A computationally efficient norm optimal iterative learning control approach for LTV systems

A unified data-driven design framework of optimality-based generalized iterative learning control

“…It is worth pointing out that the ILC (Arimoto et al, 1984) was originally proposed for nonlinear uncertain systems directly using I/O data for the controller design without requiring the exact knowledge of the system model and thus is classified as data-driven control as illustrated in Hou and Wang (2013). However, the above optimal GILC Hwang et al (1991), Amann et al (1996), Lee et al (2000), Gunnarsson and Norrlof (2001), Sun and Alleyne (2014), Freeman et al (2011), Freeman (2012, Son and Ahn (2011) and Son et al (2013) can only be categorized into "model-based control" because the knowledge of an accurate linear model of the controlled system is required for the controller design. When the model is inaccurate, the monotonic convergence of optimal GILC is no longer guaranteed, and learning transients with large, rapid growth of the error or even instability can occur.…”

“…In many of the real processes, the only available measurement is the terminal state or terminal output and the ultimate control objective is also the terminal state or terminal output instead of the entire trajectory of the system output. It is obvious that the conventional ILC (Uchiyama, 1978;Arimoto et al, 1984;Xu and Tan, 2003;Bristow et al, 2006;Ahn et al, 2007;Saab, 1994;Park et al, 1999;Sun and Wang, 2002;Tayebi, 2004;Xu and Xu, 2004;Rotariu et al, 2008;Chi et al, 2008;Hwang et al, 1991;Amann et al, 1996;Lee et al, 2000;Gunnarsson and Norrlof, 2001;Sun and Alleyne, 2014) cannot be applied to this type of control tasks because the exact measurement of the system state or output is not possible. Hence terminal iterative learning control (TILC) (Freeman, 2012;Son and Ahn, 2011;Xu et al, 1999;Gauthier and Boulet, 2009;Flores-Cerrillo and MacGregor, 2005) has been proposed to handle only terminal points at prescribed time instants rather than the whole trajectory over all time instants.…”

“…The first control scenario of tracking an entire reference trajectory is most common in practical applications (Uchiyama, 1978;Arimoto et al, 1984;Xu and Tan, 2003;Bristow et al, 2006;Ahn et al, 2007;Saab, 1994;Park et al, 1999;Sun and Wang, 2002;Tayebi, 2004;Xu and Xu, 2004;Rotariu et al, 2008;Chi et al, 2008;Hwang et al, 1991;Amann et al, 1996;Lee et al, 2000;Gunnarsson and Norrlof, 2001;Sun and Alleyne, 2014) 1994; Park et al, 1999;Sun and Wang, 2002), Lyapunov function based adaptive ILC (AILC) (Tayebi, 2004;Xu and Xu, 2004;Rotariu et al, 2008;Chi et al, 2008), and optimization based optimal ILC (OILC) (Hwang et al, 1991;Amann et al, 1996;Lee et al, 2000;Gunnarsson and Norrlof, 2001;Sun and Alleyne, 2014). The optimal ILC is most popular in practice because it can reject the undesirable large transient behavior that exists in PID-ILCs and AILCs, and has a monotonic convergence performance along the iteration direction.…”

Forgetting Factor Nonlinear Functional Analysis for Iterative Learning System with Time-Varying Disturbances and Unknown Uncertain

“…ILC algorithms can improve the trajectory tracking capability to reach better tracking results with an improving the numbers of iterations by utilizing the repetitive characteristics of the training process. As certain researches [6][7][8][9] have shown, it can guaranteed all the errors of tracking in the finite time interval are converging to zero as the learning numbers methods to infinity.…”

High Order Feedback-Feedforward Iterative Learning Control Scheme with a Variable Forgetting Factor

“…ILC algorithms can asymptotically or exponentially improve the tracking performance to achieve perfect tracking with an increasing number of iterations by utilizing the repetitive nature of the learning process. As several surveys [4][5][6] have discussed the novel ideas and development of ILC methodology, we refer the reader to these references for more information on the main concepts of ILC.…”