LPV Modeling and Robust Control of Dynamic Wireless Charging for Electric Vehicles

As to the static wireless power transfer, a small-signal model is traditionally adopted to model the back-stage dc–dc buck converter at the receiving side. The model is determined by the specific operating point as well as the component parameters. For the dynamic wireless power transfer (DWPT), however, the model is time-varying as the operating point, such as the input dc voltage of the buck is changing due to the mutual inductance fluctuation. Hence, the traditional small-signal model is not suitable for the DWPT. Alternatively, a linear parameter varying (LPV)-based data-driven identification framework is proposed to depict the dynamic behavior of DWPT in the whole mutual inductance fluctuation range. Furthermore, the dual-model, the system model and the disturbance model, identification is realized with only one data acquisition. With the obtained models, the control frame integrating the feedforward and the feedback controller is established to suppress the fluctuation of the output voltage of dc–dc. Experiments results show that the proposed controller can keep the output stable under the condition that the periodic fluctuation range of mutual inductance reaches 50%.