results show that the imposition of kinematic constraints can lead to worse results in the kinetic results if the multibody model is too simplistic. However, this kind of approach can be necessary if further studies are performed, as a force sharing problem, where the reaction and driving actions should be properly defined.