2:20pm - 2:40pm
Shape Optimization of Rotating Electric Machines using Isogeometric Analysis and Harmonic Stator-Rotor Coupling
1Institut für Angewandte Mathematik, Technische Universität Graz, Austria; 2Institut für Teilchenbeschleunigung und Elektromagnetische Felder (TEMF), Technische Universität Darmstadt, Germany; 3Centre for Computational Engineering, Technische Universität Darmstadt, Germany
This work deals with shape optimization of electric machines using isogeometric analysis. Isogeometric analysis is particularly well suited for shape optimization as it allows to easily modify the geometry without remeshing the domain. A 6-pole permanent magnet synchronous machine is modeled using a multipatch isogeometric approach and rotation of the machine is realized by modeling the stator and rotor domain separately and coupling them at the interface using harmonic basis functions. Shape optimization is applied to the model minimizing a goal function, e.g. the total harmonic distortion of the electromotive force.
2:40pm - 3:00pm
Efficiency Map Prediction of Motor Drives using Deep Learning
McGill University, Canada
In this paper, a new method of predicting efficiency maps of electric motor drives is proposed using deep learning. Due to the need for simulating a large number of finite element models to estimate the efficiency map of one motor drive topology with certain geometry dimensions and materials, incorporating the whole efficiency map into the design optimization process is an overwhelmingly time-consuming task and maybe impossible depending on the quality of computational resources. Therefore, deep learning networks are employed here to generate a fast and accurate efficiency map prediction. Basically, the two important stages of efficiency map calculations, i.e. the flux linkage and torque-speed maps, are replaced by a combination of a deep network and a feedforward neural network to account for the geometry, materials and operating point variations. The output of the proposed method, in terms of the run-time as well as the prediction accuracy, are then compared with that of the finite element solution. The results show a good match.
3:00pm - 3:20pm
Deep Learning and Reduced Models for Fast Optimization in Electromagnetics
University of Pisa, Italy
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 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 FE model or the reduced FE model. The overall computational time of the optimization procedure is significantly reduced.
3:20pm - 3:40pm
New Selection Strategy of Many-Objective Optimization Based on Genetic Algorithm for the Design of Electrical Machines
Seoul National University, Korea, Republic of (South Korea)
This paper presents new selection strategy of many-objective optimization for the design of electrical machines. In order to reduce objective functions, the characteristics of the machine are classified into a major group and a minor group. We also proposed a simple selection strategy using satisfaction index and separation distance to deal with allowable criteria of the characteristics. To verify the proposed strategy, the rotor structure of synchronous reluctance motor was designed through the finite element method.
3:40pm - 4:00pm
A Benchmark TEAM Problem for Multi-Objective Pareto Optimization in Magnetics: the time harmonic regime
1University of Pavia, Italy; 2University of Padova, Italy; 3McGill University, Montreal, Canada; 4University of Insubria, Italy; 5University of Southampton, UK
The paper reformulates and generalizes the TEAM benchmark, originally proposed for multiobjective optimization of magnetic devices in DC regime, extending it to the AC regime. A solution is furnished which has enabled an extensive search and reliable estimation of the shape of the Pareto front. Field uniformity and losses are considered with reference to a class of power inductors. It is argued that the benchmark will provide a challenging target for new algorithms, especially those involving numerical modelling using finite element codes where the number of objective function calls needs to be minimized for practical designs.