This paper systematically covers the significant developments of the last decade, including surrogate modeling of electrical machines and direct and stochastic search algorithms for both single- and multi-objective design optimization problems. The specific challenges and the dedicated algorithms for electric machine design are discussed, followed by benchmark studies comparing response surface (RS) and differential evolution (DE) algorithms on a permanent-magnet-synchronous-motor design with five independent variables and a strong nonlinear multiobjective Pareto front and on a function with eleven independent variables. The results show that RS and DE are comparable when the optimization employs only a small number of candidate designs and DE performs better when more candidates are considered.

Two new models for specific power losses in cold-rolled motor lamination steel are described together with procedures for coefficient identification from standard multifrequency Epstein or single sheet tests. The eddy-current and hysteresis loss coefficients of the improved models are dependent on induction (flux density) and/or frequency, and the errors are substantially lower than those of conventional models over a very wide range of sinusoidal excitation, from 20 Hz to 2 kHz and from 0.05 up to 2 T. The model that considers the coefficients to be variable, with the exception of the hysteresis loss power coefficient that has a constant value of 2, is superior in terms of applicability and phenomenological support. Also included are a comparative study of the material models on three samples of typical steel, mathematical formulations for the extension from the frequency to the time domain, and examples of validation from electrical machine studies.

A model of core losses, in which the hysteresis coefficients are variable with the frequency and induction (flux density) and the eddy-current and excess loss coefficients are variable only with the induction, is proposed. A procedure for identifying the model coefficients from multifrequency Epstein tests is described, and examples are provided for three typical grades of non-grain-oriented laminated steel suitable for electric motor manufacturing. Over a wide range of frequencies between 20-400 Hz and inductions from 0.05 to 2 T, the new model yielded much lower errors for the specific core losses than conventional models. The applicability of the model for electric machine analysis is also discussed, and examples from an interior permanent-magnet and an induction motor are included.

Performance trade offs for achieving high torque density and high drive-cycle efficiency in spoke-type permanent magnet (PM) motors are investigated. Unlike the traditional method, which relies on analytical equations, this trade-off study is conducted through a statistical analysis of the Pareto-optimal design candidates resulting from a large-scale design optimization of an example traction machine over a Le Mans-based drive cycle. In the design optimization process, both low-speed and extended-speed/field-weakening operating regions are evaluated using high-fidelity electromagnetic finite element (FE) simulations with an objective to simultaneously increase the torque density and decrease the power losses over the high energy-throughput-zones of the machine torque-speed plane. The resultant 3400 design candidates are utilized to investigate the impact of increasing power density on other important performance metrics such as power losses, PM demagnetization, and torque ripple. This analysis is supplemented by multiphysics simulation of three counterpart optimized designs and successful experimental verification of a prototype of one of those three designs which represents a record high power density motor in traction applications.

Computationally efficient finite-element analysis (FEA) (CE-FEA) fully exploits the symmetries of electric and magnetic circuits of sine-wave current-regulated synchronous machines and yields substantial savings of computational efforts. Motor performance is evaluated through Fourier analysis and a minimum number of magnetostatic solutions. The major steady-state performance indices (average torque, ripple and cogging torque, back-electromotive-force waveforms, and core losses) are satisfactorily estimated as compared with the results of detailed time-stepping (transient) FEA. In this paper, the CE-FEA method is presented and applied to a parametric design study for an interior-permanent-magnet machine. Significant reduction of simulation times is achieved (approximately two orders of magnitude), permitting a comprehensive search of large design spaces for optimization purposes. In this case study, the influence of three design variables, namely, stator tooth width, pole arc, and slot opening, on three performance indices, namely, average torque, efficiency, and full-load torque ripple, is examined, and design trends are derived. One hundred candidate designs are simulated in less than 20 minutes on a state-of-the-art workstation.

This paper presents an accurate and efficient method for predicting the iron losses in pulsewidth modulation (PWM) inverter-fed induction motors. The method was initially proposed for the prediction of iron losses in a nonoriented soft magnetic material with the same supply conditions. The proposed method requires the separation of the iron losses in the hysteresis and eddy-current components and the average rectified and rms of the PWM voltage values. Starting from the iron losses measured with a sinusoidal supply and the PWM waveform characteristics, a fast and reliable prediction of the iron losses in the motor with the new voltage supply can be obtained. The proposed method has been proved on an induction-motor prototype capable of providing good accuracy in the iron-loss measurement. The comparison between the measured and predicted iron losses with the PWM supply has shown excellent agreement, confirming the validity of the method.