Conference Agenda

Overview and details of the sessions of this conference. Please select a date or location to show only sessions at that day or location. Please select a single session for detailed view (with abstracts and downloads if available).

Session Overview
PB-M5: Optimization and Design
Wednesday, 17/Jul/2019:
10:50am - 12:40pm

Session Chair: Siguang An
Session Chair: Didier Trichet
Location: Patio 44-55

Show help for 'Increase or decrease the abstract text size'

Optimization of an Electromagnetic Actuator using Topology-Optimization

Philipp Seebacher1, Manfred Kaltenbacher1, Fabian Wein2, Henry Lehmann1

1Technical University of Vienna, Austria; 2University of Erlangen-Nuremberg, Germany

We apply the SIMP (Solid Isotropic Material with Penalization) model for the topology optimization of an electromagnetic actuator. Thereby the gradient of the object function is computed by the adjoint method. The objective is to maximize the electromagnetic force between yoke and anchor, including the consideration of nonlinear material behaviour.

Global Transformer Design Optimization Using Multiobjective Cheetah Algorithm

Leandro dos Santos Coelho1,2, Carlos Eduardo Klein2, Viviana Cocco Mariani2,3, Mauricio Valencia Ferreira Da Luz4, Jean Vianei Leite4

1PPGEPS, Pontifical Catholic University of Parana (PUCPR), Curitiba, PR, Brazil; 2Department of Electrical Engineering, Federal University of Parana (UFPR), Curitiba, PR, Brazil; 3PPGEM, Pontifical Catholic University of Parana (PUCPR), Curitiba, PR, Brazil; 4GRUCAD-EEL-CTC, Federal University of Santa Catarina (UFSC), Florianopolis, SC, Brazil

Optimization plays important roles in electromagnetics and related fields because several real-world problems can be essentially modeled as multiobjective optimization problems. In terms of stochastic optimization approaches, the swarm intelligence (SI) algorithms are a family of nature-inspired algorithms widely used for solving complex optimization problems. SI algorithms are well-known for their ability to solve optimization problems with multiple objectives and/or constraints that may not always be continuous and/or differentiable. In this paper, a modified Cheetah based algorithm is adapted to multiobjective optimization (MOCBA) using mechanisms of external archiving and diversity preservation to solve a transformer design optimization (TDO) problem with two competing objectives. Simulations applied to a TDO problem demonstrate the effectiveness of the proposed multiobjective MOCBA as an alternative to well established stochastic population-based techniques in global optimization.

Design of a loop-gap resonator with bimodal uniform fields using finite element analysis

Matthew Libersky1, Daniel Silevitch1, Ammar Kouki2

1Caltech, United States of America; 2École de technologie supérieure, Canada

The loop-gap resonator (LGR) was originally developed to provide a uniform microwave magnetic field on a sample for electron spin resonance (ESR) experiments. The LGR is composed one or more loops and gaps acting as inductances and capacitances respectively. Typical LGR designs produce a uniform field on a sample at a single resonant frequency, but for certain experiments it is necessary to study the response of a material to uniform fields at multiple frequencies applied simultaneously. In this work we develop an empirical design procedure using finite element method calculations to design an asymmetric loop-gap resonator with uniform fields at two frequencies in the same sample volume and optimize the field uniformity, intensity, and filling factors. We will manufacture an optimized resonator and compare the measured S-parameters to the simulation.

Implementation of Beamforming Techniques Adapted to Realistic Microstrip Linear Antenna Arrays

Zaharias D. Zaharis1, Ioannis P. Gravas1, Traianos V. Yioultsis1, Christos S. Antonopoulos1, Pavlos I. Lazaridis2, Thomas D. Xenos1

1Aristotle University of Thessaloniki, Greece; 2University of Huddersfield, U.K.

Most of the beamforming techniques applied so far on antenna arrays make estimations on a theoretical basis concerning the main lobe direction and the directions of the radiation pattern nulls, because they treat the antenna array factor as the total radiation pattern without regarding the actual non-isotropic radiation pattern of the array elements and the mutual coupling between them, which both take place in reality. The present study introduces an effective modification of two popular beamforming techniques to be applied for the synthesis of current distribution for realistic microstrip linear antenna arrays. These inverse problem techniques are the minimum variance distortionless response method and the null steering beamforming method. In both methods, the actual radiation pattern of the microstrip elements and the mutual coupling between them are taken into account. The methods are tested in several scenarios, where one desired and one or more undesired (interference) signals are received from various directions of arrival (DoA) by a microstrip linear antenna array. A statistical analysis is applied in each scenario concerning the divergence of the main lobe direction from DoA of the desired signal, as well as the direction divergence of the nulls from DoA of the respective interference signals. The analysis results demonstrate the robustness of the modified beamforming techniques as well as the validity of the proposed modification.

Topology Optimization to Improve Vibration Response of Electric Motors for Hybrid Electric Vehicles

Jaejoon Lee, Jaewook Lee

School of Mechanical Engineering, Gwangju Institute of Science and Technology, Gwangju 61005, South Korea

In this study, topology optimization is conducted to improve vibration response of electric motors for hybrid electric vehicles. Specifically, the stator geometry of electric motor is designed to maximize the fundamental natural frequency while the motor torque satisfies the target value. If the fundamental natural frequency of the structure is higher than driving frequency, the vibration resonance can be prevented. To calculate natural frequency, the vibration model is built using modal analysis with the finite element method. The model is validated by investigating natural frequencies in the benchmark model. For the torque analysis, a simplified magnetic model using d- and q-axis inductances is adopted to reduce the computing time. To achieve the optimization goal, the stator structural shape of electric motors for hybrid electric vehicles is designed in two-dimension using topology optimization. The result of this study will be useful for developing high durability electric motors of hybrid electric vehicles.

Sensitivity Analysis and Design Optimization of a New Hybrid-Excited Dual PM Generator for Stand-Alone Wind Power Generation

Xing Zhao, Shuangxia Niu, Weinong Fu

The Hong Kong Polytechnic University, Hong Kong S.A.R. (China)

This paper presents a new hybrid-excited dual PM generator for stand-alone wind power generation. Compared with the existing counterparts, the key difference in the proposed design is to arrange consequent-pole PMs in stator slots. And in this way, symmetrical flux modulation effect is achieved between stator PMs and rotor PMs, which leads to enhanced torque density. Moreover, DC field coils are artifically arranged in stator slots to achieve flexible flux weakening operation, which enables extended constant-voltage speed range. In this paper, the machine structure and operation principle are introduced. Some leading design parameters are determined based on a sensitivity analysis. A finite-element and genetic algorithm combined method is adpoted to realize a multi-objective design optimization, and the optimal design of the proposed topology is proved to be a potential generator candidate for direct-drive wind power generation.

3-D Topology Optimization of Claw-Pole Alternator Using Gaussian-Basis Function with Global and Local Searches

Yoshitsugu Otomo1, Hajime Igarashi1, Yuki Hidaka2, Taiga Komatsu2, Masaki Yamada2

1Hokkaido University, Japan; 2Mitsubishi Electric Corporation, Japan

This paper proposes a two-step topology optimization method of a claw-pole alternator, in which the rotor shape is represented by linear combination of the normalized Gaussian functions. In this method, the global search using the micro-genetic algorithm is followed by the local search using the sensitivity analysis based on the adjoint variable method. The rotor core shape is determined so that the open voltage between phase coils is maximized. The local search is shown to play an important role to improve the performance. It is shown that the claw-pole alternator with the optimized rotor core structure outperforms the conventional alternator.

Optimization of Planar Magnet Array Using Digital Annealer

Akito Maruo1, Hajime Igarashi1, Hirotaka Oshima2, Satoshi Shimokawa2

1Hokkaido University, Japan; 2Fujitsu Laboratories Ltd., Japan

This paper presents a fast optimization of two-dimensional magnetic array using Digital Annealer that is a CMOS hardware developed by Fujitsu Co. to effectively solve quadratic unconstrained binary problems. The magnetic array structure, which can be used for a vibration energy harvester and planar motor, is optimized so that the induced voltage of coils placed above the array is maximized. It is shown that the trapezoidal magnet array which covers the two-dimensional plane is one of the natural extensions of the Halback array.

Multiobjective Coyote Algorithm Applied to Electromagnetic Optimization

Juliano Pierezan1, Leandro dos Santos Coelho1,2, Viviana Cocco Mariani1,3, Luiz Lebensztajn4

1Department of Electrical Engineering, Federal University of Parana, Curitiba, PR, Brazil; 2Industrial and Systems Engineering Grad. Program (PPGEPS), Pontifical Catholic University of Parana, Curitiba, PR, Brazil; 3Mechanical Engineering Graduate Program (PPGEM), Pontifical Catholic University of Parana, Curitiba, PR, Brazil; 4Escola Politécnica da Universidade de São Paulo, Brazil

The Coyote Optimization Algorithm (COA) is a population-based nature-inspired metaheuristic for global optimization that considers the social relations of the coyote proposed originally to single-objective optimization. In this paper, the numerical results are reported to validate a novel proposed multiobjective COA (MOCOA) to solve the Testing Electromagnetic Analysis Method (TEAM) workshop benchmark problem 25. Simulation results demonstrate the validity of the proposed MOCOA to find nondominated solutions that represent good trade-offs among the objectives in the evaluated problem.

Topology Optimization of Rotor in Synchronous Reluctance Motor Considering Torque Ripple by the GA with Cluster of Materials and Cleaning Procedure

Takeo Ishikawa

Gunma University, Japan

The authors have proposed a topology optimization method to optimize the distribution of materials by the Genetic algorithm considering the cluster of material and a cleaning procedure. This paper optimizes the topology of rotor structure of synchronous reluctance motor for a high average torque and a low torque ripple by considering two materials. When the effect of torque ripple on the fitness function is chosen to be an appropriate vale, several layers can be produced in the outside in rotor to reduce the torque ripple.

Synthesis of the mutual inductor of a Wireless Power Transfer Systems: a field-circuit approach

Manuele Bertoluzzo2, Paolo Di Barba3, Michele Forzan2, Maria Evelina Mognaschi3, Elisabetta Sieni1

1University of Insubria, Italy; 2University of Padova, Italy; 3University of Pavia, Italy

Wireless power transfer systems are a viable solution to solve the problems that still delay a widespread diffusion of electric vehicles. Performances of these systems are affected by the magnetic characteristic of the coupling coils, so that they should be carefully designed. This paper presents a two-stage process for the optimal design of the coils. In the first stage, the equivalent circuit of the coupled coils is synthesized using an analytical approach for the computation of the objective functions and a genetic algorithm for their minimization. In turn, in the second stage, the optimization acts on objective functions computed by FEM analysis and identifies the geometrical parameters of the coils with the aim of achieving the circuit parameters recognized as optimal at the end of the first stage. The effectiveness of the proposed method is checked comparing the obtained coil design with a laboratory prototype.

Sequence based Environment for Topology Optimization

Arbaaz Khan, David Lowther, Chetan Midha

McGill University, Canada

This paper proposes a new Topology Optimization method which transforms the material distribution problem into a movement sequence search problem for a controller moving in the design space. This method enforces connectivity between the cells in the discretized design domain that contain the same material. This removes the need of any filtering or smoothing of the optimal result to obtain a manufacturable design. The validity of this method is proved by employing a Genetic Algorithm on a C-core electromagnetic actuator problem. The results obtained showed similar performance when compared with a conventional discrete topology optimization method. This method also provides the ability to leverage the use of different search techniques such as tree search algorithms

Random Vector Functional Link Neural Network Approaches for Bouc-Wen Hysteresis Nonlinear Modelling

Leandro dos Santos Coelho1,2, André Armstrong Janino Cizotto1, Viviana Cocco Mariani1,3, Luiz Lebensztajn4

1Department of Electrical Engineering, Federal University of Parana, Curitiba, PR, Brazil; 2Industrial and Systems Engineering Grad. Program (PPGEPS), Pontifical Catholic University of Parana, Curitiba, PR, Brazil; 3Mechanical Engineering Graduate Program (PPGEM), Pontifical Catholic University of Parana, Curitiba, PR, Brazil; 4Escola Politécnica da Universidade de São Paulo, Brazil

Hysteresis is a phenomenology frequently encountered in a variety of science and engineering fields, such as electromagnetism, control systems and identification. On the other hand, system identification of hysteretic systems is challenging because dynamic nonlinearities act upon non-measurable internal state variables. In this paper is proposed random vector functional link (RVFL) network approaches for black-box nonlinear identification of a hysteretic nonlinear system. Several RVFL with different activation functions in the hidden layer were designed and validated. In addition, a recently proposed Bouc-Wen benchmark is used as a numerical example of the proposed RVFL approach. Results to the hysteretic nonlinear system demonstrated the effectiveness of the proposed RVFL as a competitive alternative with fast learning speed and good generalization.

Analysis of a Low Cost High Performance Axial Flux Vernier Machine with SMC Cores and Ferrite Magnets

Chengcheng Liu1, Youhua Wang1, Gang Lei2, Youguang Guo2, Jianguo Zhu3

1Hebei University of Technology, China, People's Republic of; 2University of technology sydney, Australia; 3University of Sydney, Australia

In This paper, a low cost high performance axial flux Vernier machine (AFVM) with soft magnetic composite (SMC) cores and ferrite magnet is proposed and analyzed. By cooperating with spoke magnet rotor structure and 3D magnetic flux stator structure, the proposed AFVM can output higher torque compared with the traditional permanent magnet machine though low cost low co-energy ferrite magnet and low permeability SMC material are adopted. First the main electromagnetic topology and operation principle of the proposed AFVM is introduced, second its main dimensions are investigated and optimized and then its electromagnetic parameter and performance are calculated based on the deduced analytical model and 3D finite element method (FEM), lastly its main performance will be compared with both the axial flux machine and traditional Vernier machine in regarding on the side of torque ability, power factor and efficiency.

Design of an Excitation Coil and Core to Improve the Measurement Accuracy in the Electromagnetic Flowmeter

Kyoung Yun Kim, Sang Hyeon Im, Gwan Soo Park

Pusan National University, Korea, Republic of (South Korea)

The electromagnetic flowmeter is a device that measures the amount of fluid flowing inside a pipe by an electromagnetic method. To improve measurement accuracy to reflect the velocity of fluid varying according to various conditions, the magnetic flux density on the measurement surface have to keep uniformity. In this paper, we propose a new excitation coil system design that can more precisely measure the irregular drift at the pipe wall by using 3D Finite Elements Method (FEM). In addition, the newly designed excitation coil system is manufactured as a prototype and then the magnetic flux density is measured between electrodes. Finally, the magnetic flux density measured at the prototype is compared with that calculated by the FEM analysis to verify the application possibility.

Analysis of the Torque Characteristics of Dual Airgap Spoke Type Permanent Magnet Vernier Machine considering Pole Ratio Effect

Mudassir Raza Siddiqi, Zia Ullah, Jin Hur

Incheon National University, Korea, Republic of (South Korea)

This paper investigates the effect of pole ratio on the performance of dual airgap spoke type permanent magnet vernier machine (PMVM). The study covers the analysis of electromagnetic torque characteristics considering the pole ratio effect on the dual airgap spoke type PMVMs. The performance parameters such as torque, torque ripple, torque density and efficiency in the dual airgap PMVMs with dual stator winding are analyzed by the variation of design parameters such as rotor and stator slot combinations, winding pole numbers and magnet dimensions. 2D-finite element method (FEM) is used to analyze the performance of the model and verify the effect of pole ratio on the torque performance of dual airgap spoke type PMVM.

Analysis of Magnetization Characteristic Based B-H curve fitting according to the Operation Conditions for traction motor

YoungHyun KIM, DaeHee YU, JungHo LEE

Hanbat national university, Korea, Republic of (South Korea)

This paper examines the demagnetization characteristics of rare earth permanent magnets (PMs) used in electric machinery. An analytical model was designed considering a controllable shape, temperature, and the external magnetic field in PMs used in the rotor of an interior permanent magnet synchronous motor (IPMSM) for driving an electric railway vehicle (ERV). The demagnetization process was analysed based on the magnetization and demagnetization theory for PM materials in computer simulation. The result of the analysis can be used to calculate the magnetization and demagnetization phenomenon according to the input B-H curve. So, this paper presents the conditions of demagnetization by the external magnetic field in the running and stopped states and proposes a simulation method that can analyse demagnetization phenomena according to each condition and be used to design the IPMSM that maximizes efficiency and torque characteristics.