No abstract

No abstract

This paper proposes an adaptive trajectory tracking control strategy implemented on a parallel ankle rehabilitation robot with joint-space force distribution. This device is redundantly actuated by four pneumatic muscles (PMs) with three rotational degrees of freedom. Accurate trajectory tracking is achieved through a cascade controller with the position feedback in task space and force feedback in joint space, which enhances training safety by controlling each PM to be in tension in an appropriate level. At a high level, an adaptive algorithm is proposed to enable movement intention-directed trajectory adaptation. This can further help to improve training safety and encourage human-robot engagement. The pilot tests were conducted with an injured human ankle. The statistical data show that normalized root mean square deviation (NRMSD) values of trajectory tracking are all less than 2.3% and the PM force tracking being always controlled in tension, demonstrating its potential in assisting ankle therapy.

No abstract

No abstract

The promise of wearable assistive robotics cannot be realized without the development of actuators that mimic the behavior and form of biological muscles. Planar fluidic muscles known as Peano muscles or pouch motors have the potential to provide the high force and compliance of McKibben pneumatic artificial muscles with the low threshold pressure of pleated pneumatic artificial muscles. Yet they do so in a soft and slim form that can be discreetly distributed over the human body. This work is an investigation into the empirical modeling of the Peano muscle, the effect of its material on its performance, and its capabilities and limitations. We discovered that the Peano muscle could provide responsive and discreet actuation of soft and rigid bodies requiring strains between 15% and 30%. Ideally, they are made of non-viscoelastic materials with high tensile and low bending stiffnesses. While Sarosi et al's empirical model accurately captures its static behavior with an root mean square error of 10.2 N, their dynamic model overestimates oscillation frequency and damping. We propose that the Peano muscle be modeled by a parallel ideal contractile unit and viscoelastic element, both in series with another viscoelastic element.

No abstract

In this work, we explore the basic static and dynamic behavior of a hydraulically actuated Peano muscle and how its geometry affects key static and dynamic performance metrics. The Peano muscle, or pouch motor is a fluid powered artificial muscle. Similar to McKibben pneumatic artificial muscles (PAMs), it has the ability to generate the high forces of biological muscles with the low threshold pressure of pleated PAMs, but in a slim, easily distributed form. We found that Peano muscles have similar characteristics to other PAMs, but produce lower free-strains. A test rig capable of measuring high-speed flow rates with a Venturi tube revealed that their efficiency peaks at about 40% during highly dynamic movements. Peano muscles with more tubes and of a greater size do not move faster. Also, their muscle tubes should have an aspect ratio of at least 1:3 and channel width greater than 20% to maximize performance. These findings suggest that finite element modeling be used to optimize more complex Peano muscle geometries.