Dynamic modelling of one degree of freedom pneumatic muscle-based actuator for industrial applications

Pneumatski umjetni mišići spadaju u skupinu nekonvencionalnog aktuatora za različitu biomedicinsku ili industrijsku primjenu. Jedan DOF aktuator sastoji se od dva mišića koji djeluju suprotno jedan od drugog (antagonistička veza) s pojednostavnjenom regulacijskom shemom (regulacijska petlja stabilnosti je isključena kako bi se pojednostavila regulacija i kako bi se postigla što viša stabilnost). Aktuator pokazuje nelinearno ponašanje sukladno prirodi pneumatskih mišića. Model mišića temelji se na modificiranom modelu mišića s dva elementa koji se sastoji od varijable amortizera i varijable opruge spojenih paralelno. Iz kinematičke točke gledišta, to je jednostavan mehanizam s jednim stupnjem slobode (ruka s dodanom masom koja rotira oko jedne okretne osi) s ravninom gibanja paralelnom s tlom. Glavna svrha ovog modela je kontrola konstrukcije, tako da je strogost kriterija točnosti smanjena u odnosu na stvarni model. Model je potvrđen uporabom rezultata dinamičkih pokusa sa stvarnom opremom u laboratorijskim uvjetima.Pneumatic artificial muscles belong to the group of nonconventional actuators intended for various biomedical or industrial applications. The one-DOF actuator (DOF – degree of freedom) consists of two muscles acting in opposition to each other (antagonistic connection) with simplified control scheme (stiffness control loop is excluded in order to simplify the control and to achieve as high stiffness as possible). The actuator exhibits nonlinear behaviour attributable to the compliant nature of pneumatic muscles. The muscle model is based on modified two-element muscle model consisting of a variable damper and a variable spring connected in parallel. From kinematic point of view, it is a simple mechanism with one degree of freedom (arm with added mass rotating around one revolute axis) with plane of movement being parallel to the ground. The main purpose of this model is a control design, so the stringency of accuracy criteria is lowered compared to a truth model. The model is validated using the results of dynamic experiments with real plant in laboratory conditions