Search

This paper proposes a multi-objective optimization method for permanent magnet motors using a fast optimization algorithm, Covariance Matrix Adaptation Evolution Strategy (CMA-ES), and deep learning. Multi-objective optimization with topology optimization is effective in the design of permanent magnet motors. Although CMA-ES needs fewer population size than genetic algorithm for single objective problems, this is not evident for multi-objective problems. For this reason, the proposed method generates training data by solving the single-objective optimization multiple times using CMA-ES, and constructs a deep neural network (NN) based on the data to predict performance from motor images at high speed. The deep NN is then used for fast solution of multi-objective optimization problems. Numerical examples demonstrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

In this study, we propose a fast topology optimization (TO) method based on a deep neural network (DNN) that predicts the current-dependent motor torque characteristics using its cross-sectional image. The trained DNN is shown to provide the current condition that provides the maximum torque under the assumed motor control method. The proposed method helps perform TO with a reduced number of field computations while maintaining a high search capability. [ABSTRACT FROM AUTHOR]

This article proposes a method for modeling electric devices based on a Cauer circuit whose circuit parameters are directly determined from measured or computed data using the adjoint variable method. It has been shown that electric devices that are governed by the quasi-static Maxwell's equation can be modeled by the Cauer circuit. From the synthesized Cauer circuit, eddy current losses can be evaluated for a wide frequency range. Moreover, it can be embedded into a circuit simulator to perform the time-domain analysis. It is shown that the Cauer circuit whose parameters are identified using the proposed method works better for a simple numerical model than that whose parameters are identified using a genetic algorithm. Moreover, the Cauer equivalent circuit of a reactor and a power inductor is synthesized from the measured data using the proposed method. It is shown that the input impedance of the reactor and power inductor is well approximated by the Cauer circuit over the frequency domain of interest. [ABSTRACT FROM AUTHOR]

In this paper, the multi-objective topology optimizations of wireless power transfer (WPT) devices with two different coil geometries are proposed for obtaining the designs with good balance between transfer efficiency and safety. For this purpose, the proposed method adopts the normalized Gaussian network (NGnet) and Non-dominated Sorting Genetic Algorithm II (NSGA-II). In addition, the optimization under the different constraint on ferrite volume is carried out to verify its influence on optimization results. It has been shown that the proposed method successfully provides the Pareto solution to the design problem of the WPT device. [ABSTRACT FROM AUTHOR]

This article proposes a two-step topology optimization method based on the normalized Gaussian functions. The proposed method is shown to be effective for design of a claw-pole alternator. In this method, the global search using the micro-genetic algorithm ($\boldsymbol {\mu }$ GA) is followed by the local search using the sensitivity analysis based on the adjoint variable method. The 3-D structure of the rotor is optimized using the proposed method under low- and high-speed conditions. [ABSTRACT FROM AUTHOR]

This paper presents the topology optimization of the magnetic core for a wireless power transfer (WPT) device. In this optimization, the coil and magnetic-core shapes for WPT are represented by the Gaussian basis functions. They are sequentially optimized using the genetic algorithm and 3-D finite-element analysis so that the coupling coefficient of the WPT device is maximized. Moreover, a robust optimization method is proposed to keep the interlinkage flux as large as possible against misalignment in the coils. It is shown by the computational and experimental results that the optimized device outperforms the conventional devices. [ABSTRACT FROM AUTHOR]

This paper presents the fast topology optimization methods for rotating machines based on deep learning. The cross-sectional image of electric motors and their performances obtained during a multi-objective topology optimization based on the finite-element method and genetic algorithm (GA) is used for training of the convolutional neural network (CNN). Two different approaches are proposed: 1) CNN trained by preliminary optimization with a small population for GA is used for the main optimization with a large population and 2) CNN is used for screening of torque performances in the optimization with respect to the motor efficiency. [ABSTRACT FROM AUTHOR]

This paper presents robust topology optimization of electromagnetic machines by considering the magnetic degradation caused by mechanical and thermal stresses in punching, shrinking fitting, and other manufacturing processes. The topology optimization is performed using two methods: one is the robust genetic algorithm in which random noises are added to the magnetic characteristic parameters and the other takes the deviations in the objective and constraint functions due to the degradation into account. These methods are applied to optimization of the flux barrier shapes in an interior permanent magnetic motor to find that one can successfully realize robust design. [ABSTRACT FROM PUBLISHER]

This paper presents the optimization of a meander line antenna finite-element model by means of an adaptive genetic algorithm (GA). To search for optimal antenna configurations, the present method employs a GA with an adaptive method that adjusts the characteristics of its selection, crossover, and mutation operators in order to maintain a diverse set of high-quality candidate solutions during its execution. It is shown that the present method can find an optimal solution faster than the conventional GA. Moreover, the fitness values of the optimal solutions obtained by the present method are better than those obtained by conventional GA. [ABSTRACT FROM AUTHOR]

This paper presents a non-conforming finite element (FE) method with tetrahedral elements, which is applied to 3-D shape optimization. In the present method, continuity of magnetic flux is directly enforced on the interfaces of different domains in which FE meshes are independently generated. The slave unknowns in one side which are linearly dependent on the unknowns in the other side are eliminated. It is shown that the present method yields accurate solutions to 3-D nonlinear magnetostatic problems. Moreover, evolutional shape optimization of an inductor is shown to be successfully carried out using the present method. [ABSTRACT FROM AUTHOR]

This paper reports novel design method of waveguide slot array antenna with dielectric lenses by using the evolutionary optimization. We adopt the micro-genetic algorithm (MGA) with the real-coded gene as the optimization method and the finite-difference time-domain (FDTD) method to evaluate the objective function. The purpose of this study is to establish the design method of the above antenna by optimizing slot length and offset parameters. In the optimization of the antenna aperture distribution, it is shown that the optimized aperture field becomes sufficiently uniform, and the side-lobe levels (SLL) are improved in comparison with those of conventional antennas. Another optimization in which the objective function is evaluated by the array factor controls directly the side-lobe levels, to design low side-lobe antennas. By using the proposed method, the target antenna is optimized to satisfy the given SLL. It is confirmed that the proposed design method is a promising tool for designing the antennas loaded with dielectric materials. [ABSTRACT FROM PUBLISHER]

This paper presents optimization of meander line antennas (MLA) for passive IC tags which contain nonlinear circuits. In the present analysis, the tag antenna loaded by the Cockcroft–Walton circuit is analyzed by solving coupling equations for electromagnetic waves and nonlinear circuit with FDTD method. The antenna shape is optimized so that the output voltage of the Cockcroft–Walton circuit is maximized using the micro genetic algorithm (\mu\ GA). The computational burden in the optimization processes is effectively suppressed by truncating the transient computations. A criterion is introduced to evaluate the errors caused by the truncation. The present method yields the optimal MLA which has the greater output voltage in comparison with the dipole antenna. [ABSTRACT FROM PUBLISHER]

This paper presents evolutional design of antennas for UHF-band passive IC tags. The wire-antenna shapes are optimized using the genetic algorithm (GA) aiming at maximization of feeding power to the IC tag as well as minimization of the antenna size. The characteristics of the antennas printed on a dielectric substrate are evaluated using the moment method in the optimization processes. The genotypes for GA are introduced to express the shapes of lattice antennas, which include loops and floating lines, and spiral antennas. The present genotypes for the latter are also valid for optimization of spiral structure of, for example, inductors and filters. It is shown that optimized lattice and spiral antennas have smaller sizes in comparison with the conventional meander antennas. [ABSTRACT FROM PUBLISHER]

This paper presents parameter and topology optimization of inductor shapes using evolutionary algorithms. The goal of the optimization is to reduce the size of inductors satisfying the specifications on inductance values under weak and strong bias-current conditions. The inductance values are computed from the finite-element (FE) method taking magnetic saturation into account. The result of the parameter optimization, which leads to significant reduction in the volume, is realized for test, and the dependence of inductance on bias currents is experimentally measured, which is shown to agree well with the computed values. Moreover, novel methods are introduced for topology optimization to obtain inductor shapes with homogeneous ferrite cores suitable for mass production. [ABSTRACT FROM AUTHOR]

Although probabilistic optimization methods based on genetic algorithm (GA) provides accurate results, its performance is sometimes considerably sensitive to parameter changes. Moreover, the constraints are violated due to such parameter changes. A robust GA which performs random perturbation during optimization processes has been applied to some mathematical problems to show that it works as fast as the usual GAs. An adequate elite reservation technique for the robust GA is presented and applied to the robust GA for electromagnetic problems. Moreover, this method is shown to find solutions which are kept feasible against parameter changes. [ABSTRACT FROM AUTHOR]

Robust optimization searches for optimal solutions which remain feasible under the effect of uncertainties. The conventional robust optimization, however, often requires great computational cost to find robust solutions. This paper introduces a novel robust optimization method based on genetic algorithm. In this method, individuals with "good" performance are selected and tested for their robustness for reduction of computational time. It is shown that the present method works in shorter time in comparison with the conventional robust optimization. [ABSTRACT FROM AUTHOR]