The computational cost of topology optimization based on binary particle swarm optimization is greatly reduced by the use of Deep Neural Networks (DNN). A first Convolutional Neural Network (CNN) is trained with data coming from Finite Element Analysis (FEA) with the aim of correctly estimating the output quantity (a motor torque in the proposed case study). A second CNN is trained to give as output the boundary conditions (BC, expressed in terms of fields or potentials) to be used as BC of a reduced Finite Element Method (FEM) model, created in order to still be able to give the correct value of the output quantity. In the optimization phase, the torque properties are evaluated by the trained CNN, and only a reduced percentage of cases are re-evaluated by either the full FEM model or the reduced FEM model. The overall computational time of the optimization procedure is significantly reduced.

In this work a topology optimization procedure is proposed and applied to the TEAM 25 problem, i.e. a model of a die press with electromagnet for orientation of magnetic powder. The shape of the press is described as a free discretized profile, and its relation to the flux density in the cavity is simulated by Finite Element Analysis (FEA) and learned by a Deep Neural Network (DNN) model. The DNN is used as a surrogate model for optimization, aiming to obtain a desired flux density distribution in the cavity. Results are promising, as better accuracy is obtained with respect to full FEA-based optimization approach with reduced time cost. Once trained, surrogate model can be used to efficiently solve a whole family of problems where a different target field distribution is defined.

In this article, we demonstrate the near electric field reduction ability of spiral resonators arrays (SRAs) in resonant inductive wireless power transfer (WPT) applications. In particular, we show that a suitably designed planar SRA, placed in the proximity of the driver loop, can significantly decrease the electric field level at the operative frequency, thus reducing the potentially dangerous radiation (i.e., the specific absorption rate) in the case of human exposure. In addition, due to the thin structure of the SRAs, the system is able to enhance the performance with respect to a conventional simple driver–receiver system for a fixed working distance. A systematic study has been performed through electromagnetic simulations to evaluate the optimum number of array unit cells to eventually achieve a compromise between efficiency enhancement and electric field shielding on a two-coil system, selected as a test case. Finally, the numerical results have been confirmed through measurements conducted on fabricated prototypes of the proposed WPT configurations. The introduced solution can be remarkably useful in all the WPT applications requiring a close interaction with human operators where exposure is a concern, such as automotive applications, biomedical implants, and rechargeable devices.

In this paper, we introduce a general analytical procedure to unambiguously characterize a metasurface through its lumped circuital equivalent in resonant inductive Wireless Power Transfer (WPT) applications. The proposed model incorporates the finite extent of the slab, as well as the WPT near field operative regime and the presence of the particular driving/receiving coils arrangement, providing quantitative and easy-to-handle parameters which can be manipulated to achieve WPT performance enhancement. We first develop the theoretical background aimed at the lumped parameters extraction, which reveals, for WPT applications, more accurate and robust with respect to the conventional sub-wavelength homogenization theories based on infinite slab extent and impinging plane wave hypotheses. We provide some general guidelines for the design of metasurfaces for WPT performance enhancement based on the derived circuit model; afterwards, we numerically design a test-case consisting of two resonant coils (driver and receiver, respectively) with an interposed passive metasurface to verify the developed theory. Finally, we show some measurements performed on a fabricated prototype, that present an overall excellent agreement with both the lumped model and the numerical simulations.