Machine learning and artificial neural networks have shown to be applicable in modeling and simulation of complex physical phenomena as well as creating surrogate models trained with physics-based simulation data for numerous applications that require good computational performance. In this article, we review widely the surrogate modeling concept and its applications in the electrical machine context. We present comprehensively a workflow for developing data-driven surrogate models including data generation with physics-based simulation and design of experiments, preprocessing of training data, and training and testing of the surrogates. We compare neural networks and gradient boosting decision trees in modeling and simulation of torque behavior of a permanent magnet synchronous machine together with selected design of experiments approaches with respect to surrogate accuracy and computational efficiency. In addition, an approach to utilizing domain knowledge to create a hybrid surrogate model in order to improve the surrogate accuracy is shown. The accuracy of the selected hybrid neural network was better than with the gradient boosting approach and was close to the finite element simulation, whereas its run-time efficiency was significantly better compared to the finite element simulation with a speed-up factor of over 2,000. In addition, combining the sampling methods provided better results than the selected methods alone.

This paper introduces an interpolation method based on snapshot approach to reduce the order of a nonlinear model for magnetostatic problems and facilitate the evaluation of the corresponding system of equations. The solution of the reduced model for a set of given inputs is built by interpolating the right singular vectors, obtained from singular value decomposition of snapshots. To validate the efficiency of this method, this reduced model is compared with full-order finite-element model. Furthermore, the proposed method is analyzed with respect to a more conventional model order reduction that is combined with the discrete empirical interpolation method. The orthogonal interpolation method is the most effective method to reduce model of an interior permanent magnet synchronous machine especially in term of computational time.

In the control of electric drives, inaccurate estimation of the motor parameters affects the robustness of the control. This effect is observed particularly when the linear lumped parameter-based models are employed to represent a machine composed of nonlinear magnetic materials. Finite element (FE) models consider the material nonlinearity accurately. However, implementing these models in a real-time system is challenging due to the computational burden. In this paper, we propose a fast-dynamic model, based on a model order reduction method, to control a permanent magnet synchronous machine in a wide range of speed. The stator currents in the rotor frame of reference are given as inputs to a reduced FE model, which computes the nodal values of the magnetic vector potential and thereafter the flux linkages. A discrete-time model is used to control the rotor speed and the stator current components. Experiments on a 2.2 kW interior permanent magnet synchronous machine verify the viability of the proposed model.

This paper presents a pre-normative roadmap that foresees the developments in the charging of heavy-duty electric vehicles (HD-EVs). It supports and facilitates the future standardization efforts of charging technologies by creating an overview of the popularity of charging technologies and the end users’ needs. The required input for the work was collected using a comprehensive investigation on the available charging technologies and their standardization, reviewing the existing roadmaps and research work, and conducting surveys and interviews of end users and technical stakeholders. According to the findings, a pantograph on the roof of a vehicle and plug-based charging are currently the most used charging interfaces. This trend is likely to continue in the future, since (1) pantographs on vehicle roofs, (2) pantographs on infrastructure, and (3) plugs were graded as charging interfaces with the highest potential by the participants of the technical survey. Static and conductive charging technologies show more potential than dynamic and wireless charging technologies. Nevertheless, inductive charging may be a future charging solution for HD-EVs if the current bottlenecks in the technology can be addressed. These bottlenecks include high prices, slightly lower efficiency, lack of standardization, the maximum achievable power, and safety concerns. Furthermore, interoperability was repeatedly mentioned as the main challenge for today’s charging technologies.

In this paper, proper orthogonal decomposition (POD) method is employed to build a reduced-order model from a high-order nonlinear permanent magnet synchronous machine model with multiple inputs. Three parameters are selected as the multiple inputs of the machine. These parameters are terminal current, angle of the terminal current, and rotation angle. To produce the lower-rank system, snapshots or instantaneous system states are projected onto a set of orthonormal basis functions with small dimension. The reduced model is then validated by comparing the vector potential, flux density distribution, and torque results of the original model, which indicates the capability of using the POD method in the multivariable input problems. The developed methodology can be used for fast simulations of the machine.

Electromagnetic force and torque computations for a bearingless synchronous reluctance motor present multi-input complexity due to the levitating auxiliary winding located in the machine active parts. The machine model complexity is reduced through the proper orthogonal decomposition method. For this purpose, a snapshot matrix is needed to be constructed. This paper proposes a methodology of choosing efficient snapshots for the model order reduction. The reduced model is then applied to compute the torque and electromagnetic forces on the rotor of the machine. The method presented for selecting the snapshots is shown to be more efficient than uniformly distributed snapshots and it reduces the number of finite element method evaluations needed for the construction of the reduced model.