Command trajectory of fast forearm movements estimated by optimal control theory

Using optimal control theory and dynamic programming, we reconstructed the command trajectory of the fast elbow joint movement with two musculoskeletal models of the elbow joint, one with linear areflexic and the other with nonlinear reflexic muscle representation. The goal of the optimization was to seek a command trajectory that best reproduced the experimentally derived movement trajectory. The estimation results showed that the command trajectory using the model with linear areflexic muscle representation manifested the triphasic response, while the results using the model with nonlinear reflexic muscle representation showed only biphasic response. The calculated command trajectories capture the characteristics of the measured EMG. The results revealed that the optimal control theory could be used to reconstruct, from the joint angle trajectory, a command trajectory compatible with the currently accepted concept of motor control.