A Modified Method of Submovement Decomposition Based on Velocity Profile and Endpoint Position

Proper functioning of the rehabilitation and motion control systems depends on the proper recognition of the motor system and the mechanism of natural movement formation. The submovement theory states that each movement consists of building units that are called submovement and by combining a limited number of submovement, more complex movements are formed. Many evidence supports this theory. Therefore, proper submovement extraction is an important topic in movement identification and control. One of the most commonly used submovement extraction methods relies on fitting a number of overlapping basis function to the movement. Parameters of these basis functions are usually identified by optimization method. In the related papers, the adaptation between the velocity profiles of the reconstructed movement and the main movement is used as an optimization cost function criteria, while the adaptation of the movement end point to the target is very important in many movements. Accordingly, the purpose of this paper is to modify the objective function of optimization so that both the reconstructed movement paths and end points are aligned with the original movement. For this purpose, the optimization cost function was modified so that in addition to the difference in velocity profiles, the effect of the target distance at the end of the movement, is also considered. The results show that, there is good adaptation between the main and reconstructed movement trajectories and end points. Therefore, with the proposed method, more appropriate submovements are extracted.