Inverse dynamic optimization including muscular dynamics, a new simulation method applied to goal directed movements.

This paper presents a new method for estimating muscular force and activation from experimental kinematic data. The method combines conventional inverse dynamics with optimization utilizing a dynamic muscle model. The method uses only very limited computational power, which makes it a useful tool especially for complex systems like the shoulder or the locomotor system. The net torques/forces are calculated by using conventional inverse dynamics. A solution of the load sharing problem is determined by minimization of the weighted sum of squared muscle forces. The load sharing problem is solved with a dynamic constraint reflecting physiological muscle properties. This constraint takes into account the nonlinear dynamics of the contractile element (CE) and the series elastic element (SE), active state dynamics and neural excitation dynamics. This physiological constraint is determined with an inverse muscle model. With this model, muscular states and neural inputs are also estimated. The method of inverse dynamics requires position, velocity and acceleration signals as input. A method to prepare such signals from noisy measured data is presented.