Finite-Iteration Learning Tracking with Packet Losses

Abstract

Since the ILC was proposed, it has developed into a powerful control tool for theoretical study and practical applications. Because of its simplicity and effectiveness, ILC has been applied in many fields, such as fault-tolerant control Wang et al. (2016), Timoshenko Beam He et al. (2017), linear systems Amann et al. (1996); Bu and Hou (2016), nonlinear systems Bu et al. (2017); Yu and Li (2016); Wei et al. (2016), hard disk drives Graham and Callafon (2006); Wu and Tomizuka (2010), autonomous vehicles Chen and Moore (2002), data-driven control Janssens et al. (2013); Chi et al. (2015, 2016), and multi-agent systems Xiong et al. (2016); Meng et al. (2015); Meng and Moore (2016); Li and Li (2014). As an illustration, data-driven iterative learning is demonstrated in Chi et al. (2015) for a class of nonlinear discrete-time systems with high-order learning law. For discrete-time models, iterative learning is suggested in Xiong et al. (2016) by using the event-triggered technique. In Meng and Moore (2016), the associated formation among nodes converges to the anticipated one exponentially as the iteration step grows if the union set of the interactive graphs includes a spanning tree.