Conference Agenda

An iterative domain decomposition method is proposed for numerical analysis of 3-Dimensional (3D) linear magnetostatic problems taking the magnetic vector potential as an unknown function. The iterative domain decomposition method is combined with the Preconditioned Conjugate Gradient (PCG) procedure and the Hierarchical Domain Decomposition Method (HDDM) which is adopted in parallel computing. Our previously employed preconditioner was the Neumann-Neumann (NN) preconditioner. Numerical results showed that the method was only effective for smaller problems. In this paper, we consider its improvement with the Balancing Domain Decomposition DIAGonal scaling (BDD-DIAG) preconditioner. Specially, the coarse matrix solver luckily uses the direct method and numerical results show that the preconditioner works well as the simplified diagonal scaling (diag) preconditioner which is well known and is used as the default preconditioner in the free software ADVENTURE_Magnetic.

Wireless power transfer (WPT) has been widely used to charge mobile devices and electric vehicles (EV) because of its safety and convenience. However, due to the large air gap between the transmitting (Tx) and receiving (Rx) coils, this kind of loosely coupled transformer has a large leakage magnetic field which affects power transmission. To overcome this issue, a planar magnetic flux concentrator (MFC) is proposed for use in a WPT system. The magnetic field and current distributions of the Tx and Rx coils integrated with a planar MFC are investigated. The simulation results show that in the case of the Tx coil with an MFC, the magnetic flux density of the Tx and Rx coils increases around the centre hole (which can increase the power transfer) and decreases on the outer surface of the Tx and Rx coils (which can reduce leakage magnetic field). By using a small size Rx coil the receiving power can be increased. Magnetic shielding can be obtained by a short-circuited MFC (as a conducting plate). The equivalent T-circuits for the Tx and Rx coils with and without an MFC are proposed based on the impedance analysis. Finally, we optimize the MFC design to achieve a better result by using a thicker MFC with smaller slit width.

Aiming at the problems of data security and transaction fairness in existing entrusted computations schemes, considering the tamper proof, unforgeable and decentralized characteristics of blockchain technology, an entrusted computations architecture for

numerical simulation of electromagnetic comprehensive performance under the framework of blockchain is proposed. Based on the multi-level data situation and the practical requirements for entrusted computations, we propose a multi-level data processing method and use smart contracts to realize the viewing strategy for private data. A detailed transaction mechanism with multi-level data security and multi-party transaction fairness is built. The feasibility and performance are proved by two examples.

In this paper, a method based on simplified prism vector (SPV) basis functions and best uniform approximation (BUA) is proposed to solve the wideband electromagnetic scattering problems of planar thin dielectric structures. The SPV basis functions have been proved to be very effective in calculating electromagnetic scattering problems of thin dielectric structures. On this basis, the application of method BUA can reduce a lot of calculation time by adding only a small amount of memory when calculating wideband problems. Numerical results show the accuracy and efficiency of the proposed method.

Conventional nodal FEM has the problem of low efficiency in dealing with large three-dimensional eddy current fields. Based on the characteristics of parallelism and hierarchical structure of FEM, this work introduces various tensors to accelerate large-scale finite element simulations. High-order tensorized finite element method is performed and numerically verified, reducing the time of element calculation and the assembling of the stiff matrix and assembly process time by about two orders of magnitude. It can be concluded from our work that tensorization allows parallelization and efficient use of computer resources for accelerated computing.

When sweeping dielectric parameters by traditional finite-element method (FEM), discrete sampling points in the target range need to be solved one by one, which is very inefficient for large matrices. The authors propose a novel efficient local model-order reduction scheme for the tuning process of dielectric-metal mixed structure using the full-wave FEM. The local macro-modeling method (LMM) is utilized to compress the space where the media exists, which reduces the dimension of the matrix and improves the computational efficiency in the parametrized tuning problem. In this paper, the reduced-order model transmission matrix is innovatively embedded into other low-dimensional matrix blocks. Compared with the common scheme, the matrix with lower dimension is obtained and parallel computing ability is guaranteed. To ensure mathematical rigor, a brief conclusion of the convergence of the transmission matrix expansion is presented. The numerical results show that the embedded local macro-modeling method can greatly reduce the matrix dimension and ensure the calculation accuracy in a relatively wide range.

In this paper, a particular methodology is introduced based on a coupled 2D FEA and circuit approach for the representation of permanent magnets in electrical machines including eddy current losses due to slot harmonics and switching frequency of PWM inverter supplies. The governing equation is diffusion equation expressed in Cartesian two dimensional configuration in terms of magnetic vector potential while the gradient of the electric scalar potential term enables end effects consideration of eddy currents developed in the permanent magnets. Accurate representation of eddy currents flowing in well defined paths in the permanent magnets has been achieved under dissymmetrical field conditions by introducing a particular segmentation of the magnet end regions through appropriate equivalent circuit coupling. Comparison of the results obtained by 3D FEA and the introduced coupled 2D model in a C-core magnetic circuit has illustrated the accuracy and the feasibility of the proposed methodology.

We present a new geometrical numerical method that is able to deal with grids having curved, i.e. non-planar, faces. The fundamental problem with this kind of grids is that up-to-now no low-order convergent numerical method for diffusion-like problems is able to use only one discrete flux unknown for every curved face. Our new numerical scheme achieves exactly this result. The basic idea behind our method is to devise a new dual grid, which, contrary to standard approaches, is not constructed according to barycentric subdivision but instead is determined from the very geometry of curved faces. Once such a “generalized dual grid” is available, the methodology of our numerical method is exactly the same of the standard discrete geometric approaches, in the sense that the new dual grid is used in the scheme just like the standard barycentric one. Hence, all the solid foundations of geometric methods are immediately available also to our new numerical method.

The finite element method is the main performance analysis method of electromagnetic devices and systems, but it has the problem of long calculation time. To solve this problem, a magnetic field prediction method based on residual U-Net and self focusing Transformer encoder is proposed. Taking the magnetic fields of permanent magnet synchronous motor (PMSM) and amorphous alloy transformer (AMT) as the research object, the finite element models of PMSM and AMT are established and simulated. The geometric structure diagram and magnetic field cloud diagram of the finite element models of PMSM and AMT obtained from the simulation analysis are taken as the input and output of the neural network, and the corresponding magnetic field distribution is obtained by predicting the pixels of the images. The calculation example proves that ResUnet Transformer model can realize the fast and accurate prediction of magnetic field, and has a good supporting role for state calculation under complex conditions, fine simulation and topology optimization under multiple working conditions. It is also one of the key implementation methods of digital twin and virtual sensor.

The fast and accurate computation of loss and temperature-rise of gas-insulated transmission lines (GIL) is highly required for the optimal design and safe operation. To this end, a novel 2D multi-slice field-circuit coupling model for computing ohmic loss and a thermal network model for calculating temperature rise is proposed. The field-circuit coupled finite element method can quickly and accurately obtain the solid loss, which can take into account the affect of interphase shunts, skin effect and proximity effect. The 2D multi-slice equivalent method can include the structural impact of multiple interphase shunts and grounding wires corresponding to real-world setting. The calculated loss provides heat source for temperature-rise calculation, and the branch current analysis can provide guidance for the layout interval of interphase shunts in the engineering project.

A systematic technique for the precise calculation of the impact of an external point source to a pregnant female torso is developed in this paper. The generalized torso model introduces a pattern of spheres to consistently represent the muscle tissues, the organs around the abdomen, the breasts, and the uterus, while an additional concentric sphere represents the fetus(es). Based on this geometry, electric fields are rapidly extracted via the pertinent dyadic Green’s function, derived in terms of the superposition principle and the necessary surface boundary conditions. Numerical results for a Hertz dipole (i.e. antenna of a wireless network) around the waist verify the merits of the proposed formulation and characterize the sensitivity of the internal organs and fetus to the specific source.

This paper presents a way to apply the vector nodal meshless method using random distributions of vector directions and nodes. For this, a planar subdivision is created using the vector directions of the nodes. The planar subdivision is used only to choose the support nodes, and polygons with a maximum of six sides are considered. The method correctly interpolates different vector fields and finds the eigenvalues of a rectangular waveguide without spurious modes.

The temperature rise experiment of gas-insulated transmission lines (GIL) is helpful to study the overheating failure process. It is more cost-effective and convenient to use the scaled-down model for experiment research since building the prototype experiment is expensive and time-consuming. Equation analysis and dimensional analysis are used to investigate the similarity principle of the various physical fields based on mathematical model of electromagnetic-thermal-fluid coupling fields. The Ohmic loss is calculated by the finite element method. The scaled-down model’s associated geometric, physical quantity, and boundary conditions are properly designed, and the similarity criterion is optimized to improve the scaled-down model’s viability. The proposed scaled-down model is validated using numerical computation and scaled-down model experiments that compare the results of ohmic loss and temperature field distribution.

This paper proposes an original approach to efficiently calibrate individual coil currents in a closed-loop degaussing system dedicated to the magnetic silence of a ferromagnetic ship. The conventional open-loop degaussing (OLDG) is first conducted based on the off-board magnetic field data. In the best condition of OLDG, signals measured at on-board magnetic sensors are defined as target field values in an optimization problem for closed-loop degaussing (CLDG). The coil currents are then updated by solving the problem in real time to make every on-board sensor always keep its allotted target value. Through the two-stage procedure, a high-performance magnetic silent ship can be easily achieved without predicting off-board magnetic fields created by the hull magnetization. To check the validity of the proposed method, an elaborate ship mock-up equipped with 14 degaussing coils and 12 magnetic sensors on board is produced, and it is tested in scale-down magnetic treatment facilities (MTF) according to two preset scenarios.

Automobile accelerator pedal is related to people's life and property safety. The automobile active safety technology of accelerator pedal has attracted much attention of experts. This paper presents a new type of magnetorheological damper against misstep transposition. Considering the driver's misstep, the damper can respond in time and prevent traffic accidents. Combined with ergonomics and mechanical design, the structural design of the anti-mistaken stepping magnetorheological damper is carried out to simulate the operation state in different states. The suitable current is obtained by the output mechanical model, and the two dimensional electromagnetic field simulation is carried out inside the damper. The multi-physical field simulation simulates the performance of the device under real conditions. Finally, the device is numerically optimized in combination with the simulation situation. After optimization, the device realizes lightweight and greatly improves the damping force, and improves people's confidence in the safe driving technology of automobiles.