Position control of a servo-pneumatic actuator using generalized Maxwell-Slip friction model

Abstract

Servo-Pneumatic actuators introduce some advantages such as cleanliness, high specific power and actuating force with high performance, comparing to electrical and hydraulic actuators. Servo-Pneumatic actuators are employed in many applications such as non-invasive surgeries, and rehabilitation robotics. High friction in servo-pneumatic actuators, due to the cylinder tight sealing, results in low controllability and decreases the system trajectory tracking performance. Therefore, position control of these actuators is a challenging issue. This paper presents a new approach for modeling the friction force called Generalized Maxwell-Slip model. According to GMS model, friction is constituted of two stages, namely “presliding” and “sliding”. Presliding stage is referring to friction force as velocity crosses zero. Moreover, friction illustrates a hysteresis behavior commensurate to displacement; therefore friction could be defined as a function of displacement in the presliding stage. Conversely, friction force in sliding stage is more of a function of velocity. Acquiring a model for hysteresis behavior is a problematic issue, which is resolved by using Generalized Prandtl-Ishlinski model. Consequently, attaining an accurate model for friction could help us to compensate this highly nonlinear phenomenon more precisely. Eventually, to achieve desirable tracking performance, multiple surface sliding mode controller is employed. To verify the effectiveness of the proposed approach, an experimental setup is developed and the acquired results show great improvement in tracking performance.