Robust shape optimization of electric devices based on deterministic optimization methods and finiteelement analysis with affine parametrization and design elements
 316 Downloads
Abstract
In this paper, gradientbased optimization methods are combined with finiteelement modeling for improving electric devices. Geometric design parameters are considered by piecewise affine parametrizations of the geometry or by the design element approach, both of which avoid remeshing. Furthermore, it is shown how to robustify the optimization procedure, that is, how to deal with uncertainties on the design parameters. The overall procedure is illustrated by an academic example and by the example of a permanentmagnet synchronous machine. The examples show the advantages of deterministic optimization compared to standard and popular stochastic optimization procedures such as particle swarm optimization.
Keywords
Finiteelement analysis Genetic algorithms Gradient methods Electric machines Optimization methods Particle swarm optimization Permanentmagnet machines Quadratic programming1 Introduction
In almost all electric design procedures, numerical optimization is employed as one of the last design steps in order to optimize the device’s performance and efficiency, to minimize its weight and size and to save on material and manufacturing costs. Often, the quality of this optimization step indirectly determines the success of the product and, hence, the market position of the company. The reliability, accuracy and computational cost of the numerical optimization procedure becomes in itself a subject of competition. This paper illustrates that shape optimization can be improved substantially when finiteelement (FE) analysis procedures are equipped with piecewise affine parametrization or design elements, such that wellperforming deterministic optimization methods become applicable.
Impressive technical improvements have been achieved by numerical optimization on the basis of magnetic equivalent circuits or 2D and 3D FE models. All have led to highly optimized designs, for example, for permanentmagnet synchronous machines (PMSMs) in automotive applications. Since three decades, FEbased optimization has been addressed in several text books (see, e.g., [12]) and hundreds of journal articles (see, e.g., [14] and the references therein). Although originally, gradientbased methods were preferred (see, e.g., [48, 56, 58]), already for more than two decades, stochastic algorithms are more popular (see, e.g., [19, 33]). The majority of the proposed procedures opt for stochastic or populationbased optimization methods, such as genetic algorithms and particle swarm optimization (see, e.g., [34]), because they allow to use FE solvers as a black box, they can easily consider geometric parameters, their parallelization is straightforward and they are more likely to find the global optimum. Stochastic algorithms have been used for robust optimization, have been applied together with surrogate modeling and have been extended to multiobjective optimization problems [3, 23]. In particular for PMSMs, optimization with stochastic methods became the method of choice [2, 9, 51].
The trend toward stochastic optimization combined with FE analysis continues without restraint, as is illustrated by the number of according contributions at recent conferences. This paper partially counteracts this tendency by turning back to deterministic optimization algorithms. Deterministic optimization methods are known to converge faster than stochastic optimization methods, albeit possibly to a local optimum. Moreover, the analysis of gradientbased methods is more mature, allowing for a rigorous control of mesh discretization errors, for instance. The main drawback of many deterministic methods is, however, the necessity to provide derivatives, which is particularly cumbersome when optimization according to geometric parameters is pursued. This drawback is here addressed explicitly and is alleviated by piecewise affine parametrizations of the geometry or by the design element approach. The overall deterministic optimization routine is shown to outperform the most popular stochastic algorithms by factors. Moreover, the optimization method will be robustified to include uncertainties on the design parameters.
The paper is structured as follows: Sect. 2 recalls the basics of mathematical optimization. It clearly distinguishes between deterministic methods (Sect. 2.3) and particle swarm optimization as a relevant representative of stochastic methods (Sect. 2.4). Furthermore, an extension to robust optimization is discussed in Sect. 2.5. Section 3 deals with FE analysis of magnetodynamic fields. The core parts of the paper are Sect. 3.3.1 about affine parametrization and Sect. 3.3.2 about design elements, both facilitating and improving the calculation of derivatives with respect to geometric parameters. The superior performance of gradienttype deterministic optimization is illustrated for a benchmark example in Sect. 4 and for a PMSM in Sect. 5. Conclusions are formulated in Sect. 6.
2 Constrained optimization
2.1 Constrained optimization problem
The optimization is carried out with respect to I design parameters \(\mathbf{p}=(p_1,p_2,\ldots ,p_I)\) belonging to the admissible set \({\mathcal {P}}_{\mathrm {ad}} =\{\mathbf{p}\in {\mathbb {R}}^IG_m(\mathbf{p})\le 0,m=1,\ldots ,M\}\), where \(G_m(\mathbf{p})\) denote the constraints. The design parameters can be any continuous variables, for example material constants, excitation parameters and geometric sizes or positions. The constraints limit the admissible range of these parameters, for example to preserve the topology of the geometry or to set physical and operational constraints. Discrete design parameters are not considered in this work, although many methods apply, for example as part of a branchandbound technique, to mixedinteger optimization problems as well [21].
2.2 Optimization methods

Problem (1) considers a single optimization goal. For a multiobjective optimization problem, a Pareto front is calculated such that the relative importance of the optimization goals can be fixed later on [7, 12]. This paper does not further consider multiobjective optimization. Nonetheless, the developed techniques are applicable to multiobjective optimization as well.

A distinction is made between global optimization and local optimization, where the first strives to a global optimum, whereas the second may run into a local one. This paper is limited to methods for local optimization. In practice, if a global optimum is required, the methods may be repeated for different start values, or could be embedded as part of a global optimization scheme [24].

Especially when the evaluation of the objective function is computationally expensive, it is recommended to carry out the optimization method on the basis of a surrogate model (indirect optimization methods). Such a simplified model can be obtained by expert knowledge on the application [59], by design space reduction [17], by a response surface methodology [17] or by space mapping [29] or manifold mapping [15]. Here, a direct optimization procedure is used. All ideas presented here can, however, be used in combination with indirect optimization approaches as well [30].

The result from a nominal optimization is a set of optimized design parameters leading to an optimum of the objective function. The optimum may, however, become irrelevant when it is highly sensitive to uncertainties in the design parameters. One speaks about robust optimization when the optimization is carried out taking such uncertainties into account. In this paper, both nominal and robust optimization methods are considered. An approach for robustification is discussed in Sect. 2.5.

Two families of basic optimization methods exist: deterministic and stochastic methods. Among the stochastic methods, genetic algorithms [35], differential evolution [38] and particle swarm optimization (PSO) [27] are well known.
2.3 Gradientbased deterministic method
2.4 Particle swarm optimization
 1.
Maintain a part of the current velocity;
 2.
Head toward the particle’s best found point (\({\hat{\mathbf{p}}}_q\));
 3.
Head toward the swarm’s best found point (\({\hat{\mathbf{p}}}_\text {swarm}\)).
2.5 Robust optimization
In a nominal optimization procedure, one is looking for the minimum value of an objective function. However, during manufacturing, small deviations can occur on the parameters. As a consequence, the optimal solution may become suboptimal in reality. Robust optimization searches for an optimum that is not too much affected by the expected parameter deviations [41, 60].
3 Finiteelement model
The behavior of the devices under consideration is determined by magnetic field phenomena and is simulated using a FE model.
3.1 Magnetoquasistatic formulation
3.2 Finiteelement discretization
3.3 Geometry parametrization
3.3.1 Affine parametrization
This section discusses the idea of decomposing (a part of) the mesh such that geometrical changes can be represented by piecewise affine maps [30, 47]. The affine maps will be referred to by \(f^\ell _{\mathbf{p},\text {aff}}\).
3.3.2 Design element approach
For many geometry optimization tasks, the domain underlying geometry changes cannot be decomposed in triangles or tetrahedra with straight edges and faces, which excludes the use of affine parametrization. NURBS are a more general way to represent geometries and are widely used in CAD systems. Therefore, it seems natural to use the control points (and weights) of NURBS curves as design parameters [6, 49]. This approach has received considerable attention in recent years as new approaches, incorporating NURBS geometries into FE analysis, have emerged. Isogeometric analysis [25] and the NURBSenhanced FE method [50] are important examples. Here, NURBS are only used for the geometry parametrization. A triangular (tetrahedral) mesh is generated once and deformed using the concept of design elements [6, 26].
3.4 Sensitivities
4 Example 1: Die press mold
As a first example, a die press mold for radially magnetizing a segment of sintered magnetic powder (SMP) is considered [54]. This problem has been proposed as testing electromagnetic analysis methods (TEAM) benchmark problem 25 [53] and has been used in numerous papers for comparing optimization algorithms. The vast majority of these publications apply and compare stochastic optimization methods [32, 52], possibly combined with surrogate models [8], uncertainty quantification [39], multiobjective optimization or a combination of them [31]. Only a few papers (see, e.g., [1, 4]) choose deterministic methods, again possibly combined with surrogate models [20], uncertainty quantification [55] or multiobjective optimization. This paper addresses one of the main drawbacks of deterministic methods, which is the consideration of geometric parameters. For this example, the design element approach is used.
Results from the optimization of the die press mold with particle swarm optimization (PSO), trust region (TR) (with MATLAB\(^{\circledR }\)’s fmincon) and an own implementation of sequential quadratic programming (SQP) combined with the design element approach
Method  Minimizer \({\hat{\mathbf{p}}}_{\mathrm{min}}\)  Minimum  Iteration  Function calls  Time  

(in mm)  (in \(\text {T}^2\))  count  \(\mathrm {f}()\)  \(\nabla \mathrm {f}()\)  (in s)  
PSO  \(\begin{pmatrix} 5.1000\\ 16.0000 \\ 16.0000\\ 9.5000\end{pmatrix}\)  1.413498  7  280  N/A  56.63 
SQP (fmincon)  \(\begin{pmatrix} 5.1000\\ 16.0000 \\ 16.0000\\ 9.5000\end{pmatrix}\)  1.413498  4  7  7  31.61 
SQP (own implementation)  \(\begin{pmatrix} 5.1000\\ 16.0000 \\ 16.0000\\ 9.4999\end{pmatrix}\)  1.413498  2  3  2  12.84 
The performance of a standard algorithm for particle swarm optimization (PSO), of the sequential quadratic programming (SQP) method implemented in MATLAB\(^{\circledR }\)’s fmincon function [43] and of an own implementation of SQP is compared in Table 1. Both SQP implementations use the analytical gradients, the BFGS formula for updating the Hessian and a sufficient decrease condition in a merit function. For the PSO, a set of 40 particles is considered and the implementation is multithreaded, while the gradientbased methods are singlethread implementations. The termination criterion for the PSO algorithm is the number of stall iterations, which was set to 5. The PSO actually finds the optimum after 2 iterations. This is because the optimum is at a vertex of the boxshaped domain and all the particles leaving the admissible region are projected onto the boundary. All three methods converge to the same optimum. The deterministic algorithms are substantially faster than PSO, even though PSO exploits parallelization. On the same machine, an evaluation of the objective function \(J(\mathbf{p})\) is performed in 1.65 s, an analytical evaluation of the gradient \(\nabla J(\mathbf{p})\) in 4.69 s and a numerical evaluation of the gradient \(\nabla _{\text {num}} J(\mathbf{p})\) using a forward difference quotient in 7.48 s. All tests were done on a 64 GB RAM Intel\(^{\circledR }\) Xeon\(^{\circledR }\) E52630 v4 machine.
5 Example 2: Permanentmagnet synchronous machine (PMSM)
5.1 Design parameters
5.2 Objective function
5.3 Optimization problem
Numerical results obtained for a \(\varvec{\delta }=0.2~\hbox {mm}\) [5]
\(p_1\)  \(p_2\)  \(p_3\)  \(S_{\mathrm{pm}}\)  \(E_0\)  FE slv  Time  

(mm)  (mm)  (mm)  \(({\hbox {mm}^2})\)  (V)  calls  (s)  
Initial design  19.00  7.00  7.00  133  30.370  –  – 
Genetic algorithm  21.04  2.98  6.56  62.80  30.370  \(\approx 6760\)  520.5 
PSO with penalty term  20.60  3.09  5.91  63.71  30.370  \(\approx 3470\)  267.16 
PSO, own implementation  21.08  2.98  6.63  62.80  30.370  1765  217.52 
SQP, nominal optimization  21.07  2.98  6.61  62.80  30.370  34  2.0 
SQP, robust optimization  20.88  3.73  6.82  77.87  31.086  48  5.9 
5.4 Results
 1.
The first optimization run is carried out with the genetic algorithm implemented in MATLAB\(^{\circledR }\).
 2.The second optimization run is carried out with MATLAB\(^{\circledR }\)’s PSO implementation. To circumvent the restriction to boxshaped parameter domains, the admissible set is enforced by a penalty turn. The new objective function readswhere \(f(t)=e^{(4t^{0.1})}1\) was chosen heuristically such that \(J_\text {pen}\) grows exponentially if one of the constraints is violated. The function \(J_\text {pen}\) was called 4740 times, but was organized as to only evaluate the nonlinear constraint if all other constraints were satisfied. The number of particles was set to 30, the maximum number of stall iterations to \(N_\text {stall}=15\) and the function change tolerance to \(10^{6}\). The PSO characteristic constants are chosen to be \(\omega _0=0.5\) and \(\omega _1=\omega _2=1.49\). The algorithm took 157 iterations before termination.$$\begin{aligned} J_\text {pen}(\mathbf{p})&=J(\mathbf{p}) +2J(\mathbf{p}) \big ( f\left( \max (p_2+p_315,0)\right) \nonumber \\&\quad +f\left( \max (3p_12p_350,0)\right) \nonumber \\&\quad +f\left( \max (g(x),0)\right) \big ) \text {,}\end{aligned}$$(45)
 3.
The third optimization is carried out with an own PSO implementation, for the original objective function \(J(\mathbf{p})\) and applying the nonlinear constraints directly. Here, it is assumed that the admissible set is convex such that points inside the convex hull formed by all previous points do not need to be checked. Fifty particles were used. Termination was enforced after maximally \(N_\text {it,max}=100\) steps or when \(N_\text {stall,max}=15\) stall iterations were observed.
 4.
The fourth run was done with the deterministic method described in Sect. 2.3, relying upon FE simulations equipped with an affine parametrization of the geometry as described in Sect. 3.3.1.
 5.
The fifth run was done with the deterministic method for robust optimization expressed by (11) in Sect. 2.5, again with affine parametrization of the geometry.
The results of all optimization procedures are compared with the values of the initial design (Fig. 5). All routines achieve a substantial decrease in the PM size from \(133~\hbox {mm}^2\) up to about \(63~\hbox {mm}^2\). The price for robustness is a slightly larger size of about \(77~\hbox {mm}^2\). The deterministic methods outperform the stochastic ones by two orders of magnitude. This impressively illustrates the major message of this paper stating that deterministic optimization methods accompanied by FE analysis providing gradients with respect to geometric parameters should be favored over stochastic methods, at least for the here considered class of problems.
6 Conclusion
Affine parametrization and design element approaches are capable of parametrizing the geometry of finiteelement models such that accurate derivatives with respect to geometric parameters become available. This alleviates one of the major drawbacks of gradienttype deterministic optimization methods. For the example of a die mold press, standard sequential programming combined with the design element approach outperforms particle swarm optimization by more than a factor ten. The second example illustrates the applicability of gradienttype robust optimization combined with an affine parametrization of the geometry for a permanentmagnet synchronous machine. Supported by the substantial improvement in computational efficiency, this paper stands up for a revival of deterministic methods for numerical optimization in electrotechnical design procedures.
Notes
Acknowledgements
This work is supported by the German BMBF in the context of the SIMUROM project (grant nr. 05M2013) and the PASIROM project (grant nr. 05M2018), by the ’Excellence Initiative’ of the German Federal and State Governments and by the Centre and Graduate School Computational Engineering at TU Darmstadt.
References
 1.Alotto P, Nervi M (2001) An efficient hybrid algorithm for the optimization of problems with several local minima. Int J Numer Meth Eng 50:847–868CrossRefGoogle Scholar
 2.Bash ML, Pekarek SD (2011) Modeling of salientpole woundrotor synchronous machines for populationbased design. IEEE Trans Energy Convers 26(2):381–392CrossRefGoogle Scholar
 3.Baumgartner U, Magele C, Renhart W (2004) Pareto optimality and particle swarm optimization. IEEE Trans Magn 40(2):1172CrossRefGoogle Scholar
 4.Berkani M, Giurgea S, Espanet C, Coulomb JL, Kieffer C (2013) Study on optimal design based on direct coupling between a FEM simulation model and LBFGSB algorithm. IEEE Trans Magn 49(5):2149–2152CrossRefGoogle Scholar
 5.Bontinck Z, Lass O, Schöps S, De Gersem H, Ulbrich S, Rain O (2018) Robust optimization formulations for the design of an electric machine. IET Sci Meas Technol (accepted). arXiv:1712.01536
 6.Braibant V, Fleury C (1984) Shape optimal design using Bsplines. Comput Methods Appl Mech Eng 44(3):247–267. https://doi.org/10.1016/00457825(84)901324 CrossRefzbMATHGoogle Scholar
 7.Brisset S, Vizireanu D, Brochet P (2008) Design and optimization of a ninephase axialflux PM synchronous generator with concentrated winding for directdrive wind turbine. IEEE Trans Ind Appl 44(3):707–715. https://doi.org/10.1109/TIA.2008.921379 CrossRefGoogle Scholar
 8.Canova A, Gruosso G, Repetto M (2003) Magnetic design optimization and objective function approximation. IEEE Trans Magn 39(5):2154–2162CrossRefGoogle Scholar
 9.Cassimere BN, Sudhoff SD (2009) Populationbased design of surfacemounted permanentmagnet synchronous machines. IEEE Trans Energy Convers 24(2):338–346CrossRefGoogle Scholar
 10.Clemens M (2005) Large systems of equations in a discrete electromagnetism: formulations and numerical algorithms. IEE Proc Sci Meas Technol 152(2):50–72. https://doi.org/10.1049/ipsmt:20050849 CrossRefGoogle Scholar
 11.Delfour M, Zolésio JP (2011) Shapes and geometries: metrics, analysis, differential calculus, and optimization, 2 edn. In: Advances in design and control. SIAM, Philadelphia. https://doi.org/10.1137/1.9780898719826
 12.Di Barba P (2010) Multiobjective shape design in electricity and magnetism, lecture notes in electrical engineering. Springer, BerlinCrossRefGoogle Scholar
 13.Diehl M, Bock HG, Kostina E (2006) An approximation technique for robust nonlinear optimization. Math Program 107(1–2):213–230. https://doi.org/10.1007/s1010700506851 MathSciNetCrossRefzbMATHGoogle Scholar
 14.Duan Y, Ionel DM (2013) A review of recent developments in electrical machine design optimization methods with a permanentmagnet synchronous motor benchmark study. IEEE Trans Ind Appl 49(3):1268–1275. https://doi.org/10.1109/TIA.2013.2252597 CrossRefGoogle Scholar
 15.Echeverría D, Lahaye D, Encica L, Lomonova EA, Hemker PW, Vandenput AJA (2006) Manifoldmapping optimization applied to linear actuator design. IEEE Trans Magn 42(4):1183–1186. https://doi.org/10.1109/TMAG.2006.870969 CrossRefGoogle Scholar
 16.Gangl P, Amstutz S, Langer U (2016) Topology optimization of electric motor using topological derivative for nonlinear magnetostatics. IEEE Trans Magn 52(3):1–4. https://doi.org/10.1109/TMAG.2015.2496172 CrossRefGoogle Scholar
 17.Gillon F, Brochet P (2000) Screening and response surface method applied to the numerical optimization of electromagnetic devices. IEEE Trans Magn 36(4):1163–1167. https://doi.org/10.1109/20.877647 CrossRefGoogle Scholar
 18.Graeb HE (2007) Analog design centering and sizing. Springer, DordrechtGoogle Scholar
 19.Hameyer K, Kasper M (1993) Shape optimization of a fractional horsepower DCmotor by stochastic methods. In: Hernandez S, Brebbia C (eds) Computer aided optimimum design of structures III: optimization of structural systems and applications, pp 15–30Google Scholar
 20.Hemker PW, Echeverría D (2007) A trustregion strategy for manifoldmapping optimization. J Comput Phys 224:464–475MathSciNetCrossRefGoogle Scholar
 21.Hemker T, von Stryk O, De Gersem H, Weiland T (2008) Mixedinteger nonlinear design optimization of a superconductive magnet with surrogate functions. IEEE Trans Magn 44(6):1110–1113CrossRefGoogle Scholar
 22.Hinze M, Kunkel M, Vierling M (2011) POD model order reduction of driftdiffusion equations in electrical networks. In: Benner P, Hinze M, ter Maten EJW (eds) Model reduction for circuit simulation. Lecture notes in electrical engineering, vol 74. Springer, Netherlands, pp 177–192Google Scholar
 23.Ho SL, Yang S, Ni Guangzheng LEWC, Wong HC (2005) A particle swarm optimizationbased method for multiobjective design optimizations. IEEE Trans Magn 41(5):1756–1759. https://doi.org/10.1109/TMAG.2005.846033 CrossRefGoogle Scholar
 24.Horst R, Pardalos PM, Thoai NV (2000) Introduction to global optimization. In: Nonconvex optimization and its applications, vol 48. SpringerGoogle Scholar
 25.Hughes TJR, Cottrell JA, Bazilevs Y (2005) Isogeometric analysis: CAD, finite elements, NURBS, exact geometry and mesh refinement. Comput Methods Appl Mech Eng 194:4135–4195. https://doi.org/10.1016/j.cma.2004.10.008 MathSciNetCrossRefzbMATHGoogle Scholar
 26.Imam MH (1982) Threedimensional shape optimization. Int J Numer Methods Eng 18(5):661–673. https://doi.org/10.1002/nme.1620180504 CrossRefzbMATHGoogle Scholar
 27.Kennedy J, Eberhart R (1995) Particle swarm optimization. IEEE Int Conf Neural Netw 4:1942–1948Google Scholar
 28.Koch S, De Gersem H, Weiland T, Fischer E, Moritz G (2008) Transient 3D finite element simulations of the SIS100 magnet considering anisotropic, nonlinear material models for the ferromagnetic yoke. IEEE Trans Appl Supercond 18(2):1601–1604CrossRefGoogle Scholar
 29.Koziel S, Leifsson L (2013) Surrogatebased modeling and optimization: applications in engineering. Springer, BerlinCrossRefGoogle Scholar
 30.Lass O, Ulbrich S (2017) Model order reduction techniques with a posteriori error control for nonlinear robust optimization governed by partial differential equations. SIAM J Sci Comput 39(5):S112–S139. https://doi.org/10.1137/16M108269X MathSciNetCrossRefzbMATHGoogle Scholar
 31.Lebensztajn L, Rondini Marretto CA, Caldora Costa M, Coulomb JL (2004) Kriging: a useful tool for electromagnetic device optimization. IEEE Trans Magn 40(2):1196–1199CrossRefGoogle Scholar
 32.Lei G, Liu C, Zhu J, Guo Y (2015) Techniques for multilevel design optimization of permanent magnet motors. IEEE Trans Energy Convers 30(4):1574–1584CrossRefGoogle Scholar
 33.Lok CL, Vengadaesvaran B, Ramesh S (2017) Implementation of hybrid pattern searchgenetic algorithm into optimizing axialflux permanent magnet coreless generator (AFPMG). Electr Eng 99:751–761. https://doi.org/10.1007/s0020201604439 CrossRefGoogle Scholar
 34.Ma C, Qu L (2015) Multiobjective optimization of switched reluctance motors based on design of experiments and particle swarm optimization. IEEE Trans Energy Convers 30(3):1144–1153CrossRefGoogle Scholar
 35.Man KF, Tang KS, Kwong S (2012) Genetic algorithms: concepts and designs. Springer, BerlinzbMATHGoogle Scholar
 36.Monk P (2003) Finite element methods for Maxwell’s equations. Oxford University Press, OxfordCrossRefGoogle Scholar
 37.Müller G, Vogt K, Ponick B (1974) Berechnung elektrischer Maschinen, 6th edn. WileyVCH, WeinheimGoogle Scholar
 38.Neri F, Tirronen V (2010) Recent advances in differential evolution: a survey and experimental analysis. Artif Intell Rev 33(1–2):61–106. https://doi.org/10.1007/s1046200991372 CrossRefGoogle Scholar
 39.Nishida N, Takahashi Y, Wakao S (2008) Robust design optimization approach by combination of sensitivity analysis and sigma level estimation. IEEE Trans Magn 44(6):998–1001CrossRefGoogle Scholar
 40.Nocedal J, Wright SJ (2006) Numerical optimization. Springer series in operations research and financial engineering, 2nd edn. Springer, New YorkGoogle Scholar
 41.Omekanda AM (2006) Robust torque and torqueperinertia optimization of a switched reluctance motor using the Taguchi methods. IEEE Trans Ind Appl 42(2):473–478CrossRefGoogle Scholar
 42.Pahner U, Mertens R, De Gersem H, Belmans RJM, Hameyer K (1998) A parametric finite element environment tuned for numerical optimization. IEEE Trans Magn 34(5):2936–2939. https://doi.org/10.1109/20.717685 CrossRefGoogle Scholar
 43.Powell MJD (1978) A fast algorithm for nonlinearly constrained optimization calculations, vol 630. Springer, Berlin, pp 144–157zbMATHGoogle Scholar
 44.Powell MJD (2002) UOBYQA: unconstrained optimization by quadratic approximation. Math Program B92:555–582. https://doi.org/10.1007/s101070100290 MathSciNetCrossRefzbMATHGoogle Scholar
 45.Powell MJD (2009) The BOBYQA algorithm for bound constrained optimization without derivatives. In: Technical report DAMTP 2009/NA06. Cambridge University PressGoogle Scholar
 46.Rahman MA, Zhou P (1991) Determination of saturated parameters of PM motors using loading magnetic fields. IEEE Trans Magn 27(5):3947–3950. https://doi.org/10.1109/20.104967 CrossRefGoogle Scholar
 47.Rozza G, Huynh DBP, Patera AT (2008) Reduced basis approximation and a posteriori error estimation for affinely parametrized elliptic coercive partial differential equations. Arch Comput Methods Eng 15(3):229–275. https://doi.org/10.1007/s1183100890199 MathSciNetCrossRefzbMATHGoogle Scholar
 48.Russenschuck S (1990) Mathematical optimization techniques for the design of permanent magnet synchronous machines based on numerical field calculation. IEEE Trans Magn 26(2):638–641CrossRefGoogle Scholar
 49.Ryu JS, Yao Y, Koh CS, Yun S, Kim DS (2005) Optimal shape design of 3D nonlinear electromagnetic devices using parameterized design sensitivity analysis. IEEE Trans Magn 41(5):1792–1795. https://doi.org/10.1109/TMAG.2005.845982 CrossRefGoogle Scholar
 50.Sevilla R, FernándezMéndez S, Huerta A (2008) NURBSenhanced finite element method (nefem). Ph.D. Thesis, Universitat Politècnica de Catalunya (2008)Google Scholar
 51.Sizov GY, Zhang P, Ionel DM, Demerdash NAO, Rosu M (2013) Automated multiobjective design optimization of PM AC machines using computationally efficient FEA and differential evolution. IEEE Trans Ind Appl 49(5):2086–2096CrossRefGoogle Scholar
 52.Sonoda S, Takahashi Y, Kawagishi K, Nishida N, Wakao S (2007) Application of stepwise multiple regression to design optimization of electric machine. IEEE Trans Magn 43(4):1609–1612CrossRefGoogle Scholar
 53.Takahashi N (1997) Optimization of die press model (TEAM workshop problem 25). Technical report, International Compumag Society. http://www.compumag.org/jsite/images/stories/TEAM/problem25.pdf. Accessed 11 Mar 2018
 54.Takahashi N, Ebihara K, Yoshida K, Nakata T, Ohashi K, Miyata K (1996) Investigation of simulated annealing method and its application to optimal design of die mold for orientation of magnetic powder. IEEE Trans Magn 32(3):1210–1213. https://doi.org/10.1109/20.497461 CrossRefGoogle Scholar
 55.Takahashi RHC, Ramírez JA, Vasconcelos JA, Saldanha RR (2001) Sensitivity analysis for optimization problems solved by stochastic methods. IEEE Trans Magn 37(5):3566–3569. https://doi.org/10.1109/20.952663 CrossRefGoogle Scholar
 56.Takorabet N, Laporte B, Vinsard G (1997) On the optimization of linear induction devices. Electr Eng 80:221–226CrossRefGoogle Scholar
 57.Trältzsch F (2009) Optimale Steuerung Partieller Differentialgleichungen: Theorie. Springer, Verfahren und AnwendungenCrossRefGoogle Scholar
 58.Weeber K, Hoole SRH (1992) Geometric parametrization and constrained optimization techniques in the design of salient pole synchronous machines. IEEE Trans Magn 28(4):1948–1960CrossRefGoogle Scholar
 59.Wrobel R, Lukaniszyn M, Jagiela M, Latawiec K (2003) A new approach to reduction of the cogging torque in a brushless motor by skewing optimization of permanent magnets. Electr Eng 85:59–69. https://doi.org/10.1007/s0020200201444 CrossRefGoogle Scholar
 60.Yoon SB, Jung IS, Hyun DS, Hong JP, Kim YJ (1999) Robust shape optimization of electromechanical devices. IEEE Trans Magn 35(3):1710–1713. https://doi.org/10.1109/20.767356 CrossRefGoogle Scholar
Copyright information
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.