“…The IPMC’s high potential for various applications in robotic actuators (Bar-Cohen et al, 2000; McDaid et al, 2010; Manley et al, 2009; Sadeghipour et al, 1992; Santos et al, 2010; Takagi et al, 2006), artificial muscles (Annabestani et al, 2016a, 2018; Shahinpoor and Kim, 2000), and dynamic sensors (Shahinpoor, 2003) has made the behavior identification and prediction of this smart material as a field of interest. Precision of the most models that were introduced to predict IPMC’s behavior (such as Volterra series (Kothera and Leo, 2005), neural networks (Truong and Ahn, 2011), Box–Jenkins and auto-regressive moving average with exogenous inputs (ARMAX) methods (Yun and Kim, 2006a, 2006b), nonlinear auto-regressive with exogenous inputs (NARX) structure fuzzy model and particle swarm optimization methods (Chi et al, 2011; Nam and Ahn, 2012), ANFIS-NARX model (Annabestani and Naghavi, 2014a, 2014b), and ANFIS-NARX with hysteresis modification (Zamyad et al, 2018)) depends on the output feedback (Zamyad and Naghavi, 2018).…”