Surrogate-Based Acoustic Noise Prediction of Electric Motors

Ibrahim, Issah; Silva, Rodrigo C. P.; Mohammadi, Mohammad Hossain; Ghorbanian, Vahid; Lowther, D.A.

doi:10.1109/tmag.2019.2945407

Cited by 30 publications

(13 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…All the studied speed Overall, a moderate correlation strength can be observed between L M and L S for all the studied motor structures. It must be emphasized that the higher the value of r, the closer the sound pressure level data points fall to a straight line, which means that even the cheapest regression models trained in Ibrahim et al (2020) can fit tolerably well to accurately predict L M . Such models, however, will not fit well at lower r values due to the randomized behavior of the sound pressure peaks resulting from resonance phenomena in most of the full motor samples, which seems to be the case in this observation.…”

Section: Resultsmentioning

confidence: 99%

“…Once the EM and structural CAD models are realized, they are then parameterized using the approach in Ibrahim et al (2020). Parameterization helps to reduce the time spent on model abstraction and geometry editing.…”

Section: Geometric Modeling and Parameterizationmentioning

confidence: 99%

“…The issues with acoustic noise have been studied extensively, and a survey of literature (Gieras et al , 2005; Al Nahlaoui et al , 2011) covers most of the important aspects up to date. While many of these works try to reduce the simulation time by focusing on sound power radiated directly from the stator (Islam and Husain, 2010; Mohammadi et al , 2018; Ibrahim et al , 2020; Vermaand and Balan, 1994; Al Nahlaoui et al , 2011; Ibrahim et al , 2019; Mohammadi et al , 2018), the overall acoustic behavior of a typical electric motor is a cumulative effect of other components such as the rotor, the housing, the support, etc. As far as the current state of the art is concerned, there are no known correlations between the stator sound pressure level (L S ) and the sound pressure level of the complete electric motor (L M ).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Correlating structural complexity and acoustic noise performance of electric motors

Ibrahim

Mohammadi

Ghorbanian

et al. 2022

COMPEL

Self Cite

View full text Add to dashboard Cite

Purpose Acoustic noise is a crucial performance index in the design of electrical machines. Due to the challenges associated with modelling a complete motor, the stator is often used to estimate the sound power in the prototyping stage. While this approach greatly reduces lengthy simulations, the actual sound power of the motor may not be known. But, from the acoustic noise standpoint, not much is known about the correlation between the stator and complete motor. This paper, therefore, aims to use the sound pressure levels of the stator and the full motor to investigate the existence of correlations in the interior permanent magnet synchronous motor. Design/methodology/approach A multiphysics simulation framework is proposed to evaluate the sound pressure levels of multiple motor geometries in a given design space. Then, a statistical analysis is performed on the calculated sound pressure levels of each geometry over a selected speed range to compare the correlation strength between the stator and the full model. Findings It was established that the stator and the complete motor model are moderately correlated. As such, a reliance on the stator sound power for design and optimization routines could yield inaccurate results. Originality/value The main contribution involves the use of statistical tools to study the relationship between sound pressure levels associated with the stator geometry and the complete electric motor by increasing the motor sample size to capture subtle acoustic correlation trends in the design space of the interior permanent magnet synchronous motor.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Geometric Modeling and Parameterizationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Correlating structural complexity and acoustic noise performance of electric motors

Ibrahim

Mohammadi

Ghorbanian

et al. 2022

COMPEL

Self Cite

View full text Add to dashboard Cite

show abstract

“…T HE rapid advancement in automotive electrification requires higher standards for Noise and Vibration (NVH) target setting and thus development of new methods for vibration mitigation. These can be categorized as either earlydesign stage methods, where the design space is ample, represented by machine cross-section optimization [1], [2], or latedesign stage methods where few modifications can be made. For the latter, NVH improvements can be either achieved through machine design methods [3], where small modification, such as rotor notching are made to the cross-section in order to shape the air-gap forces, or control methods where different pulse width modulation (PWM) techniques are adopted [4] or harmonic current control is performed.…”

Section: Introductionmentioning

confidence: 99%

PMSM Current Shaping for Minimum Joule Losses While Reducing Torque Ripple and Vibrations

et al. 2021

View full text Add to dashboard Cite

This paper presents a current shaping method for torque ripple and mean air-gap radial forceshape harmonics reduction, under minimum Joule losses, used for air-born and structure-born vibration reduction in 3-phase PMSM electric drives. The main source for structure-born noise in electric powertrains is the torque ripple, while the main source for air-born noise are the radial air-gap forces. The proposed method uses a Fourier-decomposed LUT model obtained from updating the 2D electromagnetic FE model using test-bench results. Modal testing is used to update the structural model, and the vibration response on the machine full RPM range is determined using the vibration-synthesis method. The proposed current shaping algorithm is deterministic and can be used on any topology of 3-phase PMSMs. The torque and mean air-gap force harmonics that are intended for reduction can be arbitrary selected and their minimization can be fully completed if the maximum current and voltage constraint are respected. On the machine under test, vibration reduction is accomplished, especially for the mechanical orders that interact with the second mode-shape.

show abstract

“…The main goal of experiments reported in this paper is to use an acoustic simulation approach for the automated optimization of a Schroeder diffuser based on machine learning. Acoustic simulations have very broad use in acoustics in tasks such as auralization [1], noise prediction [2], simulation-driven optimization of room acoustics, and acoustic devices [3][4][5][6][7]. Since the rise of GPU-accelerated computations, they are also being carried out more efficiently with respect to time and resources [8,9].…”

Section: Introductionmentioning

confidence: 99%

Reinforcement Learning Algorithm and FDTD-Based Simulation Applied to Schroeder Diffuser Design Optimization

Kurowski

Kostek

2021

IEEE Access

View full text Add to dashboard Cite

The aim of this paper is to propose a novel approach to the algorithmic design of Schroeder acoustic diffusers employing a deep learning optimization algorithm and a fitness function based on a computer simulation of the propagation of acoustic waves. The deep learning method employed for the research is a deep policy gradient algorithm. It is used as a tool for carrying out a sequential optimization process the goal of which is to maximize the fitness function based on parameters characterizing the autocorrelation diffusion coefficient of the designed acoustic diffuser. As the autocorrelation acoustic diffusion coefficients are calculated from the polar response of a diffuser, the FDTD (finite-difference timedomain) simulation method is used to obtain a set of impulse responses necessary to calculate the polar responses of the optimized Schroeder diffusers. The results obtained from optimization based on the deep learning algorithm were compared with the outcomes of an analogous algorithm employing a genetic algorithm, and based on random selection of the Schroeder diffuser well depth pattern. We found that the best result was achieved by the deep policy gradient, as it produced outcomes, which, in terms of the provided autocorrelation diffusion coefficient, were statistically better than properties of designs provided by two other baseline approaches.

show abstract

Surrogate-Based Acoustic Noise Prediction of Electric Motors

Cited by 30 publications

References 7 publications

Correlating structural complexity and acoustic noise performance of electric motors

Correlating structural complexity and acoustic noise performance of electric motors

PMSM Current Shaping for Minimum Joule Losses While Reducing Torque Ripple and Vibrations

Reinforcement Learning Algorithm and FDTD-Based Simulation Applied to Schroeder Diffuser Design Optimization

Contact Info

Product

Resources

About