Minimizing Human-exoskeleton Interaction Force by Using Global Fast Sliding Mode Control

Minimizing Human-exoskeleton Interaction Force by Using Global Fast Sliding Mode Control

A critical issue in the model-based control of performance-augmenting exoskeleton systems is the unknownnonlinear dynamic properties of the systems or the uncertainties. An improper estimation of the systemdynamics can cause instabilities in the system and generate considerable human-exoskeleton interaction forcesduring human motions. Thus, the controller of such exoskeleton systems needs to add robustness to stabilize itagainst the uncertainties. In this paper, we propose a global fast sliding mode control algorithm integrated in ahybrid controller for each exoskeleton leg to minimize human-exoskeleton interaction forces. By doing so, theproposed algorithm does not require an exact estimation of the dynamic properties of the exoskeleton system, butstill minimizes the physical human-exoskeleton interaction (pHEI) forces. Finally, the performance of the proposedalgorithm is verified by experiments on our lower exoskeleton system, which is used for human power augmentationand called “PRMI” exoskeleton. Our experimental results show that the proposed control algorithm provides agood control quality for the PRMI exoskeleton. The PRMI exoskeleton can support a wearer carrying heavy loadwhile tracking the rapid movements of the wearer without obstructing them.

A critical issue in the model-based control of performance-augmenting exoskeleton systems is the unknownnonlinear dynamic properties of the systems or the uncertainties. An improper estimation of the systemdynamics can cause instabilities in the system and generate considerable human-exoskeleton interaction forcesduring human motions. Thus, the controller of such exoskeleton systems needs to add robustness to stabilize itagainst the uncertainties. In this paper, we propose a global fast sliding mode control algorithm integrated in ahybrid controller for each exoskeleton leg to minimize human-exoskeleton interaction forces. By doing so, theproposed algorithm does not require an exact estimation of the dynamic properties of the exoskeleton system, butstill minimizes the physical human-exoskeleton interaction (pHEI) forces. Finally, the performance of the proposedalgorithm is verified by experiments on our lower exoskeleton system, which is used for human power augmentationand called “PRMI” exoskeleton. Our experimental results show that the proposed control algorithm provides agood control quality for the PRMI exoskeleton. The PRMI exoskeleton can support a wearer carrying heavy loadwhile tracking the rapid movements of the wearer without obstructing them.