Premium
An analysis of simplified muscle activation profile parameterization
Author(s) -
Strobach Daniel,
Kecskeméthy Andrés
Publication year - 2006
Publication title -
pamm
Language(s) - English
Resource type - Journals
ISSN - 1617-7061
DOI - 10.1002/pamm.200610052
Subject(s) - biceps , computer science , ankle , pelvis , gait , rotation (mathematics) , reduction (mathematics) , mathematics , simulation , anatomy , physical medicine and rehabilitation , artificial intelligence , geometry , medicine
This paper analyzes a simplified method for rough identification of muscle activation profiles of general motor tasks by means of dynamic optimization. Muscle activation profiles are parameterized with six parameters per muscle, using linear combinations of two smooth C ∞ functions closely related to the GAUSSian distribution function used in stochastics and fuzzy control. The method is applied to a simplified subsystem of the human leg consisting of pelvis, thigh shank and foot, interconnected by planar joints at hip, knee and ankle. The system comprises one antagonistic muscle pair at the knee for knee flexion and extension ( vastus intermedius and biceps femoris caput brevis ). To simulate the swing phase of gait, rheonomic constraints are imposed on pelvis (translation and rotation), hip (rotation) and ankle (rotation). The optimization results show that, (1) the method is suitable to map typical muscle activation time histories that are recorded via EMG, (2) the method can reduce the number of design parameters and CPU‐time consumption significantly in comparison to other parameterizations and (3) this reduction in CPU‐time consumption additionally coinncides with an improved approximation quality to the target motion. (© 2006 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)