Human Synergy Based Analysis to Control of Rehabilitation Robot for Adaptive Assistance

Abstract

Adaptive assistance is needed to support mobility of disabled person due to weakened muscles and loss of joint control. Assistive robot requires coordinated motion trajectory of its joint, similar to human joint trajectory. A robot trajectory planning mechanism, based on comparative analysis between robot trajectory and human joint trajectory data is important to find coordinated trajectory for assistive robot joint. To perform the comparative analysis, human motion trajectory is measured by using motion capture data, and robot trajectory is generated by extracting via points from human motion trajectory. Further, polynomial methods (cubic, quintic and linear segment with parabolic blends) are used to generate robot trajectory between via-points. A four-link Sit to Stand (STS) model is used for comparative analysis of human and robot trajectories. To generate the robot trajectory using via-point method, estimation of maximum number of via-point is well known problem. Analysis is aimed to find the robot trajectory, that is close to human motion trajectory, by finding minimum energy and minimum error trajectory and to find the maximum number of via-point for trajectory with minimum energy and maximum coordination. The paper proposed robot trajectory generation method that uses via-points of human trajectory and Principle Component Analysis (PCA) based synergy calculation method. The main research contributions are towards finding robot trajectory with minimum energy and minimum error and to estimate maximum number of via-point based on minimum error and maximum coordination.