Académique Documents
Professionnel Documents
Culture Documents
Abstract—This paper presents the particularization of a re- modulation technique of [12] takes advantage of redundancy,
cent multilevel multiphase space vector pulse-width modulation which allows extending the linear region of the modulation
algorithm for three-phase converters. This modulation technique index, reducing the number of switchings, and balancing
takes advantage of the switching state redundancy, which permits
to achieve different goals like extending the modulation index capacitors in multilevel converters.
range and reducing the number of switchings. The particularized Those two multiphase SVPWM algorithms are valid for any
algorithm is then compared with an existing multilevel space number of phases and consequently they can be applied to
vector modulation technique showing many similarities. Finally, the standard three-phase converters. In [13], the multiphase
the algorithm is implemented in a low cost FPGA and it is tested
algorithm in [11] was particularized to three-phase converters,
in laboratory with a real prototype by using a neutral point
clamped inverter. and it was shown that in this case the obtained algorithm
is nearly the same as the former generalized 3D SVPWM
I. I NTRODUCTION algorithm presented in [9]. In this paper, the multilevel mul-
The space vector pulse-width modulation (SVPWM) is a tiphase SVPWM algorithm with switching state redundancy
very common modulation technique that is still in focus of the in [12] is particularized for three-phase converters. The re-
research community attention [1]–[5] because it offers a great sulting algorithm is compared with the fast 2D SVPWM
flexibility to optimize the switching waveforms. This technique technique presented in [8]. Finally, it is implemented in a
represents the reference voltage and the switching states of field-programmable gate array (FPGA) and it is tested with
the converter in a vectorial space. In this space, the reference a neutral-point clamped (NPC) inverter feeding and induction
vector is approximated by means of a sequence of switching motor.
vectors. Complexity and computational cost of traditional mul-
tilevel SVPWM techniques increase with the number of levels II. M ULTILEVEL M ULTIPHASE SVPWM A LGORITHM
of the converter, and most of them use trigonometric functions W ITH S WITCHING S TATE R EDUNDANCY
or precomputed tables [6], [7]. In the two-dimensional (2D)
and the three-dimensional (3D) SVPWM algorithms proposed In multiphase converters the SVPWM is a multidimensional
in [8] and [9], respectively, the calculation of the switching problem where the vector selection can be carried out directly
vectors and the switching times is reduced to a few comparison in a multidimensional space. In [12], the modulation problem
and addition operations, which do not depend on the number of a 𝑃 -phase converter is formulated in a 𝑃 -dimensional
of levels. Consequently, those multilevel techniques are very space and it is solved for multilevel topologies in which
suitable for real-time implementation [10]. the output level of every phase is an integer multiple of
Recently, in [11] and [12], two new multiphase SVPWM a fixed voltage step 𝑉dc . Flying capacitor, diode-clamped,
techniques with low computational complexity, which makes cascaded full-bridge and hybrid converters are included in such
them suitable for on-line implementation using low-cost de- topologies. The solution is an algorithm based on a two-level
vices, were presented. Both techniques can be used with the multiphase SVPWM modulator that is valid for any number
typical multilevel topologies such as diode-clamped, flying of levels and phases. This multiphase modulation technique
capacitor, cascaded full-bridge or even hybrid converters. The is able to handle all switching states of the inverter, without
modulation technique of [11] is suitable for converters with discard any one, and it provides a sorted switching vector
neutral wire. In the case of converters with floating neutral, sequence that minimizes the number of switchings. In addition,
this algorithm shows a poor performance because it does the algorithm proves suitable for real-time implementation due
not take into account the switching state redundancy. The to its low computational complexity.
3) Decompose the reference space vector into the sum of 9) Select 𝑃 consecutive integer numbers {𝑞𝑚 } within the
its integer part 𝝎 𝑖 and its fractional part 𝝎 𝑓 : interval [𝑞max , 𝑞min ] according to any desired modula-
tion strategy.
𝝎 𝑖 = integ(𝝎 𝑟 ) (5) 10) Calculate the values of 𝑛𝑚 and 𝑗𝑚 that correspond to
𝝎𝑓 = 𝝎𝑟 − 𝝎𝑖 . (6) each selected index 𝑞𝑚 by means of
( )
4) From the fractional part, obtain the sequence of dis- 𝑞𝑚 − 𝑞𝑖
𝑛𝑚 = integ (17)
placed space vectors 𝝎 𝑑𝑗 = [𝜔𝑑 1 , 𝜔𝑑 2 , . . . , 𝜔𝑑 𝑃 −1 ]T 𝑃
and their dwell times 𝜏𝑗 by means of the two-level 𝑗𝑚 = 𝑞𝑚 − 𝑞𝑖 − 𝑛𝑚 𝑃 + 1. (18)
multiphase SVPWM in [11]. The displaced space vector
11) Obtain the vectors of the switching sequence {v𝑠𝑚 } and
sequence is extracted from the matrix
⎡ ⎤ their switching times {𝑡𝑚 } by means of
1 1 ... 1 v𝑠𝑚 = Tv (𝝎 𝑖 + 𝝎 𝑑𝑗 ) + 𝑛𝑚 [1, 1, . . . , 1]T (19)
⎢ 1
⎢ 𝜔𝑑 1 𝜔𝑑 12 . . . 𝜔𝑑 1𝑃 +1 ⎥
⎥
⎢ ⎥ 𝑡𝑚 = 𝜏𝑗𝑚 , (20)
⎢ 2 2 2
D = ⎢ 𝜔𝑑 1 𝜔𝑑 2 . . . 𝜔𝑑 𝑃 +1 ⎥ (7)
⎥
⎢ . respectively, where Tv is the following 𝑃 × (𝑃 − 1)
⎢ . .. .. .. ⎥
⎣ . . . . ⎦
⎥ transformation matrix:
⎡ ⎤
𝜔𝑑 𝑃
1 𝜔𝑑 𝑃
2 ... 𝜔𝑑 𝑃
𝑃 +1 1 0 ... 0
⎢0 1 ... 0⎥
⎢ ⎥
that is calculated as
⎢. .. .. ⎥
⎢ ⎥
Tv = ⎢ .. .. . (21)
D = PT D̂ (8) ⎢ . . .⎥⎥
⎣0 0 ... 1⎦
⎢ ⎥
where P is the permutation matrix that sorts the com-
ponents of 𝝎 𝑓 in descending order and D̂ is a upper 0 0 ... 0
Voltage (p.u.)
1
in (29). Dwell times corresponding to the three nearest space b
vector are obtained from the fractional parts of the reference 0
−4
v𝑠3 : [𝑣𝑠 𝑔3 , 𝑣𝑠 ℎ3 ]T = [3, −3]T → 𝜏3 = 0.32. 0 0.5 1
Above vectors are the same as the previous ones, except for Difference between both algoritms
1
the way in which they are sorted. Fig. 1 shows that if space Error g
vectors of the fast multilevel SVPWM in [8] algorithm are 0.5 Error h
Diff. (%)
Phase b: 1 2 2 2 1
Motor
150 V
150 V NPC
Phase c: 0 0 1 0 0
inverter
(a) Diagram
ing state redundancy,” IEEE Trans. Ind. Electron., vol. 56, no. 3, pp. [14] O. López, J. Doval-Gandoy, C. M. Peñalver, J. Rey, and F. D. Freijedo,
792–804, Mar. 2009. “Redundancy and basic switching rules in multilevel converters,” Int.
[13] O. López, J. Alvarez, J. Doval-Gandoy, F. D. Freijedo, A. Nogueiras, Review of Electrical Engineering, vol. 0, no. 0, pp. 66–73, Jan.–Feb.
and C. M. Peñalver, “Multilevel multiphase space vector PWM algorithm 2006.
applied to three-phase converters,” in Proc. IEEE Industrial Electronics
Society Conf. IECON, Orlando, FL, 10–13 Nov. 2008, pp. 3288–3293.