Comprehensive Modeling and Simulation of an Anthropomorphic Robotic Exoskeleton for Rehabilitation

摘要

This paper presents a novel approach for dynamic modeling of the lower limb rehabilitation exoskeletons in view of requirements of repetitive movements with accuracy, and precise tasks in rehabilitation processes. It is necessary to achieve optimal control performance to realize well rehabilitation results in this kind of exoskeletons. There have been numerous exoskeletons designed in the past which have been modelled using Euler- Lagrange method. This method requires complex computations with large number of variables and is less flexible in addition and subtraction of variables. Therefore, the methodology of bond graph modeling is employed to build a dynamic model with very less computations and much flexibility. Modeling and simulation of an active three d.o.f single leg exoskeleton has been carried out and the dynamic responses have been analyzed to verify the effectiveness and performance advantages of the proposed method. Future research work will focus on designing an intelligent and optimized control strategy for lower limb rehabilitation exoskeletons to achieve desired performance.