Performances of Hill-type and neural network muscle models - Toward a myosignal-based exoskeleton

Muscle models are the essential components of any musculoskeletal simulatio n. In addition, muscle models which are incorporated in neural-based prosth etic and orthotic devices may significantly improve their performance. The aim of the study was to compare the performances of two types of muscle mod els in terms of predicting the moments developed at the human elbow joint c omplex based on joint kinematics and neuromuscular activity. The performanc e evaluation of the muscle models was required to implement them in a power ed myosiganal-driven exoskeleton (orthotic device). The experimental setup included a passive exoskeleton capable of measuring the joint kinematics an d dynamics in addition to the muscle myosignal activity (EMG). Two types of models were developed and analyzed: (i) a Hill-based model and (ii) a neur al network. The task, which was selected for evaluating the muscle models p erformance, was the flexion-extension movement of the forearm with a hand-h eld weight. For this task the muscle model inputs were the normalized neura l activation levels of the four main flexor-extensor muscles of the elbow j oint, and the elbow joint angle and angular velocity Using this inputs, the muscle model predicted the moment applied on the elbow joint during the mo vement. Results indicated a good performance of the Hill model, although th e neural network predictions appeared to be superior. Relative advantages a nd shortcomings of both approaches were presented and discussed. (C) 1999 A cademic Press.