New Modulation Technique to Mitigate Common Mode Voltage Effects in Star-Connected Five-Phase AC Drives

energies Article New Modulation Technique to Mitigate Common Mode Voltage Effects in Star-Connected Five-Phase AC Drives Markel Fernandez 1,∗, Andres Sierra-Gonzalez 2, Endika Robles 1, Iñigo Kortabarria 1, Edorta Ibarra 1and Jose Luis Martin 1 1Derpartment of Electronic Technology, University of the Basque Country (UPV/EHU), Plaza Ingeniero Torres Quevedo 1, 48013 Bilbao, Spain; [email protected] (E.R.); [email protected] (I.K.); [email protected] (E.I.); [email protected] (J.L.M.) 2Tecnalia Research and Innovation, C. Mikeletegi 7, 20009 Donostia, Spain; [email protected] *Correspondence: [email protected] Received: 29 November 2019; Accepted: 29 January 2020; Published: 31 January 2020   Abstract: Star-connected multiphase AC drives are being considered for electromovility applications such as electromechanical actuators (EMA), where high power density and fault tolerance is demanded. As for three-phase systems, common-mode voltage (CMV) is an issue for multiphase drives. CMV leads to shaft voltages between rotor and stator windings, generating bearing currents which accelerate bearing degradation and produce high electromagnetic interferences (EMI). CMV effects can be mitigated by using appropriate modulation techniques. Thus, this work proposes a new Hybrid PWM algorithm that effectively reduces CMV in five-phase AC electric drives, improving their reliability. All the mathematical background required to understand the proposal, i.e., vector transformations, vector sequences and calculation of analytical expressions for duty cycle determination are detailed. Additionally, practical details that simplify the implementation of the proposal in an FPGA are also included. This technique, HAZSL5M5-PWM, extends the linear range of the AZSL5M5-PWM modulation, providing a full linear range. Simulation results obtained in an accurate multiphase EMA model are provided, showing the validity of the proposed modulation approach. Keywords: multiphase electric drives; CMV; modulation techniques; PWM 1. Introduction AC electric drives are used in a wide variety of industrial applications such as in compressors [ 1 ], in electric vehicle propulsion systems [ 2 , 3 ] and in more electric aircraft (MEA) [ 4 ], among others. Although three-phase systems dominate the AC drive market, multiphase solutions are gaining popularity [ 5 – 7 ]. Multiphase systems are preferable for applications where high fault tolerance is required [ 8 ], such as for MEA applications, where electromechanical actuators (EMA) for control surfaces, fuel pumps, landing gears, environmental control systems and starter-generators need to be operated [ 9 – 11 ]. Apart from their intrinsic fault tolerance, other benefits of multiphase drives include a reduced current per phase (reducing copper losses and increasing efficiency) [ 4 ], noise and electromagnetic interference (EMI) minimization [ 12 , 13 ], higher power density and lower torque ripple [ 14 ], making them attractive for transport electrification. Among the multiphase topologies available in the scientific literature, star-connected five-phase technologies (Figure 1) can be highlighted, as they provide a good trade-off between system complexity and fault tolerance [ 15 , 16 ]. Specifically, multiphase permanent magnet synchronous machines (PMSM) are being considered for aircrafts due to their superior power density [17,18]. Energies 2020,13, 607; doi:10.3390/en13030607 www.mdpi.com/journal/energies Energies 2020,13, 607 2 of 19 In general, AC electric drives can experience issues due to the common-mode voltage (CMV) [ 19 ] and common mode currents (CMC) [ 20 ]. CMV variations are generated by the commutation of the power converter devices, producing EMI [ 21 ] and bearing currents that can compromise the integrity of the electric machine [ 22 ]. Such voltage variations create new capacitive paths through the motor bearings, leading to premature aging. Capacitive currents, electrostatic discharge machine (EDM) currents, circulating currents and rotor-to-ground currents can flow through the bearings [ 23 , 24 ] (Figure 2), and their harmful effects depend on the type of bearing, size of the machine and how the machine is used. e d c 0 b a VCM n Vdc Z Figure 1. CMV in a five-phase power system. CMV dv/dt at motor terminals CMV mirrored over the bearings Common Mode Currents Circulating currents Shaft voltage EDM currents Capacitive currents Rotor to ground currents Figure 2. Cause-and-effect chain of common mode voltage (CMV) (adapted from [25]). Operating at higher switching frequencies can entail more severe CMV related issues (additional EMI generation and larger number of dv/dt) [ 26 ]. As considerable efforts are being carried out to widespread the usage of wide bandgap (WBG) devices in AC drives, much higher switching frequencies are expected in the future, making the investigation on CMV mitigation a popular topic [ 13 , 19 , 20 ]. Thus, a wide variety of solutions have been proposed in the literature. Such solutions can be classified as passive or active. Passive solutions are those which mitigate or eliminate the harmful effects generated by CMV, while active solutions are intended to reduce or totally avoid CMV generation. Among passive solutions, Faraday shielding [ 27 , 28 ], ceramic and hybrid bearings [ 22 , 29 ], shielded cables [ 23 , 30 ] and shaft grounding rings [ 27 , 31 ] can be highlighted. On the other hand, modulation techniques and new inverter topologies such as multilevel inverters [ 32 ], single-phase transformerless inverters [ 33 , 34 ] and three-phase inverters [ 35 ] among others [ 36 ] are the most common active solutions. Among all these solutions, modulation algorithms can be considered for CMV reduction in star-connected five-phase AC drives due to their ease of implementation, low cost, and because no additional hardware is needed. Energies 2020,13, 607 3 of 19 In [ 37 ], the authors initially proposed a CMV reduction modulation technique for five-phase inverters, named AZSL5M5-PWM. However, the proposal has been only considered for passive loads (star-connected RL loads) and solely validated in open-loop. From the obtained results, it has been concluded that the linear range of the original AZSL5M5-PWM is limited, which can prevent the utilization of this technique in electric drives where operation close to the base speed (without entering in field weakening region) is desirable, as is the case in most EMA systems. Thus, a hybrid AZSL5M5-PWM technique (HAZSL5M5-PWM) that provides the same linear range as conventional space vector PWM (SV-PWM) is proposed in this work, and its performance is evaluated in an EMA system. This manuscript is organized as follows. First of all, conventional SV-PWM for star-connected five-phase power systems is presented, where the harmonic projection of the stator voltages into their corresponding orthogonal subspaces by means of Clarke transformation is mathematically justified. After that and considering the third harmonic elimination constraint, it is shown how CMV variations are generated in the multiphase drive. Secondly, the most relevant reduced common-mode voltage PWM (RCMV-PWM) modulation techniques are briefly described, focusing on their limitations. After that, the proposed Hybrid AZSL5M5-PWM modulation technique is presented providing the required tools for duty cycle calculation, and validated by means of simulation. The target of the proposed modulation technique is to effectively reduce CMV in star connected multiphase systems, while the hybridization is performed to cover the whole operation range of the drive. Open-loop and detailed five-phase EMA simulations are conducted to perform the validation, where not only CMV reduction is verified, but other figures such as total harmonic distortion (THD) and efficiency are evaluated in order to demonstrate that the achieved CMV reduction does not significantly penalize other relevant drive figures. 2. Influence of the SV-PWM Technique in the CMV of a Star-Connected Five-Phase AC Drive SV-PWM is one of the most used modulation techniques in three-phase and multiphase power systems thanks to its easy digital implementation and optimum DC bus voltage utilization. As a star-connected five-phase system has four degrees of freedom, stator voltages and currents can be represented into two separated two-dimensional planes, α - β and x - y , and one homopolar component by means of the following amplitude invariant Clarke transformation [38]:        vα vβ vx vy v0        =2 5        1cos(2π/5)cos(4π/5)cos(6π/5)cos(8π/5) 0sin(2π/5)sin(4π/5)sin(6π/5)sin(8π/5) 1cos(4π/5)cos(8π/5)cos(12π/5)cos(16π/5) 0sin(4π/5)sin(8π/5)sin(12π/5)sin(16π/5) 1 21 21 21 21 2               va vb vc vd ve        . (1) The Clarke transformation allows us to decouple the 5-dimensional voltage vector in the abcde reference frame into three orthogonal subspaces ( α - β , x - y and 0). For a surface-mounted permanent magnet synchronous machine (SM-PMSM), this decoupling is done through the diagonalization of the inductance matrix L(2). L=       L11 L12 L13 L14 L15 L21 L22 L23 L24 L25 L31 L32 L33 L34 L35 L41 L42 L43 L44 L45 L51 L52 L53 L54 L55        . (2) For SM-PMSMs, the elements of L can be considered invariant with respect to the rotor angular position, as the surface placed magnets have a permeability near that the one of the air. Therefore, a SM-PMSM behaves like a non-salient pole synchronous machine [ 39 ]. As the windings in each phase are manufactured identically, the mutual inductances between any pair of phases separated with the Energies 2020,13, 607 4 of 19 same electrical angle are equal, i.e., L12 =L15 =L21 =L23 =L51 =Ljk (if |j−k|= 1) or L13 =L31 = L25 =L52 =Ljk (if |j−k|= 2). Similarly, all the self-inductances are equal ( L11 =L22 =··· =L55 ). This type of matrix is known as a circulant matrix, and it has some special properties [ 40 ]. For example, it guarantees that L is orthogonally diagonalizable by a transformation represented by a 5 × 5 real matrix [41,42]. The circulant matrices are diagonalized by the Fourier Matrix [ 40 , 43 ]. Therefore, the Clarke transformation decomposes the 5-dimensional vectors according to their harmonic components. In the α - β sub-space, the h= 5 (l− 1 )± 1 harmonic components are projected while, in the x - y sub-space, the h= 5 (l− 1 )± 3 ones are projected, being l∈ { 1, 3, 5, ... } [ 41 , 44 ]. The harmonic components of order h= 5 l are projected into the zero-sequence or homopolar sub-space. In Table 1, the odd harmonics associated with each sub-space according to the Clarke transformation of (1) are presented for a 5-phase machine. Table 1. Sub-space harmonics mapping for a five-phase machine. Sub-Space Harmonics α−βh=1, 9, 11, 19... x−y h =3, 7, 13, 17... zero-sequence h=0, 5,15, 25... The number of possible switching states or space vectors is 2 5 , where 30 are active vectors and two are zero vectors (Figure 3). Active vectors can be classified depending on their magnitude as: • Large vectors, where |Vl|= 4 / 5 VDC cos(π/ 5 ) , which correspond to the outer decagon of Figure 3. •Medium vectors, where |Vm|=2/5VDC, which correspond to the middle decagon of Figure 3. • Small vectors, where |Vs|= 4 / 5 VDC cos( 2 π/ 5 ) , which correspond to the inner decagon of Figure 3. (11100) (01000) (10100) (11000) (11101) (11010) (11001) (10000) (01001) (10001) (11011) (10101) (10010) (00001) (10011) (00011) (10111) (01011) (01101) (11110) (01100) (01110) (00100) (01010) (10110) (01111) (00110) (00101) (00010) (00111) 1 2 3 4 5 6 7 8 9 10 α β (a) y x (10101) (00100) (10001) (10100) (10111) (11100) (10110)(10000) (00110) (10010) (11110) (10011) (11000) (00010) (11010) (01010) (11011) (01110) (00111) (11101) (00101) (01101) (00001) (01100) (11001) (01111) (01001) (00011) (01000) (01011) 1 2 3 4 5 6 7 8 9 10 (b) Figure 3. Five-phase SV-PWM: α - β and x - y planes with their corresponding space vectors and switching states (a ‘1’ in a switching state represents that the top switch of a given phase is activated, while a ‘0’ represents that its complementary switch is activated). (a)α-βvector plane. (b)x-yvector plane. For an n -phase system, n− 1 active vectors must be applied at each commutation period in order to achieve a sinusoidal output [ 45 ]. Thus, four active vectors must be used in a star-connected Energies 2020,13, 607 5 of 19 five-phase system. Although various possible active vector combinations are possible to produce a given output voltage vector, the most common alternative consists on using two large and two medium adjacent vectors. As an illustrative example, Figure 3shows the vectors used to synthesize a given reference voltage vector located in the first sector of the α - β plane, being the following the application sequence that minimizes switching losses: 00000, 10000, 11000, 11001, 11101, 11111, 11101, 11001, 11000, 10000, 00000. If the third harmonic component needs to be eliminated, the application-time ratio between medium and large vectors must satisfy (3), as with this ratio the sum of the applied vectors in the x-yplane is zero (Figure 3) [46]. tlarge tmedium =1.618. (3) As a result, the maximum achievable output voltage following this modulation approach is: Vomax =4 5cos π 5cos π 10=0.6155VDC, (4) and the CMV generated in the five-phase system is: VCM(t) = 1 5[Va0(t) + Vb0(t) + Vc0(t) + Vd0(t) + Ve0(t)]. (5) When using SV-PWM and applying the vector sequence that corresponds to the first sector of the α - β plane, the CMV waveform of Figure 4is obtained. From (5), it can be deduced that all large and short vectors generate CMV levels of ± 0.3 VDC , while CMV levels are of ± 0.1 VDC for medium vectors and of ± 0.5 VDC for null vectors. When evaluating the impact of the CMV, the difference between the maximum and minimum CMV levels ( ∆CMV , Figure 4) must be considered, and the number of CMV variations for each commutation period (NCMV) must also be taken into account. S1 Tsw S3 S5 SV-PWM 00000 11000 10000 11001111011111111101110011100010000 00000 S7 S9 CMV [V] Tsw 0.5Vdc -0.5Vdc 0.3Vdc -0.3Vdc 0.1Vdc -0.1Vdc ΔCMV Tsw Tsw Shaft voltage [V] Bearing current [A] Figure 4. CMV waveform of SV-PWM technique (adapted from [23]). Energies 2020,13, 607 6 of 19 3. RCMV-PWM Techniques As zero vectors are responsible for generating the maximum CMV levels (Figure 4), most of the RCMV-PWM techniques avoid the application of these vectors to reduce ∆CMV and NCMV . In [ 38 ], an extension of the three-phase active zero state PWM (AZS-PWM) [ 47 ] modulation technique to the five-phase scenario is proposed. This technique replaces zero vectors by applying two active vectors with the opposite phase at the same time. In this work, this technique will be named AZSL2M2-PWM as, apart from the active vectors that substitute zero vectors, two large (L2) and two medium (M2) vectors are used at each modulation period. This technique shows a good harmonic performance and DC bus utilization, being its linear range 0 ≤m≤ 1. However, ∆CMV is not greatly reduced, as only ± 0.5 VDC CMV levels are avoided (Table 2). Similarly, a modulation algorithm that employs four large active vectors in conjunction with two active vectors with opposite phases (AZSL4-PWM) is proposed in [ 38 ]. This technique has the same linear range as SV-PWM and AZSL2M2-PWM, and considerably reduces ∆CMV, as only applies large vectors. Nonetheless, NCMV remains as for SV-PWM (Table 2). M5-PWM [ 48 ] and L5-PWM [ 49 ] techniques completely eliminate ∆CMV and NCMV by only using odd or even medium (M5-PWM), or odd or even large (L5-PWM) active vectors. However, this is achieved at the cost of introducing additional power losses, significantly reducing the linear range up to 0.5257 for M5-PWM (Table 2), and generating high harmonic distortion for L5-PWM, making them inappropriate for many industrial applications. Authors in [ 48 , 49 ] also propose variants that use ten medium (M10-PWM) or ten large (L10-PWM) vectors. These techniques enhance the linear range by increasing the available vectors, but do not reduce ∆CMV and NCMV as much as with M5-PWM and L5-PWM (Table 2). Table 2. Summary of the most relevant features of SV-PWM and RCMV-PWM techniques. Modulation ∆CMV [V] NCMV ∆CMV Reduction [%] NCMV Reduction [%] vCM Waveform Linear Range SV-PWM VDC 10 - - 0.5VDC -0.5VDC 0.3VDC 0.1VDC -0.1VDC -0.3VDC 0≤m≤1 AZSL2M2-PWM 0.6VDC 6−40% −40% 0.3VDC 0.1VDC -0.1VDC -0.3VDC 0≤m≤1 AZSL4-PWM 0.2VDC 10 −80% 0% 0.1VDC -0.1VDC 0≤m≤1 M5-PWM 0 0 −100% −100% -0.3VDC 0≤m≤0.5257 M10-PWM 0.6VDC 2−40% −80% 0.3VDC -0.3VDC 0≤m≤0.618 L5-PWM 0 0 −100% −100% 0.1VDC 0≤m≤0.8507 L10-PWM 0.2VDC 6−80% −40% 0.1VDC -0.1VDC 0≤m≤1 AZSL5M5-PWM 0.4VDC 2−60% −80% 0.1VDC -0.3VDC 0≤m≤0.8507 Among the reviewed techniques, AZSL2M2-PWM and L10-PWM best suit for industrial applications, as they keep the linear range with a reasonable THD while effectively reducing ∆CMV . However, NCMV reduction by means of such modulation algorithms is limited. Thus, a new RCMV-PWM technique that further reduces NCMV while keeping an extended linear range is proposed in the following section. Energies 2020,13, 607 7 of 19 4. Proposed RCMV-PWM Technique 4.1. Active Zero State L5M5 PWM Technique (AZSL5M5-PWM) The main part of the proposed RCMV-PWM technique, named active zero state L5M5 PWM (AZSL5M5-PWM), is based on the AZSL2M2-PWM technique [ 38 ]. However, and unlike AZSL2M2-PWM, the proposed scheme only uses odd or even vectors to further reduce the CMV voltage variations. For example, if odd vectors are only considered, this leads to the sector distribution of Figure 5a. Thus, five medium vectors and five large vectors are exclusively used to synthesize the reference voltage (Figure 5), and CMV varies between − 0.3 VDC and 0.1 VDC (if only even vectors are used, CMV varies between − 0.1 VDC and 0.3 VDC ). Consequently, ∆CMV = 0.4 VDC and NCMV =2 (Table 2). (11100) (01000) (11001) (10000) (00001) (10011) (01110) (00100) (00010) (00111) 1 2 5 3 4 0 < VoMAX < 0.5236VDC (a) 1 2 5 3 4 0 < VoMAX < 0.5236VDC (11000) (01100) (00110) (00011) (10001) (11101) (01111) (11110) (10111) (11011) (b) Figure 5. Sector distribution of AZSL5M5-PWM modulation scheme in the α - β plane. ( a ) AZSL5M5-PWM implemented with odd vectors. (b) AZSL5M5-PWM implemented with even vectors. For the sake of simplicity, all the following analyses are conducted considering the AZSL5M5-PWM variant which exclusively uses odd vectors. The procedure is analogous for even vectors. The application time of each vector can be easily calculated by solving the following system:      V∗ α V∗ β V∗ x V∗ y      =     VmrαVllαVlrαVmlα VmrβVllβVlrβVmlβ Vmrx Vllx Vlrx Vmlx Vmry Vlly Vlry Vmly           δ1 δ2 δ3 δ4      , (6) where V∗ α , V∗ β , V∗ x and V∗ y are the reference voltage projections in the α - β and x - y planes, δ1 , δ2 , δ3 and δ4 are the duty cycles for each vector, and the 4 × 4 matrix is composed of the magnitudes of the vectors to be applied in each sector, where Vmr refers to the modulus of the medium vector on the right side of the reference vector ( Vre f ) (Figure 6, Â ), Vml refers to the modulus of the medium vector on the left (Figure 6, Ã ), and Vll and Vlr refer to the modulus of the large vectors on both left and right sides, respectively (Figure 6,Áand À). Consequently, a 4 ×4 matrix should be defined for each sector. In general, V∗ x and V∗ y are set to zero in order to cancel the voltage third harmonic. From (6), Energies 2020,13, 607 8 of 19 it is possible to explicitly determine the values of the duty cycles δ1 , δ2 , δ3 and δ4 with respect to the reference voltages V∗ αand V∗ β: δ1=α1V∗ α/VDC +β1V∗ β/VDC =−a1sin 2sπ 5V∗ α/VDC +a1cos 2sπ 5V∗ β/VDC, δ2=α2V∗ α/VDC +β2V∗ β/VDC =−a2sin 2(s−1)π 5V∗ α/VDC +a2cos 2(s−1)π 5V∗ β/VDC, δ3=α3V∗ α/VDC +β3V∗ β/VDC =a2sin 2sπ 5V∗ α/VDC −a2cos 2sπ 5V∗ β/VDC, δ4=α4V∗ α/VDC +β4V∗ β/VDC =a1sin 2(s−1)π 5V∗ α/VDC −a1cos 2(s−1)π 5V∗ β/VDC, (7) where being s={1, 2...5} the corresponding sector in the αβ plane, and being a1= ( −5+√5)/q2(5+√5)and a2=q(10/(5+√5). Regarding the practical implementation of the proposed technique, the 2 × 4 matrix Ms can be defined as in (8), where the elements of such matrix can be precalculated for each sector and stored into look-up tables (LUT). In this way, the computational burden and implementation complexity of the algorithm are greatly reduced. Table 3summarizes the values of Msfor each sector s={1, 2...5}. Ms=     α1β1 α2β2 α3β3 α4β4      . (8) Table 3. Values of Msdepending on the αβ plane sector. Sector 1 (M1) Sector 2 (M2) Sector 3 (M3) Sector 4 (M4) Sector 5 (M5)     0.691 −0.224 0 1.176 1.118 −0.363 0 0.726        0.427 0.588 −1.118 0.363 0.691 0.951 −0.691 0.224        −0.427 0.588 −0.691 −0.951 −0.691 0.951 −0.427 −0.588        −0.691 −0.224 0.691 −0.951 −1.118 −0.363 0.427 −0.588        0−0.726 1.118 0.363 0−1.176 0.691 0.224     (11100) (01000) (11000) (11101) (11001)(10000) (10001) (11011) (00001) (10011) (00011) (10111) (11110) (01100) (01110) (00100) (01111) (00110) (00010) (00111) 16 25 29 24 8 28 12 3014 4 10 15 2 723 319 1 27 17 6 5 2 1 4 3 Vref Figure 6. AZSL5M5-PWM modulation vector sequence ( À – Å ) for sector 1 when only odd vectors are applied. The main difference of the AZSL5M5-PWM technique over the AZSL2M2-PWM one is that, there are no strictly phase-opposite vector pairs, as only odd/even vectors can be used (Figure 6). To solve this problem, three active vectors (two medium and one large) are used to replace a zero vector. Energies 2020,13, 607 9 of 19 First, the large vector on the right side of the sector (Figure 6, À ) is applied during t0/ 3, and each medium vector (Figure 6,Äand Å) is applied during t0/3. Since the applied vector sequence has a great impact on the CMV and on the switching losses, a sequence with minimum commutations has been chosen for each sector. For instance, when the reference voltage vector lays in sector 1, the next vector sequence is applied: 11001, 11100, 10000, 01000, 00100, 00010, 01000, 10000, 11100 and 11001 (Figure 6). Odd vector variant AZSL5M5-PWM vector sequences depending on the reference voltage sector are given in Table 4. It is important to note that, for AZSL5M5-PWM, more than one commutation is produced at each vector change. Table 4. AZSL5M5-PWM vector sequences (odd vectors). AZSL5M5-PWM Vector Sequence Sector 1 11001 11100 10000 01000 00100 00010 01000 10000 11100 11001 Sector 2 11100 01110 01000 00100 00010 00001 00100 01000 01110 11100 Sector 3 01110 00111 00100 00010 00001 10000 00010 00100 00111 01110 Sector 4 00111 10011 00010 00001 10000 01000 00001 00010 10011 00111 Sector 5 10011 11001 00001 10000 01000 00100 10000 00001 11001 10011 As the α - β plane is divided into five sectors instead of ten (Figure 5), the linear range of AZSL5M5-PWM is slightly reduced, being the maximum achievable output voltage: VoMAX =4 5cos π 5cos π 5=0.5236VDC. (9) For applications where achieving full linear range is mandatory, an hybrid modulation that extends AZSL5M5-PWM’s linear range is proposed in the following. 4.2. Hybridization of the Proposed Modulation Algorithm Three operation areas have been differentiated in Figure 7to carry out the hybrid modulation algorithm and extend the linear range of AZSL5M5-PWM. The AZSL5M5-PWM hybrid variant (HASZL5M5-PWM) that uses odd vectors (white area) has been chosen as the main modulation scheme. When Vre f steps over the boundaries of the white pentagon, two things may occur. On the one hand, Vre f might remain within the shadowed boundaries. In such case, AZSL5M5-PWM with even vectors would be applied. On the other hand, if Vre f is out of the limits of AZSL5M5-PWM with even or odd vectors, SV-PWM can be used to fulfill the remaining area, marked with lines. This modification extends the linear range of AZSL5M5-PWM up to 26.8%. Further variants with greater CMV reduction could also be considered if a full range RCMV-PWM technique, such as AZSL2M2-PWM, is used instead of SV-PWM. (11100) (01000) (11001) (10000) (00001) (10011) (01110) (00100) (00010) (00111) 16 25 8 28 14 4 2 7 19 1 (11000) (11101) (10001) (11011) (00011) (10111) (11110) (01100) (01111) (00110) 29 24 12 30 10 15 23 3 27 17 Figure 7. HAZSL5M5-PWM. White: AZSL5M5-PWM with odd vectors; shadowed: AZSL5M5-PWM with even vectors; lines: SV-PWM. Energies 2020,13, 607 16 of 19 Author Contributions: M.F. concept for the article, writing, proposed modulation technique development, open-loop and closed-loop models simulations and analysis. A.S-G. EMA model development and support on closed-loop model simulations. E.R. IGBT and thermal model development, support on open-loop simulations and review. I.K. support on modulation technique development, conceptual support, review and supervision. E.I. simulation platform development, review and supervision. J.L.M. review. All authors have read and agreed to the published version of the manuscript. Funding: This work has been supported in part by the Government of the Basque Country within the fund for research groups of the Basque University system IT978-16 and in part by the Government of the Basque Country within the research program ELKARTEK as the project ENSOL (KK-2018/00040). Conflicts of Interest: The authors declare no conflict of itnerest. Abbreviations The following abbreviations are used in this manuscript: AC Alternating current AZS-PWM Active Zero-State Pulse Width Modulation AZSL2M2-PWM Active Zero-State Two Large Two Medium AZSL4-PWM Active Zero-State Four Large AZSL5M5-PWM Active Zero-State Five Large Five Medium CMC Common mode current CMV Common mode voltage DC Direct current EDM Electric discharge machining EMA Electromechanical actuator EMI Electromagnetic interferences FEM Finite element method GaN Gallium nitride HVDC High-voltage direct current IGBT Insulated gate bipolar transistor L5-PWM Five Large Pulse Width Modulation L10-PWM Ten Large Pulse Width Modulation M5-PWM Five Medium Pulse Width Modulation M10-PWM Ten Medium Pulse Width Modulation MEA More Electric Aircrafts PI Proportional Integral PMSM Permanent Magnet Synchronous Machine PWM Pulse Width Modulation RCMV-PWM Reduced Common Mode Voltage Pulse Width Modulation SiC Silicon Carbide SV-PWM Space Vector Pulse Width Modulation THD Total Harmonic Distortion VSI Voltage Source Inverter WBG Wide Bandgap References 1. Uzhegov, N.; Smirnov, A.; Park, C.H.; Ahn, J.H.; Heikkinen, J.; Pyrhönen, J. Design Aspects of High-Speed Electrical Machines With Active Magnetic Bearings for Compressor Applications. IEEE Trans. Ind. Electron. 2017,64, 8427–8436. [CrossRef] 2. Riba, J.R.; López-Torres, C.; Romeral, L.; Garcia, A. Rare-earth-free propulsion motors for electric vehicles: A technology review. Renew. Sustain. Energy Rev. 2016,57, 367–379. [CrossRef] 3. Kumar, M.S.; Revankar, S.T. Development scheme and key technology of an electric vehicle: An overview. Renew. Sustain. Energy Rev. 2017,70, 1266–1285. [CrossRef] 4. Riveros, J.A.; Barrero, F.; Levi, E.; Durán, M.J.; Toral, S.; Jones, M. Variable-Speed Five-Phase Induction Motor Drive Based on Predictive Torque Control. IEEE Trans. Ind. Electron. 2013 ,60, 2957–2968. [CrossRef] Energies 2020,13, 607 17 of 19 5. Negahdari, A.; Yepes, A.G.; Doval-Gandoy, J.; Toliyat, H.A. Efficiency Enhancement of Multiphase Electric Drives at Light-Load Operation Considering Both Converter and Stator Copper Losses. IEEE Trans. Power Electron. 2019,34, 1518–1525. [CrossRef] 6. Liu, Z.; Li, Y.; Zheng, Z. A review of drive techniques for multiphase machines. CES Trans. Electr. Mach. Syst. 2018,2, 243–251. [CrossRef] 7. Diana, M.; Ruffo, R.; Guglielmi, P. PWM Carrier Displacement in Multi-N-Phase Drives: An Additional Degree of Freedom to Reduce the DC-Link Stress. Energies 2018,11, 443. [CrossRef] 8. Zheng, P.; Sui, Y.; Zhao, J.; Tong, C.; Lipo, T.A.; Wang, A. Investigation of a Novel Five-Phase Modular Permanent-Magnet In-Wheel Motor. IEEE Trans. Magn. 2011 ,47, 4084–4087.10.1109/TMAG.2011.2150207. [CrossRef] 9. Cao, W.; Mecrow, B.C.; Atkinson, G.J.; Bennett, J.W.; Atkinson, D.J. Overview of Electric Motor Technologies Used for More Electric Aircraft (MEA). IEEE Trans. Ind. Electron. 2012,59, 3523–3531. [CrossRef] 10. Bozhko, S.; Hill, C.I.; Yang, T. More-Electric Aircraft: Systems and Modeling. In Wiley Encyclopedia of Electrical and Electronics Engineering; American Cancer Society: New York, NY, USA, 2018; pp. 1–31. [CrossRef] 11. Wheeler, P.W.; Clare, J.C.; Trentin, A.; Bozhko, S. An overview of the more electrical aircraft. J. Aerosp. Eng. 2012,227, 578–585. [CrossRef] 12. Deng, Q.; Wang, Z.; Chen, C.; Czarkowski, D.; Kazimierczuk, M.K.; Zhou, H.; Hu, W. Modeling and Control of Inductive Power Transfer System Supplied by Multiphase Phase-Controlled Inverter. IEEE Trans. Power Electron. 2019,34, 9303–9315. [CrossRef] 13. Belkhode, S.; Jain, S. Optimized Switching PWM Technique With Common-Mode Current Minimization for Five-Phase Open-End Winding Induction Motor Drives. IEEE Trans. Power Electron. 2019 ,34, 8971–8980. [CrossRef] 14. Liu, Z.; Zheng, Z.; Xu, L.; Wang, K.; Li, Y. Current Balance Control for Symmetrical Multiphase Inverters. IEEE Trans. Power Electron. 2016,31, 4005–4012. [CrossRef] 15. Prieto, B. Design and Analysis of Fractional-Slot Concentrated-Winding Multiphase Fault-Tolerant Permanent Magnet Synchronous Machines. Ph.D. Thesis, Tecnum Universidad de Navarra, Donostia, Spain, 2015. 16. Iqbal, A.; Rahman, K.; Abdallah, A.A.; Moin, K.A.S.; Abdellah, K. Current Control of a Five-phase Voltage Source Inverter. In Proceedings of the International Conference on Power Electronics and their Applications (ICPEA), Elazig, Turkey, 27 September 2013. 17. Bennett, J.W.; Atkinson, G.J.; Mecrow, B.C.; Atkinson, D.J. Fault-Tolerant Design Considerations and Control Strategies for Aerospace Drives. IEEE Trans. Ind. Electron. 2012,59, 2049–2058. [CrossRef] 18. Bojoi, R.; Cavagnino, A.; Tenconi, A.; Tessarolo, A.; Vaschetto, S. Multiphase electrical machines and drives in the transportation electrification. In Proceedings of the IEEE 1st International Forum on Research and Technologies for Society and Industry Leveraging a better tomorrow (RTSI), Torino, Italy, 18 September 2015. [CrossRef] 19. Takahashi, S.; Ogasawara, S.; Takemoto, M.; Orikawa, K.; Tamate, M. Common-Mode Voltage Attenuation of an Active Common-Mode Filter in a Motor Drive System Fed by a PWM Inverter. IEEE Trans. Ind. Appl. 2019,55, 2721–2730. [CrossRef] 20. Karampuri, R.; Jain, S.; Somasekhar, V.T. Common-Mode Current Elimination PWM Strategy Along with Current Ripple Reduction for Open-Winding Five-Phase Induction Motor Drive. IEEE Trans. Power Electron. 2019,34, 6659–6668. [CrossRef] 21. Espina, J.; Balcells, J.; Arias, A.; Ortega, C. Common mode EMI model for a direct matrix converter. IEEE Trans. Ind. Electron. 2011,58, 5049–5056. [CrossRef] 22. Muetze, A.; Binder, A. Don’t lose your bearings. IEEE Ind. Appl. Mag. 2006,12, 22–31. [CrossRef] 23. Hadden, T.; Jiang, J.W.; Bilgin, B.; Yang, Y.; Sathyan, A.; Dadkhah, H.; Emadi, A. A Review of Shaft Voltages and Bearing Currents in EV and HEV Motors. In Proceedings of the Industrial Electronics Society (IECON), Florence, Italy, 24 October 2016; pp. 1578–1583. 24. Mütze, A. Thousands of hits: On inverter-induced bearing currents, related work, and the literature. Elektrotechnik Inf. 2011,128, 382–388. [CrossRef] 25. Muetze, A. On a New Type of Inverter-Induced Bearing Current in Large Drives With One Journal Bearing. IEEE Trans. Ind. Appl. 2010,46, 240–248. [CrossRef] Energies 2020,13, 607 18 of 19 26. Morya, K.; Gardner, M.C.; Anvari, B.; Liu, L.; Yepes, A.G.; Doval-Gandoy, J.; Toliyat, H.A. Wide Bandgap Devices in AC Electric Drives: Opportunities and Challenges. IEEE Trans. Transp. Electrif. 2019 ,5, 3–20. [CrossRef] 27. Oh, W.; Willwerth, A. Shaft Grounding—A Solution to Motor Bearing Currents. Am. Soc. Heating Refrig. Air Cond. Eng. Trans. 2008,114, 246–251. 28. Muetze, A.; Binder, A. Calculation of influence of insulated bearings and insulated inner bearing seats on circulating bearing currents in machines of inverter-based drive systems. IEEE Trans. Ind. Appl. 2006,42, 965–972. [CrossRef] 29. Muetze, A. Bearing Currents in Inverter-Fed AC-Motors. Ph.D. Thesis, Der Technischen Universitaet Darmstadt, Darmstadt, Germany, 2004. 30. Schiferl, R.F.; Melfi, M.J. Bearing current remediation options. IEEE Ind. Appl. Mag. 2004 ,10, 40–50. [CrossRef] 31. Muetze, A.; Oh, W. Application of Static Charge Dissipation to Mitigate Electric Discharge Bearing Currents. IEEE Trans. Ind. Appl. 2008,44, 135–143. [CrossRef] 32. Nguyen, N.; Nguyen, T.; Lee, H. A reduced siwtching loss PWM strategy to eliminate common-mode voltage in multilevel inverters. IEEE Trans. Power Electron. 2015,30, 5425–5438. [CrossRef] 33. Syed, A.; Kalyani, S. Evaluation of single phase transformerless photovoltaic inverters. Electr. Electron. Eng. Int. J. 2015,4, 25–39. [CrossRef] 34. Freddy, T.K.S.; Rahim, N.A.; Hew, W.P.; Che, H.S. Comparison and Analysis of Single-Phase Transformerless Grid-Connected PV Inverters. IEEE Trans. Power Electron. 2014,29, 5358–5369. [CrossRef] 35. Freddy, T.K.S.; Rahim, N.A.; Hew, W.P.; Che, H.S. Modulation Techniques to Reduce Leakage Current in Three-Phase Transformerless H7 Photovoltaic Inverter. IEEE Trans. Ind. Electron. 2015 ,62, 322–331. [CrossRef] 36. Kouro, S.; Leon, J.I.; Vinnikov, D.; Franquelo, L.G. Grid-Connected Photovoltaic Systems: An Overview of Recent Research and Emerging PV Converter Technology. IEEE Ind. Electron. Mag. 2015 ,9, 47–61. [CrossRef] 37. Fernandez, M.; Robles, E.; Kortabarria, I.; Andreu, J.; Ibarra, E. Novel modulation techniques to reduce the common mode voltage in multiphase inverters. In Proceedings of the IEEE Industrial Electronics Society Conference (IECON), Lisabon, Portugal, 14 October 2019; pp. 1898–1903. 38. Durán, M.J.; Prieto, J.; Barrero, F.; Riveros, J.A.; Guzman, H. Space-Vector PWM with Reduced Common-Mode Voltage for Five-Phase Induction Motor Drives. IEEE Trans. Ind. Electron. 2013 ,60, 4159–4168. [CrossRef] 39. Rahman, A. Power Electronics and Motor Drives; CRC Press: Boca Raton, FL, USA, 2016; pp. 5–10. 40. Gray, R. Toeplitz and Circulant Matrices: A Review. Found. Trends Commun. Inf. Theory 2006 ,2, 155–239. [CrossRef] 41. Semail, E; Bouscayrol, A.; Hautier, P.J. Vectorial formalism for analysis and design of polyphase synchronous machines. Eur. Phys. J. Appl. Phys. 2003,22, 207–220. [CrossRef] 42. Tang, K.T. Mathematical Methods for Engineers and Scientists 1; Springer: Berlin, Germany, 2006. 43. Davis, P. Circulant Matrices; John Wiley & Sons Inc.: Hoboken, NJ, USA, 1979. 44. Zhar, H.; Gong, J.; Semail, E.; Scuiller, F. Comparison of Optimized Control Strategies of a High-speed Traction Machine with Five Phases and Bi-Harmonic Electromotive Force. Energies 2016,9, 952. [CrossRef] 45. Kelly, W.; Strangas, E.G.; Miller, J.M. Multiphase Space Vector Pulse Width Modulation. IEEE Power Eng. Rev. 2002,22, 53–53. [CrossRef] 46. Iqbal, A.; Levi, E. Space vector modulation schemes for a five-phase voltage source inverter. In Proceedings of the European Conference on Power Electronics and Applications (ECPEA), Barcelona, Spain, 10 September 2005; pp. 1–12. [CrossRef] 47. Yen-Shin, L.; Po-Sheng, C.; Hsiang-Kuo, L.; Chou, J. Optimal common-mode voltage reduction PWM technique for inverter control with consideration of the dead-time effects-part II: Applications to IM drives with diode front end. IEEE Trans. Ind. Appl. 2004,40, 1613–1620. [CrossRef] 48. Iqbal Alammari, R.; Mosa, M.; Abu-Rub, H. Finite set model predictive current control with reduced and constant common mode voltage for a five-phase voltage source inverter. In Proceedings of the International Symposium on Industrial Electronics (ISIE), Istambul, Turkey, 4 June 2014; pp. 479–484. [CrossRef] Energies 2020,13, 607 19 of 19 49. Munim, N.A.; Ismail, M.F.; Abidin, A.F.; Haris, H.M. Multi-phase inverter Space Vector Modulation. In Proceedings of the International Power Engineering and Optimization Conference (PEOCO), Langkawi, Malaysia, 22 July 2013; pp. 149–154. [CrossRef] 50. Robles, E.; Fernandez, M.; Ibarra, E.; Andreu, J.; Kortabarria, I. Mitigation of Common Mode Voltage Issues in Electric Vehicle Drive Systems by Means of an Alternative AC-Decoupling Power Converter Topology. Energies 2019,12, 3349. [CrossRef] 51. Sadigh, A.K.; Dargahi, V.; Corzine, K. Analytical determination of conduction power loss and investigation of switching power loss for modified flying capacitor multicell converters. IET Power Electron. 2016 ,9, 175–187. [CrossRef] 52. Wintrich, A.; Nicolai, U.; Tursky, W.; Reimann, T. Application Manual Power Semiconductors; Semikron: Nürnberg, Germany, 2017. 53. Matallana, A.; Robles, E.; Ibarra, E.; Andreu, J.; Delmonte, N.; Cova, P. A methodology to determine reliability issues in automotive SiC power modules combining 1D and 3D thermal simulations under driving cycle profiles. Microelectron. Reliab. 2019,102, 1–9. [CrossRef] 54. Parsa, L.; Toliyat, H. Five-Phase Permanent-Magnet Motor Drives. IEEE Trans. Ind. Appl. 2005 ,41, 30–37. [CrossRef] c 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).