scholarly journals Optimization of surface-mounted permanent magnet brushless AC motor using analytical model and differential evolution algorithm

2019 ◽  
Vol 70 (3) ◽  
pp. 208-217 ◽  
Author(s):  
Mohd Rezal Mohamed ◽  
Dahaman Ishak
Keyword(s):  
Finite Element ◽  
Finite Element Model ◽  
Differential Evolution ◽  
Permanent Magnet ◽  
Differential Evolution Algorithm ◽  
Element Model ◽  
Torque Ripple ◽  
Domain Model ◽  
Computational Time ◽  
Evolution Algorithm

Abstract This paper discusses the optimization of surface-mounted permanent magnet brushless AC (PMBLAC) motor using Analytical Sub-domain model with Differential Evolution Algorithm (ASDEA). Only two regions were considered in this analytical sub-domain model, ie magnet and airgap regions, with assistance of Complex Relative Permeance Function (CRPF) to account for the stator slotting effect. Five machine parameters were chosen to be optimized, namely the magnet arc-pole-pitch ratio, slot opening width, magnet thickness, airgap length and stator inner radius. The optimization process has four objectives, ie minimum torque ripple, low cogging torque, high efficiency, and high output torque. The results from the optimized ASDEA were compared with the Analytical Sub-domain Genetic Algorithm (ASGA) and further validated against 2-D finite element model (FEM). Results show a good agreement between analytically optimized models and finite element model. The ASDEA has faster computational time compared to ASGA, and this provides benefit in terms of reducing the machine design parameterization time and less redundancy work required to achieve motor design specifications.

Methods for Reducing a Thermal Finite Element Model Including the Advection Network Contribution

2006 ◽  
Author(s):  
Daniele Botto ◽  
Stefano Zucca ◽  
Muzio M. Gola
Keyword(s):  
Finite Element ◽  
Finite Element Model ◽  
Real Time ◽  
Residual Life ◽  
Mathematical Formulation ◽  
Time Integration ◽  
Element Model ◽  
Computational Time ◽  
Fatigue Damage Accumulation ◽  
Secondary Air

The life monitoring concept needs on-line calculation to evaluate stresses and temperatures on aircraft engine components, in order to asses fatigue damage accumulation and residual life. Due to the amount of computational time required it is not possible for a full finite element model to operate in real time using the on-board CPU. Stresses and temperatures are then evaluated by using simplified algorithms. In the present work Guyan reduction and component mode synthesis have been applied to a thermal finite element model, including the cooling stream flow — the so called advection network — in order to reduce the size of the solving equation system. The appropriate mathematical formulation for the advection network reduction has been developed. Two reduction methods have been performed, discussed and subsequently applied to a thermal finite element model of a real low pressure turbine disk. The reduced system includes both the disk and the correlated fluid network model, simulating turbine secondary air system. The finite element model is axi-symmetric, with constant convective coefficients. Results of time integration for the reduced and the complete models have been compared. Results show that the proposed techniques gives models with a reduced number of degrees of freedom and at the same time good accuracy in temperature calculation. The reduced models are then suitable for real time computation.

An H-Adaptive Finite Element Model for Constructing 3-D Wind Fields

2004 ◽  
Author(s):  
Xiuling Wang ◽  
Darrell W. Pepper ◽  
Yitung Chen ◽  
Hsuan-Tsung Hsieh
Keyword(s):  
Finite Element ◽  
Finite Element Model ◽  
Wind Field ◽  
Compressible Flows ◽  
Element Model ◽  
Computational Time ◽  
Individual Element ◽  
Adaptive Finite Element ◽  
Wind Fields ◽  
Irregular Terrain

Calculating wind velocities accurately and efficiently is the key to successfully assessing wind fields over irregular terrain. In the finite element method, decreasing individual element size (increasing the mesh density) and increasing shape function interpolation order are known to improve accuracy. However, computational speed is typically impaired, along with tremendous increases in computational storage. This problem becomes acutely obvious when dealing with atmospheric flows. An h-adaptation scheme, which allows one to start with a coarse mesh that ultimately refines in high gradients regions, can obtain high accuracy at reduced computational time and storage. H-adaptation schemes have been shown to be very effective in compressible flows for capturing shocks [1], but have found limited use in atmospheric wind field simulations [2]. In this paper, an h-adaptive finite element model has been developed that refines and unrefines element regions based upon velocity gradients. An objective analysis technique is applied to generate a mass consistent 3-D flow field utilizing sparse meteorological data. Results obtained from the PSU/NCAR MM5 atmospheric model are used to establish the initial velocity field in lieu of available meteorological tower data. Wind field estimations for the northwest area of Nevada are currently being examined as potential locations for wind turbines.

Multilevel Image Thresholding Based on Improved Expectation Maximization (EM) and Differential Evolution Algorithm

2021 ◽  
pp. 2150013
Author(s):  
Ehsan Ehsaeyan ◽  
Alireza Zolghadrasli
Keyword(s):  
Em Algorithm ◽  
Differential Evolution ◽  
Expectation Maximization ◽  
Differential Evolution Algorithm ◽  
Gaussian Mixture ◽  
Standard Test ◽  
Computational Time ◽  
Image Thresholding ◽  
Evolution Algorithm ◽  
Better Than

Multilevel image thresholding is an essential step in the image segmentation process. Expectation Maximization (EM) is a powerful technique to find thresholds but is sensitive to the initial points. Differential Evolution (DE) is a robust metaheuristic algorithm that can find thresholds rapidly. However, it may be trapped in the local optimums and premature convergence occurs. In this paper, we incorporate EM algorithm to DE and introduce a novel algorithm called EM+DE which overcomes these shortages and can segment images better than EM and DE algorithms. In the proposed method, EM estimates Gaussian Mixture Model (GMM) coefficients of the histogram and DE tries to provide good volunteer solutions to EM algorithm when EM converges in local areas. Finally, DE fits GMM parameters based on Root Mean Square Error (RMSE) to reach the fittest curve. Ten standard test images and six famous metaheuristic algorithms are considered and result on global fitness. PSNR, SSIM, FSIM criteria and the computational time are given. The experimental results prove that the proposed algorithm outperforms the EM and DE as well as EM+ other natural-inspired algorithms in terms of segmentation criteria.

Displacement Back Analysis Based on Return Mapping Algorithm and Differential Evolution Algorithm

2011 ◽  
Vol 301-303 ◽  
pp. 564-568
Author(s):  
Jun Xiang Wang ◽  
An Nan Jiang
Keyword(s):  
Finite Element ◽  
Differential Evolution ◽  
Differential Evolution Algorithm ◽  
Implicit Integration ◽  
Integration Algorithm ◽  
Mapping Algorithm ◽  
Return Mapping ◽  
Return Mapping Algorithm ◽  
Calculation Step ◽  
Evolution Algorithm

Differential evolution algorithm is a new global optimization algorithm. DE does not require an initial value, and it has rapid convergence, strong adaptability to a nonlinear function, the features of parallelcalculation, especially in adoption to the complex problem of multivariable optimization. The constitutive integration algorithm affecting the incremental calculation step, and convergence and accuracy of the results is a key of finite element analysis. It is usually divided into an explicit and implicit integration. Return mapping algorithm is an implicit integration to avoid solving the equivalent plastic strain directly so that we achieve a fast and accurate solution for the constitutive equations. Making use of DE and return mapping algorithm to program, the elasticplastic finite element simulation and parameter inversion, the inversion and simulation results are verificated, the results show that it is closed to the actual situation, indicating usefulness and correctness of the program.

Sign in / Sign up

Export Citation Format

Share Document