Search

This article considers the data-driven modeling of a class of phase-controlled wireless power transfer (WPT) systems, where the load may vary slowly with respect to time. The dominant mode analysis suggests that a model of the Hammerstein type, which consists of a static nonlinearity function, followed by a linear time-varying model with a pure time delay, is the best structure to describe the input–output relationship of the system. On this basis, we derive a small-signal model that is linear in the variables in order to aid control design and allow the associated model parameters to be estimated from sampled input–output data using the standard refined instrumental variable (RIV) method. In the presence of a time-varying load, however, the plant model parameters may not be correctly estimated if the load response is not removed. In order to address this problem, a new recursive RIV method is proposed, in which an effective technique is introduced to track the load response, so allowing the parameters and time delay of the time-varying model to be accurately estimated. The effectiveness of the proposed method is verified by applying it to both a simulation model and a laboratory system. [ABSTRACT FROM AUTHOR]

The wireless power transfer (WPT) system is an interesting alternative for charging electric vehicles and has a broad prospect in public transport applications. In order to obtain optimal power transfer, the existing WPT systems often require perfect alignment between the primary and secondary coils, which can hardly be met in real applications. To relieve this problem and achieve the goal of free positioning, a one-to-multiple WPT topology being able to improve the misalignment tolerance at the secondary side is proposed in this letter. Based on circuit analysis and MATLAB simulation, a mixed mode is proposed to obtain a steady power transfer with horizontal misalignment. To verify the proposed topology, a one-to-two WPT system with a three winding transformer is developed in this letter, showing that proposed topology with the mixed mode can significantly increase the positioning adaptability. [ABSTRACT FROM AUTHOR]

This brief investigates recursive instrumental variable (IV) identification of time-delayed continuous-time (CT) models from the sampled input–output data. The refined IV method is extended for the first time to identify recursively both the parameters and time delay of a CT transfer function model. It is shown that the proposed method is able to yield consistent parameter estimation. Due to the nonlinear nature of the loss function to be optimized, time-delay estimation suffers from local minima. Thus, the global convergence performance relies on the quality of initial parameters. To increase the chance of converging to the global optimum, several guidelines are suggested to help the choice of starting values. The main results derived in this brief are confirmed by numerical examples and an experimental application. [ABSTRACT FROM AUTHOR]

This paper is concerned with control-oriented modeling of a class of wireless power transfer systems with a phase-shift-controlled inverter. Two methods are investigated: the analytical modeling method, which is based on physical laws, and the refined instrumental variable method, which is based on sampled data. It shows that the former provides physical insights into the system, and that the resulting models will be accurate if the true circuit component parameters are a priori known. Otherwise, the model accuracy may decline. By contrast, the latter provides a cheap solution for accurate modeling, based only on a set of sampled input–output data, and the resulting model can capture very well the dynamic behavior of the system within a defined operating range. In real applications, it is always hard to know the actual circuit component parameters, since they may drift from the nominal values, due to some reasons such as the aging of circuit components and the variation of coil placement, making the analytical modeling method not very accurate. Therefore, to preserve the virtues of both methods, it is suggested to use them as a combination, that is, using the analytical modeling method to determine the model structure while using the data-driven modeling method to estimate the model parameters. All results derived in this paper are verified by means of both numerical simulation and experimental validation. [ABSTRACT FROM AUTHOR]