Observer-based Predictive Control of Nonlinear Clutchless Automated Manual Transmission for Pure Electric Vehicles: an LPV Approach

This study develops a novel robust control approach for a nonlinear clutchless automated manual transmission (CAMT) in pure electric vehicles. The developed approach comprises model predictive control (MPC), observer, and polytopic linear parameter varying (LPV) model with inexactly measured scheduling parameters. The stability of the online-designed MPC and the offline-designed observer, as well as the overall closed-loop system, is guaranteed by presenting the quadratic Lyapunov function stability conditions in terms of linear matrix inequalities (LMIs). Moreover, by the means of sampler and zero-order-hold (ZOH), the discrete-time LPV-MPC is merged with continuous-time LPV-observer to construct the overall observer-based controller. The first contribution of the presented approach is that the issue of inexactly measured scheduling parameters is considered, which avoids the simplifying separation principal technique. The second contribution is assuring the stability of the closed-loop system with hybrid continuous- and discrete-time systems. Also, the LPV-observer design procedure utilizes the singular value decomposition (SVD) method to reduce the conservativeness; and, a constrained one-step-ahead PV-MPC is suggested. Finally, to illustrate the performance improvement and optimality of the developed controller, it is applied to a nonlinear CAMT system and comparison numerical results are given.