“…Solving a nonlinear optimisation problem with constraints is computationally intensive, which is the main obstacle that blocks the real-time application of MPC to nonlinear systems. To reduce the computational burden of online optimisation, the differential flatness property has been exploited in many trajectory planning algorithms which are actually MPC problems (Berry, Howitt, Gu, & Postlethwaite, 2011;Cowling, Yakimenko, Whidborne, & Cooke, 2010;Faulwasser, Hagenmeyer, & Findeisen, 2014;Flores & Milam, 2006;Mahadevan, Agrawal, & Doyle, 2001;Prodan et al, 2013;Suryawan, De Don, & Seron, 2012). Loosely speaking, the differential flatness means that the flat system can be completely characterised by flat outputs and their higher order derivatives (Fliess, Lvine, Martin, & Rouchon, 1995).…”