Model Predictive Control of Wireless Power Transfer System in Seawater Environment

Wireless power transfer (WPT)systems have received more and more attention in undersea applications in recent years. The seawater between the transmitter and the receiver is a highly-conductive medium, which can be regarded as an unknown topology that will introduce eddy current loss and unknown dynamic behavior during the power transfer process. Hence, the dynamic modeling and control design methods for WPT systems in air may not apply. In this article, we propose an observer-free model predictive control (MPC) strategy for WPT systems in seawater environment. Considering the unknown topology introduced by the seawater medium, the control system is designed using a data-driven dynamic model obtained by simplified refined instrumental variable (SRIV) method. In a further step, the SRIV-based dynamic model is converted to a special state-space model by choosing a set of state variables corresponding to the input and output variables. Thus, the state observer design is avoided. Besides, operational constraints are imposed into the MPC algorithm to guarantee that the control input is implemented in an appropriate range. Experiments are performed to demonstrate that the proposed MPC system has superior performance in both reference tracking and parametric robustness in comparison to a proportional-integral control system.