Design and Double-Stage Optimization of Synchronous Reluctance Motor for Electric Vehicles

Abstract In this study, a high-power synchronous reluctance motor (SynRM) was designed for the traction motor of electric vehicle (EV) and its double-stage optimization was performed. Genetic algorithm and sensitivity analysis methods were used to obtain the best design parameters. Double-stage optimization was carried out to minimize the torque ripple and obtain the targeted torque, speed, and power values of the SynRM. In the first stage, the genetic algorithm method was used to improve the design parameters of the stator and rotor. With the improved design parameters, it was observed that the torque ripple decreased. In the second stage, the sensitivity analysis method was used. In this method, the effect of changing the skew angle of the stator on the torque ripple was investigated. The performance of the designed motor was examined in the optimization process. It was observed that the targeted torque, power, speed, efficiency, and torque ripple minimization values are successfully achieved with the best stator and rotor parameters. The results showed that SynRM produces high torque and high power with high efficiency and low torque ripple over wide speed range. It is quite proper to use the designed SynRM as a traction motor of new generation electric vehicles.


INTRODUCTION
Nowadays, electric vehicles and their traction motors have become one of the most popular topics.It is also the subject of various academic studies.In addition, there is increasing interest in environmental-friendly vehicles such as electric vehicles to strengthen environmental regulations due to excessive greenhouse gas emissions.One of the most important parts of EVs is electric motors.The balance between the expected performance of the traction system and the cost is essential in selecting the electric motor to be used in EV.
Besides four basic parameters such as power density, reliability, efficiency, and cost, many secondary parameters such as fault tolerance, longevity, thermal limit, and torque ripple are also important in traction motor selection [1].Various motor types such as DC motor, permanent magnet synchronous motor (PMSM), brushless DC motor (BLDC), switched reluctance motor (SRM), and the induction motor (IM) are used as traction motors in EV [2].It is desired that the traction motor of the EV has fast torque response, high power density, high efficiency, wide speed range, high reliability, and low-cost specifications [3].The characteristics of DC motor, PMSM, and BLDC used in EV are presented in a review study.DC motors have the advantage of the ease of speed control but also have disadvantages such as low reliability, large mass, and low-efficiency due to the brush and commutator.PMSM and BLDC have permanent magnets.PMSM has high efficiency, high power factor, and low torque ripple, while BLDC has more noise due to electrical commutation [4].SRM has a simple structure.The rotor of the SRM has no windings, shorting bars, or magnets.So, it is a simple, robust, low-cost motor.SRM has advantages such as high efficiency and simple control.However, there are also disadvantages such as torque ripple, noise, and vibration [5].
IM is widely used in EVs.IM has low-cost, simple, and robust structure, high reliability, low torque ripple, and is maintenance-free.However, these motors' efficiency and power density are low compared to other motors [6].
Motors based on high-performance permanent magnets (PMSM, BLDC), which are mostly produced with rare earth materials, are used as traction motors in EVs.The disadvantages of magnet motors are the high/variable cost of raw materials for magnets, their limited availability in nature, and the inaccessibility of magnet reserves for all countries of the world [7].Also, some conditions limit the performance of motors using magnets.When the temperature of a magnet motor exceeds the operating conditions, the performance of the motor is adversely affected due to the demagnetization effect.
Therefore, there is an increasing interest in electrical machines that do not contain magnets or contain a reduced number of magnets [8].Synchronous Reluctance Motor (SynRM) is an alternative motor that can be used as a traction motor for EVs.The stator of SynRM is similar or identical to IM or PMSM.The rotor of SynRM has a simple and robust structure.Thanks to the anisotropy structure of the rotor, torque is produced by the reluctance force [9], so its rotor does not have any windings or magnets.SynRM has advantages such as structural simplicity, lowcost, high torque, long life, field weakening capability, and disadvantages such as high torque ripple and low power factor [10].
In recent years, many researchers have carried out studies on SynRM.M. Sanada et al. performed symmetrical and asymmetrical flux barrier design in SynRM's rotor.It has been observed that the asymmetrical flux barrier produces less torque ripple than the symmetrical flux barrier.Torque ripple has been reduced by around 10% [11].N. Bianchi et al. researched different flux barrier geometries to reduce torque harmonics [12].Bomela et al. investigated the effects of design parameters such as stator winding type and rotor skew on torque ripple, average torque, and power [13].E. Armando et al. investigated the rotor skew technique in reducing torque ripple.In their studies, they applied the continuous skew technique instead of step skewing.The continuous skew technique gave better results compared to step skew in torque ripple [14].In another study, salient pole and flux barrier rotor types of SynRM were compared.The effects of skew on salient pole rotor were investigated.It has been observed that the skew method reduces torque ripple [15].P. Lazari et al. investigated the skew effect on the rotor of a permanent magnetassisted synchronous reluctance motor (PMaSynRM).3-D and 2-D analyses were performed, and the results were compared.The skew technique reduced torque ripple from 19.5 to 7.8% over 21 Nm average torque [16].Y. Deshpande et al. investigated the effects of the skew technique on the stator of outer rotor PMaSynRM [17].The skew technique has been applied to the stator of SynRM, but the designed motor has high torque ripple and low efficiency.After applying the skew technique, the efficiency of the motor is 76.04% and the torque ripple is 30.18%[18].S. Taghavi et al. designed SynRM for EV.In the study, the geometric parameters of SynRM were optimized.Analyses were carried out using finite element method (FEM) and a 45 kW 48 pole SynRM was designed.The designed SynRM was seen to produce a peak 300 Nm torque up to 5000 rpm [19].A. Credo et al. proposed a new SynRM design for the EV.Different topologies including symmetrical and asymmetrical rotor shapes were designed.Then these topologies were optimized and compared.The bridge structures in the rotor of SynRM have been removed and the necessary strength has been provided by filling liquid epoxy resin inside the flux barriers [20].A. Bozkurt et al. designed an outer rotor permanent magnet assisted synchronous reluctance motor (PMaSynRM) for light EV.In the study, flux barriers were optimized using sensitivity analysis.The designed 1 kW PMaSynRM produced 12.87 Nm of torque at an operating speed of 750 rpm.Torque ripple was calculated as 6.96% and efficiency as 91.30% [21].The motor torque characteristics of some commercialized new generation electric vehicles are given as follows.A brand in China produces 315 Nm peak torque, GM bolt produces 360Nm peak torque, Germany Bosch produces 310 Nm peak torque, and Tesla Model 3 produces 416 Nm peak torque [4].When some studies focusing on the optimization techniques are examined, it is seen that Tae-hee et al. performed multi-stage optimization to increase the average torque and reduce torque ripple.Genetic algorithm and ON-OFF method have been used in their study [22].Min-su et al. carried out a two-stage optimization involving random forest and genetic algorithm in their study.These algorithms, realized respectively, allowed them to increase the average torque by 12% and reduce torque ripple by 36% [23].Abakar et al. used Multi-Objective Genetic Algorithm to optimize the design parameters of a 5.5 kW SynRM.Using the optimized parameters, the efficiency has been increased from 91.2 to 91.8% and torque ripple has been reduced from 43 to 10% [24].Praveenkumar et al. proposed a hybrid optimization technique combining Cauchy Particle Swarm Optimization and Moth Flame Optimization to find the optimum design parameters for an electric motor.In their study, the efficiency of the motor has been increased by about 2% thanks to the hybrid optimization technique [25].
The proposed study aims to design an effective SynRM that meets the traction motor requirements of new generation electric vehicles.For this aim, a 3-phase, 4-pole, 36 stator slot SynRM has been designed by using ANSYS Maxwell electric motor design software.A double-stage optimization technique is proposed instead of the most common single optimization technique in the study.Thus, a more effective and more appropriate optimization process is aimed.A double-stage optimization process is more advantageous if it does not exceed time and processing capacity constraints.Because by utilizing the superior features of each selected algorithm, faster, more comprehensive, and more accurate results can be obtained.A double-stage optimization consisting of a genetic algorithm and sensitivity analysis was carried out.First, the stator and rotor parameters were optimized using a genetic algorithm to obtain targeted torque, power, efficiency, and ripple values.Then, the effects of the skewed stator technique on torque ripple were observed using the sensitivity analysis method.Analysis was achieved using the FEM.It has been observed that high efficiency and low torque ripple are obtained when compared with the work using skewed stator in the literature.The proposed SynRM offers high torque, high power, and high efficiency over a wide speed range with low torque ripple.The results showed that the designed SynRM is valuable, applicable, and effective for the traction motor of the new-generation EV.
The main contribution of the study; using double-stage optimization to obtain the best stator and rotor parameters, using the stator skew technique to reduce torque ripple, and improving the efficiency with high power and high torque at a wide speed range of the motor.

TARGETED FEATURES OF SYNRM
SynRM generates torque using the concept of reluctance and rotating magnetic field.The connection between the rotor and the stator is the magnetic field passing through the air gap.While the magnetic field of the stator rotates at synchronous speed, the rotor will be forced by the magnetic field to rotate at this synchronous speed.Thus, the rotor rotates at synchronous speed after being aligned with the magnetic field of the stator.The rotor of SynRM has no magnets or windings.The equivalent circuit of SynRM is shown in Figure 1.SynRM equations are given Eqs. (1)-(3).
where V s is the stator voltage and i s is the stator current.R s , x r , L s1 , e m , k, i cs , i m are winding resistance, angular velocity of the rotor, total winding leakage inductance, air gap electromotive voltage, air gap leakage flux, stator core current, and magnetizing current, respectively.R cmÀs and R cmÀr are stator core resistance, rotor core resistance, respectively.
The equation where L s1 and L m are summed is given in Eq. (4).
Equations ( 5) and ( 6) are derived when Eq. ( 1) is arranged with respect to the d-q axis reference, and the flux is given in Eq. ( 7).
where V d and V q are dq-axis voltages of the stator.i d , i q , i md , and i mq are the dq-axis currents and dq-axis magnetizing currents, respectively.k d , k q , L d , and L q are the air gap dq-axis linkage fluxes and the dq-axis inductances, respectively.In this motor, torque is produced by the force of reluctance [26].The torque expression (T e ) for SynRM is given Eq. ( 8).
The torque expression shown in Eq. ( 8) depends on the number of poles (p), the d-q axis inductances (L d , L q ), and air gap leakage flux (k).T e is related to the difference of the dq-axis inductances.The average torque will increase as the difference of dq-axis inductances (L d À L q ) increases but torque ripple will also increase.The axis inductances, axis currents, and saliency ratio are very important in improving the torque and torque ripple of the motor.Saliency ratio (S r Þ and torque ripple (T rip ) expressions are given in Eqs. ( 9) and (10) [27].
The study aims to design synchronous reluctance motors for new generation electric vehicles.In design, focused on improving average torque, torque ripple, and efficiency (g).
The efficiency expression given in Eq. ( 11) relates to motor losses [28].P out is the output power of the SynRM.P loss is total losses.Total losses are the sum of core losses, winding losses, and other losses.Core losses are the sum of eddy current losses and hysteresis losses.These expressions were analyzed in the ANSYS Maxwell package program using the FE.
The requirements for the targeted motor are shown in Table 1.The peak torque, power, torque ripple, and efficiency shown in Table 1 are defined based on the Tesla Model-S electric car [29].The first step was to determine the stator and rotor geometric parameters to design the targeted motor.The stator and rotor structure of SynRM are shown in Figures 2 and 3.
The parameter values in Table 2 were accepted as the initial values for the targeted SynRM.These parameters were selected according to the similar studies presented in the literature.According to previous studies in the literature, it has been observed that the distributed winding structure produces lower torque ripple than the concentric winding structure [30].In another study, the single-layer and double-layer structures of the stator windings were examined.It has been observed that the double-layer structure supports more slot/pole combinations and has lower losses [31].Due to its advantages, the double-layer distributed winding type is preferred in this study.

DESIGN PARAMETERS OF SYNRM
The scope of this study is limited to electromagnetic field analysis.The effects of design parameters on the thermal or mechanical structure are ignored.When considering the electromagnetic field, the design parameters of the SynRM affect the output torque, torque ripple and efficiency.To obtain the values required by the new generation electric vehicles, the parameter improvement process has been carried out.The process consists of 2 stages.In the first stage, torque ripple and average torque were improved by optimizing stator parameters (Bs0, Bs1, Bs2, Hs0, Hs1, Hs2, Rs) and rotor parameters (W, R, H, Rb, B0, Y0, R0) using genetic algorithm (GA).In the second stage, the skew was added to the stator of SynRM optimized by a genetic algorithm.The effects of skew angle (0-7 ) on torque ripple were investigated using the sensitivity analysis method.Each stage's analyses were carried out using FEM.Because nonlinear effects and the leakage flux of windings need to be considered for SynRM.Nonlinear analysis techniques are used to obtain electromagnetic performances because of nonlinear problems like the BH-curve and the magnetic permeability of nonlinear reluctances.FEM can offer a comprehensive and accurate solution for nonlinear problems.

Design of SynRM Using Initial Parameters
The initial parameters of SynRM are shown in Table 2.
The torque of SynRM using these parameters has been examined.The obtained torque graph is given in Figure 4.In this design, the average torque of the motor is 367 Nm and the torque ripple is 40%.As can be seen from the results, torque ripple is very high because any parameter optimization has not been made yet.To reduce torque ripple, motor parameters should be improved and optimized.
To achieve parameter optimization genetic algorithm and sensitivity analysis methods are operated.The genetic algorithm reduces torque ripple and increases the average   torque value, where sensitivity analysis is used for torque minimization.

Optimization Using Genetic Algorithm
Genetic algorithm is an artificial intelligence algorithm that aims to find a solution close to the best solution using genetic code [32].The solution feature of the genetic algorithm was used to improve the parameters of the SynRM.
The goal is to minimize torque ripple without reducing the output torque.For this, the fitness function of the genetic algorithm is calculated over-torque ripple and average torque.Two fitness functions given in Equation ( 12) and Equation ( 13) are used for optimization.The fitness function 1 was defined as the average torque greater than 380 Nm, and the fitness function 2 as the torque ripple less than 15%.In GA, the number of individuals and children was chosen as 30, and the maximum number of iterations was 300.
Fitness 1 Fitness 2 The flow chart of the genetic algorithm is shown in Figure 5.The algorithm creates the initial population and performs selection, crossover, and mutation operations.It tries to reach the fitness value up to the defined maximum number of iterations.The algorithm chooses the best value obtained during the iteration as a result.In the first step, optimization was made on the stator parameters.In the second step, the rotor parameters were optimized by using the most suitable stator parameters.In the third step, both stator and rotor parameters were optimized again by a genetic algorithm.Thus, a lower torque ripple was achieved than in the previous step.
The stator parameters have been optimized to find a solution to the high torque ripple problem of SynRM.Optimization was carried out using a genetic algorithm.The optimized parameters and their limits are shown in Table 3.The limits of parameters are chosen in such a way that they do not change the number of slots of the stator and are physically possible peak values.The torque graph of SynRM after optimization of the stator is shown in Figure 6.The average torque for this design is 399 Nm and the torque ripple is 35.6%.
With the stator optimization, the average torque increased, and the torque ripple decreased when compared to the initial design.Optimized stator parameters will be used in the next step.
After the stator optimization of SynRM, the optimization process was performed for the rotor.The optimized parameters and their limits are shown in Table 4.The limits of parameters are chosen in such a way that they do not change the number of flux barriers and are physically possible peak values.
The torque graph after optimization of the rotor is shown in Figure 7.The average torque obtained in this step is 401 Nm, and the torque ripple is 21%.The results show that torque and torque ripple are improved with rotor optimization, but the torque ripple is still higher than the  targeted torque ripple.Next, the rotor and stator are optimized again to achieve the targeted torque ripple.At this step, the previously optimized parameters were chosen as the optimization's initial value, and again the stator and rotor parameters were optimized by a genetic algorithm.With this optimization, the torque is increased from 432 Nm, and the torque ripple is reduced to 16.2%.The torque graph in this step is shown in Figure 8.
The new stator parameters, rotor parameters, and results in the optimization process are given in Table 5.
As can be seen from Table 5, torque and torque ripple were improved but the saliency ratio (S r ) decreased after each optimization using a genetic algorithm.As can be seen from ( 8), the torque also depends on the S r and it is desired to be high, but the S r has been optimized and reduced to decrease torque ripple.The disadvantage caused by the reduction of the S r was compensated by the increase of i d and s currents.Figure 9 shows the magnetic flux density distributions in the optimization process using genetic algorithm.It is seen that increasing i d and i q currents increased the magnetic flux density.Still, the torque ripple obtained was not yet at the expected value.To solve this problem skew technique has been applied, called the sensitivity analysis method.Furthermore, the skew technique has been applied to the stator of the SynRM.

Optimization Using Sensitivity Analysis
Sensitivity analysis is an analysis method in which the output expression is observed by changing the input parameters within certain limits.This method has many options for input parameters.Many parameter optimization studies of electric motors focused on the skew technique in the literature.Some of these studies made a skewed rotor design [14][15][16] and some made a skewed stator design [17,18].In this study, unlike other studies, genetic algorithm is applied in the first stage and stator skewing technique is used in the second stage to minimize torque ripple using sensitivity analysis.Thus, it is aimed to obtain better results by using a double optimization technique.The stator skew angle parameter has been changed from 0 to 7 in one-degree increments.
Figure 10 illustrates the skewing concept of SynRM's stator.Where, h Skew is skew angle, B s is the slot width and H s is slot height of SynRM's stator [18].In this study, the torque and torque ripple were investigated according to the change in the stator skew angle, keeping all input parameters constant except the skew angle.
The results of the sensitivity analyses are shown in Table 6.The highest torque ripple value is reached when the skew angle is 1 , and the lowest torque ripple value is reached when it is 5 Figure 11 shows torque and torque ripple according to skew angle values.The targeted torque value was achieved at all angles, where 2, 5, 6, and 7 degrees skew angles met the targeted torque ripple.
According to the sensitivity analysis results, while the skew angle was 1 , the torque was 547 Nm, and the torque ripple was 24.3%.As seen in Figure 12, this value was observed as the highest torque ripple in the analysis.The torque graph, while the skew angle is 5 , is shown in Figure 13.The lowest torque ripple was seen when the skew angle was 5. Torque at this skew angle is 426 Nm and torque ripple is 9.1%.According to these results, the stator skew angle was chosen as 5 in the final design and analyses of the SynRM.
Figure 14 shows the magnetic flux densities of the two models that produce the highest torque ripple Figure 14(a) and the lowest torque ripple Figure 14(b).When Table 6 is examined, it is seen that the i q current of the Figure 14(a) is higher.Although an increase in i q current causes an increase in torque, it has a major disadvantage that it increases torque ripple.Figure 14(b), which has a relatively lower i q current, has both torque and torque ripple at an acceptable level.

DESIGN OF SYNRM WITH OPTIMIZED PARAMETERS
The average torque value was improved, and torque ripple was minimized by the parameters optimization presented in section 3. When the obtained results are examined, it is seen that the targeted torque and torque ripple have been reached.The final design parameters of SynRM are given in Table 7.
The 3D view of the stator and rotor of the designed SynRM with optimized parameters are shown in Figures 15(a) and 15(b), respectively.The stator of SynRM has a 3-phase, 4pole, 36 stator slots, and double-layer distributed winding structure where the rotor of SynRM has 4-pole, 4-flux barriers.The barrier structure is hyperbolic polyline.The steel type of both the stator and the rotor is M19-24G and the stacking factor is 0.95.The B-H curve of this steel type is shown in Figure 16.The skew technique is used in the stator design.The skew angle is defined as 5 .Torque ripple has been reduced to 9.1% at this skew angle.Figure 17 shows the cross-section view of the stator.The stator skew angle is 0 in Figure 17(a), while the skew angle is 5 in Figure 17(b).

PERFORMANCE OF THE DESIGNED SYNRM
SynRM has advantages such as efficiency, simplicity, low cost, and low moment of inertia.However, it has the disadvantage of a high torque ripple [33].Nevertheless, torque ripple is significantly reduced concerning optimization of stator and rotor parameters and stator skew technique.SynRM was designed with the optimized parameters, and performance analyses have been implemented under different operating conditions to show validity and capacity of the proposed method.Analyses results of the designed SynRM are given as follows.
5.1.Performance analyses of the designed SynRM at 0-10,000 RPM The performance of the designed SynRM was simulated using the Ansys Maxwell/Machine Toolkit plugin.SynRM was simulated with the parameters given in Table 8.
With the help of simulation results, torque-speed-power graphics, efficiency map, and total loss map were obtained.
Output torque and power versus speed graphics are given in Figure 18.
As can be seen in Figure 18, the designed SynRM can produce a constant 470 Nm torque up to 5500 rpm and 141-470 Nm torque out of the constant torque region in which motor speed differs from 6000 to 10,000 rpm.The peak power produced by the SynRM is 285 kW at 6000 rpm.Compared with the motors designed for electric vehicles presented in the literature [19][20][21], it can be said that the designed SynRM produces high torque and high power at a wide speed range successfully.The efficiency map shows the efficiency of the electric motor along the axes of torque and speed.It is desirable for an electric motor to provide maximum efficiency in any torque-speed combination, but the losses of the motor can vary according to the torque-speed combinations, so the efficiency of the motor can also vary. Figure 19 shows the efficiency map of the designed SynRM.
The peak efficiency of SynRM is 97.3%.And the average efficiency of the SynRM is approximately 96%.As seen from the map, efficiency is considerably high.The factor that ensures high efficiency is the improvement of the design parameters of the SynRM by using double-stage optimization techniques.
SynRM's total losses map is shown in Figure 20.The total losses are the sum of the winding, core, and mechanical losses.Most of the losses are due to winding losses.As seen in Figure 19, it is seen that the motor has more loss in the high torque region.This is because the torque produced is dependent on the current drawn, so an increase in the current drawn increases the winding losses.SynRM reaches a peak power of 285 kW at 6000 rpm.At this point, SynRM's winding losses are 8.78 kW, mechanical losses are 0.18 kW, core losses are 1.82 kW, and total losses are 10.8 kW.

Performance Analyses of the Designed SynRM at 3000 RPM Constant Speed
The designed SynRM can provide a constant 470 Nm of torque up to 5500 rpm and generate sufficient torques out of the constant torque region.The performance of the designed SynRM was analyzed for the 3000 rpm constant speed, and this point is given in Figure 21.At this point, the SynRM produces 133 kW power and 426 Nm torque.
The main electrical and mechanical characteristics of the designed SynRM at 3000 rpm operating conditions are shown in Table 9.It is seen that the designed SynRM has low torque ripple, high output torque, and high efficiency    at 3000 rpm.The graphs of torque and current produced by SynRM at 3000 rpm are shown in Figures 22 and 23, respectively.The torque in Figure 22 has a ripple of approximately 38 Nm.This is about 9.1% of the average torque.The obtained torque ripple is reasonably lower than the target torque ripple defined as <15%.The Phase current of the motor is 344 A, as seen in Figure 22.The RMS value of this current is 243.2 A. It is clearly seen that the current is quite lower than the target current defined as   The analysis results clearly show that the designed motor meets the targeted torque, power, current, torque ripple, and efficiency values.The motor torque characteristics of some commercialized traction motors used in new-generation electric vehicles are given as follows.A brand in China produces 315 Nm peak torque, GM bolt produces 360 Nm peak torque, Germany Bosch produces 310 Nm peak torque, and Tesla Model 3 produces 416 Nm peak torque [27].The output torque is obtained as 470 Nm in the proposed study.The designed SynRM meets the torque value of traction motors used in commercialized electric vehicles successfully.When the studies on the optimization of SynRM are examined, the superiority of the proposed method is seen.In a study, it has been tried to obtain the optimum design parameter of SynRM with the help of   optimization.The average torque of the SynRM has been improved by 12%.Torque ripple has also been reduced by 78.7% [34].In another study, particle swarm optimization technique has been used to reduce the torque ripple of SynRM.The optimization reduced torque ripple by 71.5% [35].In this study, the average torque of the SynRM increased by 16% by using the proposed method.Also, the torque ripple is reduced by 77.2%.Thus, the torque ripple is reduced while the average torque is successfully increased.

CONCLUSION
This paper presents the design and double-stage optimization of the SynRM used as a traction motor of a new-generation EV.SynRM, which does not contain magnets, is thought to be an alternative to the traction motor of electric vehicles using magnets, thanks to its low-cost, non-demagnetization, high torque, and high power produced with high efficiency in a wide speed range.For this aim, high power, high torque, and high-speed SynRM have been designed by using ANSYS Maxwell electric motor design software.For SynRM to reach the targeted output values, a doublestage optimization consisting of a genetic algorithm and sensitivity analysis was carried out.In the first stage, the stator and rotor parameters are optimized using a genetic algorithm to achieve the targeted torque, power, efficiency, and torque ripple.It has been observed that the torque of the SynRM designed using the new design parameters increased by 17.7% and the torque ripple decreased by 59.5%.In the second stage, the stator skew technique is performed to minimize the torque ripple of the SynRM by using sensitivity analysis.At this stage, the torque ripple decreased by 43.8% compared to the previous stage.Analysis was achieved using the FEM.Test results show that the optimizations have greatly improved the performance values of SynRM.Compared to the other traction motors of EVs, it has been observed that SynRM produces high torque and meets the required torque.SynRM produces high torque and high power with high efficiency over a wide speed range.The torque ripple was obtained as 9.1% and the peak efficiency is 97.3%.SynRM produces 470 Nm torque and provides maximum of 285 kW power.The designed motor can be operated at constant torque between 0 and 5500 rpm.Furthermore, it can be operated at a relatively lower torque values up to 10,000 rpm rotational speed.The analysis results proved that the designed motor successfully meets the targeted torque, power, current, torque ripple, and efficiency values.The results revealed that the designed SynRM is effective, successful, and applicable for traction motor of the new generation EVs.

FIGURE 4 .
FIGURE 4. Torque graph when using the initial parameters.

FIGURE 5 .
FIGURE 5. Flow chart of genetic algorithm.

FIGURE 9 .
FIGURE 9. Magnetic flux density distribution in the optimization process using genetic algorithm.

FIGURE 11 .
FIGURE 11.Torque and torque ripple graph by changing the skew angle.

FIGURE 14 .
FIGURE 14. Magnetic flux density distribution according to skew angle 1 and 5 .

TABLE 1 .
Motor requirements for the target application.

TABLE 2 .
Initial parameters of the target application.

TABLE 3 .
Stator parameter limit values and new values after optimization.

TABLE 4 .
Rotor parameter limit values and new values after optimization.FIGURE 6. Torque graph after optimization of the stator.FIGURE 7. Torque graph after optimization of rotor.FIGURE 8. Torque graph using improved parameters after optimization.Bekiroglu and Esmer: Design and Double-Stage Optimization of Synchronous Reluctance Motor for Electric Vehicles 2563

TABLE 5 .
Parameter change in the optimization process.

TABLE 6 .
Torque and torque ripple values according to skew angle.

TABLE 7 .
Parameters optimized for the target application.Bekiroglu and Esmer: Design and Double-Stage Optimization of Synchronous Reluctance Motor for Electric Vehicles 2567 <500 A. Motor efficiency is 94.1, and it is greater than the target efficiency defined as >90%.Output torque is calculated as 426 Nm, and it is superior to the target output torque defined as >380Nm.

TABLE 8 .
Parameters used in machine toolkit.Efficiency map of SynRM.Bekiroglu and Esmer: Design and Double-Stage Optimization of Synchronous Reluctance Motor for Electric Vehicles 2569 FIGURE 18. Torquespeedpower graph of SynRM.FIGURE 19.