Abstract
This research formulates a mathematical model that simulates human arm's reaching movements for capturing a uniformly and linearly moving object, and examines the model's effectiveness. Actually, the research measures and analyzes moving-object capturing movements, and then models the human arm's moving-object capturing mechanism as a feedforward control model composed of three links (upper arm, forearm, hand) and three joints (shoulder, elbow, and hand joints) based on a moving-object capturing strategy: the strategy is to optimally predict the capturing point where the human arm's reaching movement duration becomes equal to the object's uniform and linear movement duration. Consequently, the following results are obtained: (1) the measured capturing movements imply the existence of the human learning mechanism of object's motion characteristics and the human arm's feedforward control mechanism; (2) the proposed model can capture the uniformly and linearly moving object; (3) the proposed model can optimize the reaching movement duration to the moving object; (4) the proposed model can basically reproduce the measured capturing trajectories accurately, although it predicts trajectories different from the measured ones as for some measured capturing movements. These results suggest that the proposed feedforward control model can be a more generalized model of the human arm control mechanism as well as an effective model of the human arm's moving-object capturing mechanism.
