Power electronic converters and pulse width modulation inverters


Power electronic convertors are a household of electrical circuits which convert electrical energy from one degree of voltage/current/frequency to another utilizing semiconductor-based electronic switch. The indispensable feature of these types of circuits is that the switches are operated merely in one of two provinces – either to the full ON or to the full OFF – unlike other types of electrical circuits where the control elements are operated in a additive active part.

As the power electronics industry has developed, assorted households of power electronic convertors have evolved, frequently linked by power degree, exchanging devices, and topological beginnings. Application countries of power convertors got immense betterments in semiconducting material engineering, which offer higher electromotive force and current evaluations every bit good as better exchanging features. On the other manus, the chief advantages of modem power electronic convertors, such as high efficiency, low weight, little dimensions, fast operation, and high power densenesss.

The procedure of exchanging the electronic devices in a power electronic convertor from one province to another is called ‘modulation ‘ . Each household of power convertors has preferred transition schemes associated with it that aim to optimise the circuit operation for the mark standards most appropriate for that household. Parameters such as exchanging frequence, deformation, losingss, harmonic coevals, and velocity of response are typical of the issues which must be considered when developing transition schemes for a peculiar household of convertors [ 1 ] .

In modern convertors, PWM is a high- velocity procedure runing depending on the rated power from a few kHz ( motor control ) up to several MHzs ( resonating convertors for power supply ) . Therefore foremost we discuss about the rule and different topologies sing PWM.


The Pulse breadth modulated ( PWM ) inverters are among the most used power-electronic circuits in practical applications. These inverters are capable of bring forthing ac electromotive forces of variable magnitude every bit good as variable frequence. The quality of end product electromotive force can besides be greatly enhanced, when compared with those of square moving ridge inverters. The PWM inverters are really normally used in adjustable velocity Acs motor drive tonss where one needs to feed the motor with variable electromotive force, variable frequence supply. For broad fluctuation in drive velocity, the frequence of the applied ac electromotive force demands to be varied over a broad scope. The applied electromotive force besides needs to change about linearly with the frequence. PWM inverters can be of Single stage every bit good as three stage types. Their rule of operation remains similar. [ 2 ]

Principle of Pulse Width Modulation ( PWM ) :

The District of Columbia input to the inverter is chopped by exchanging devices in the inverter. The amplitude and harmonic content of the Ac wave form is controlled by the responsibility rhythm of the switches. The cardinal electromotive force v1 has max. Amplitude = 4Vd/p for a square moving ridge end product but by making notches, the amplitude of V1 is reduced.

Normally, the ON and OFF provinces of the power switches in one inverter leg are ever face-to-face. Therefore, the inverter circuit can be simplified into three 2-position switches. Either the positive or the negative District of Columbia coach electromotive force is applied to one of the motor phases for a short clip. Pulse width transition ( PWM ) is a method whereby the switched electromotive force pulsations are produced for different end product frequences and electromotive forces. A typical modulator produces an mean electromotive force value, equal to the mention electromotive force within each PWM period. Sing a really short PWM period, the mention electromotive force is reflected by the fundamental of the switched pulsation form. [ 3 ]

There are several different PWM techniques, differing in their methods of execution. However in all these techniques the purpose is to bring forth an end product electromotive force, which after some filtering, would ensue in a good quality sinusoidal electromotive force wave form of coveted cardinal frequence and magnitude. For the inverter topology considered here, it may non be possible to cut down the overall electromotive force deformation due to harmonics but by proper exchanging control the magnitudes of lower order harmonic electromotive forces can be reduced, frequently at the cost of increasing the magnitudes of higher order harmonic electromotive forces. Such a state of affairs is acceptable in most instances as the harmonic electromotive forces of higher frequences can be satisfactorily filtered utilizing lower sizes of filters and capacitances. Many of the tonss, like motor tonss have an built-in quality to stamp down high frequence harmonic currents and therefore an external filter may non be necessary. To judge the quality of electromotive force produced by a PWM inverter, a elaborate harmonic analysis of the electromotive force wave form needs to be done [ 2 ] .

In fact, after taking 3rd and multiples of 3rd harmonics from the pole electromotive force wave form one obtains the corresponding burden stage electromotive force wave form. The pole electromotive force wave forms of 3-phase inverter are simpler to visualise and analyse and hence the harmonic analysis of burden stage and line electromotive force wave forms is done via the harmonic analysis of the pole electromotive forces. It is inexplicit that the burden stage and line electromotive forces will non be affected by the 3rd and multiples of 3rd harmonic constituents that may be present in the pole electromotive force wave forms.

Nature of Pole Voltage Waveforms Output By PWM Inverters:

Unlike in square moving ridge inverters the switches of PWM inverters are turned on and off at significantly higher frequences than the cardinal frequence of the end product electromotive force wave form. The typical pole electromotive force wave form of a PWM inverter is shown in. 1 over one rhythm of end product electromotive force. In a three-phase inverter the other two pole electromotive forces have indistinguishable forms but they are displaced in clip by one tierce of an end product rhythm. Pole electromotive force wave form of the PWM inverter alterations polarity several times during each half rhythm. The clip cases at which the electromotive force mutual oppositions reverse have been referred here as notch angles. It may be noted that the instantaneous magnitude of pole electromotive force wave form remains fixed at half the input District of Columbia electromotive force ( Edc ) . When upper switch ( SU ) , connected to the positive District of Columbia coach is on, the pole electromotive force is + 0.5 Edc and when the lower switch ( SL ) , connected to the negative District of Columbia coach, is on the instantaneous pole electromotive force is – 0.5 Edc. The exchanging passage clip has been neglected in conformity with the premise of ideal switches. It is to be remembered that in electromotive force beginning inverters, meant to feed an inductive type burden, the upper and lower switches of the inverter pole behavior in a complementary mode. That is, when upper switch is on the lower is off and vice-versa. Both upper and lower switches should non stay on at the same time as this will do short circuit across the District of Columbia coach. On the other manus one of these two switches in each pole ( leg ) must ever carry on to supply continuity of current through inductive tonss. A sudden break in inductive burden current will do a big electromotive force spike that may damage the inverter circuit and the burden.

The followers are some major concerns when comparing different PWM techniques:

* Good use of DC power supply, that is to present a higher end product electromotive force with the same DC supply ;

* Good one-dimensionality in electromotive force and or current control ;

* Low harmonics contents in the end product electromotive force and or currents, particularly in the low-frequency part ;

* Low shift losingss.

There are several other PWM techniques, the of import 1s are: –


In many industrial applications, sinusoidal Pulse breadth transition ( SPWM ) , besides called sine coded pulse width transition, is used to command the inverter end product electromotive force. SPWM maintains good public presentation of the thrust in the full scope of operation between nothing and 78 per centum of the value that would be reached by square moving ridge operation. If the transition index exceeds this value, additive relationship between transition index and end product electromotive force is non maintained and the over-modulation methods are required.

Sinusoidal PWM refers to the coevals of PWM end products with sine moving ridge as the modulating signal. The ON and OFF blink of an eyes of a PWM signal in this instance can be determined by comparing a mention sine moving ridge ( the modulating moving ridge ) with a high frequence triangular moving ridge ( the bearer moving ridge ) as shown in. 3. Sinusoidal PWM technique is normally used in industrial applications and is abbreviated here as SPWM. The frequence of the modulating moving ridge determines the frequence of the end product electromotive force. The peak amplitude of modulating moving ridge determines the transition index and in bend controls the RMS value of end product electromotive force. The RMS value of the end product electromotive force can be varied by altering the transition index. This technique improves deformation factor significantly compared to other ways of multi-phase transition. It eliminates all harmonics less than or equal to 2p-1, where “ P ” is defined as the figure of pulsations per half rhythm of the sine moving ridge. The end product electromotive force of the inverter contains harmonics. However, the harmonics are pushed to the scope around the bearer frequence and its multiples.

degree Fahrenheit = frequence of pulse width transition ( triangular moving ridge )

f1 = frequence of cardinal moving ridge ( command wave ) )

The pole electromotive force VA0 is +Vdc/2, when Vcontrol & A ; gt ; Vtri and -Vdc/2 when Vcontrol & A ; lt ; Vtri. The line electromotive force VAB= VAO-VBO

Amplitude transition ratio ( mom ) :

Frequency transition ratio ( medium frequency ) :

& A ; Oslash ; medium frequency should be an uneven whole number

* if medium frequency is non an whole number, there may be bomber harmonics at end product electromotive force

* if medium frequency is non uneven, DC constituent may be and even harmonics are present at end product electromotive force

& A ; Oslash ; medium frequency should be a multiple of 3 for three-phase PWM inverter

* An uneven multiple of three and even harmonics are suppressed.

The advantages of SPWM for the convertor application are pulling sinusoidal line currents with low harmonic contents, high power factor, District of Columbia nexus electromotive force ordinance, and possible bidirectional power flow [ ] .

Ranjan K. Behera at al. , [ ] presented generalised analytical solutions for different multilevel pulsation width transitions schemes like In-phase Sinusoidal Pulse Width Modulations ( IPSPWM ) , Phase Opposite Sinusoidal Pulse Width Modulation ( POSPWM ) and dipolar transition techniques for three-level NPC inverter. They identified that IPSPWM has a superior spectral public presentation compared to other two transition strategies.

Giuseppe Carrara et al. , [ ad ] analyzed the multilevel transition processes with a powerful and mathematically strict method that provides the analytical looks of the end product stage electromotive forces of the inverter. The betterments in the harmonic contents due to the increased figure of degrees were highlighted. However, several considerations on the existent construction of the inverter and on the system in which it has to be employed should be done instance by instance to find the practical convenience of this solution. On the other manus it is deserving observing that the multilevel attack is the merely allowable when both reduced harmonic contents and high power are required.

Lazhar Ben-Brahim at al. , [ ] described a new PWM control methods based on 1 ) adding a prejudice to the mention, and 2 ) shift form for GTO minimal on-pulse compensation which improved the end product wave forms without increasing the shift losingss. These methods had contributed to the betterment of the features of a GTO based NPC inverter.

Adrian Schiop [ ] presented a method based on the sinusoidal PWM for patterning and simulation of the single-phase rectifying tube clamped multilevel inverters and capacitance clamped multilevel inverters. The theoretical accounts presented are used to execute a harmonic analysis of the end product electromotive force of these multilevel inverters.

P. K. Chaturvedi at al. , [ ] investigated the constructs of sinusoidal pulsation breadth transition, optimized harmonic stepped wave form, and selective harmonic riddance techniques. These techniques well reduced the lower order harmonics in three stage three-level and five-level rectifying tube clamped inverters.

Joachim [ ] evaluated the province of the art in pulse breadth transitions for ac thrusts fed from three-phase electromotive force beginning inverters. He described feed frontward and feedback pulse width transition strategies for industrial applications and described secondary effects such as transients in synchronised pulsation breadth transition strategies and equal compensation methods.


The SVPWM method is an advanced, calculation intensive PWM method and is perchance the best among all the PWM techniques for variable frequence thrust applications. Because of its superior public presentation features, it has been happening broad spread application in recent old ages.


The infinite vector pulse breadth transition ( SVPM ) technique is more popular than conventional technique because of the undermentioned first-class characteristics:

* It achieves the broad additive transition scope associated with PWM, third-harmonic injection automatically.

* It has lower base set harmonics than regular PWM or other sine based transition methods, or otherwise optimizes harmonics.

* 15 % more end product electromotive force so conventional transition, i.e. better DC-link use.

* More efficient usage of District of Columbia supply electromotive force

* SVM increases the end product capableness of SPWM without falsifying line-line end product electromotive force wave form.

* Advanced and calculation intensive PWM technique.

* Higher efficiency.

* Prevent un-necessary exchanging hence less commuting losingss.

* A different attack to PWM transition based on infinite vector representation of the electromotive forces in the ?-? plane.

F. Profumo et al. , [ ac ] focused a general overview of the SPWM and SVPWM techniques. Particular accent has been done on PWM jobs due to the secondary effects. The complete analysis and the obtained simulation consequences have been reported. The sinusoidal PWM wave forms have cleaner spectra than the infinite vector PWM, but the quality factors are higher with regard to the infinite vector PWM holding the same peak value of the electromotive force mention. As a effect, if the shift frequence is high plenty, the losingss due to the harmonics can be about neglected, and the Space Vector PWM seems to be the best solution in footings of end product electromotive force, harmonic losingss and figure of exchanging per rhythm.

Zhenyu Yu at al. , [ ] described and reviewed the three normally used PWM techniques, sinusoidal PWM technique, infinite vector PWM techniques and hysteresis PWM techniques. They implemented these techniques with digital processor such as TMS320C240 and presented the better use of DC supply and decrease of harmonics of infinite vector PWM vs. sinusoidal PWM.

Keliang Zhou at al. , [ ] investigated the relationship between the carrier-based PWM and space-vector transition. They described the relationships between the transition signals ( include cardinal signals and zero-sequence signal ) and infinite vectors, between the transition signals and space-vector sectors, and between the exchanging form of space-vector transition and the type of bearer.

Abdul Hamid Bhat et al. , [ ab ] described an improved public presentation three stage, neutral-point clamped bidirectional rectifier with simplified control strategy. A complete mathematical theoretical account of the rectifier utilizing PWM accountant is developed. A comparative analysis of two-level and three-level convertor is evaluated. They discussed the virtues and demerits of both types of convertors.

Variable electromotive force and frequence supply to ac thrusts is constantly obtained from a three-phase electromotive force beginning inverter ( VSI ) . A figure of Pulse width transition ( PWM ) strategies are used to obtain variable electromotive force and frequence supply. AtifIqbal et al. , [ BB ] presented a measure by measure development of Matlab/Simulink theoretical account to implement SVPWM for three stage VSI. A brief reappraisal of VSI theoretical account is besides reported based on infinite vector representation.

Jang-Hwan Kim at al [ ] analyzed the relationship between the infinite vector PWM and the carrier-based PWM method, and they proposed a fresh carrier-based PWM scheme to equilibrate the impersonal point potency. They analytically described the electromotive force mistake of the transition caused by the imbalance of the impersonal point potency.


The construct of infinite vector is derived from the revolving field of AC machine which is used for modulating the inverter end product electromotive force. In this transition technique the three stage measures can be transformed to their tantamount two-phase measure either in synchronously revolving frame ( or ) stationary frame. From this two-phase constituent the mention vector magnitude can be found and used for modulating the inverter end product. The procedure of obtaining the revolving infinite vector is explained in the undermentioned subdivision, sing the stationary mention frame.

2.3.1 Principle of Space Vector PWM

The SVPWM treats the sinusoidal electromotive force as a changeless amplitude vector revolving at changeless frequence. This PWM technique approximates the mention electromotive force Vref by a combination of the eight exchanging forms ( V0 to V7 ) . A three-phase electromotive force vector is transformed into a vector in the stationary d-q co-ordinate frame which represents the spacial vector amount of the three-phase electromotive force.

2.4 Space Vector PWM for two degree Inverter:

The circuit theoretical account of a typical three-phase electromotive force beginning PWM inverter is shown in.2.2. S1 to S6 are the six power switches that shape the end product, which are controlled by the shift variables a, a? , B, b? , degree Celsius and c? . When an upper transistor is switched ON, i.e. , when a, B or degree Celsius is 0 the matching lower transistor is switched OFF, and when a? , b? or c? is 1 the corresponding upper transistor is switched OFF. Therefore, the ON and OFF provinces of the upper transistors S1, S3 and S5 can be used to find the end product electromotive force.

As illustrated in.2.2, there are eight possible combinations of ON and OFF forms for the three upper power switches. The ON and OFF provinces of the lower power devices are opposite to the upper one and so are easy determined once the provinces of the upper power transistors are determined.

To implement the infinite vector PWM, the electromotive force equations in the rudiment mention frame can be transformed into the stationary d-q mention frame that consists of the horizontal ( vitamin D ) and perpendicular ( Q ) axes as depicted in. 2.3

As described in. 2.3, this transmutation is tantamount to an extraneous projection of [ a, B, hundred ] onto the planar perpendicular to the vector [ 1, 1, 1 ] ( the equivalent d-q plane ) in a 3-dimensional co-ordinate system. As a consequence, six non-zero vectors and two nothing vectors are possible. Six nonzero vectors ( V1 – V6 ) shape the axes of a hexangular as depicted in 2.4 and feed electric power to the burden. The angle between any next two non-zero vectors is 60 grades. Meanwhile, two nothing vectors ( V0 and V7 ) are at the beginning and use zero electromotive force to the burden. The eight vectors are called the basic infinite vectors and are denoted by V0, V1, V2, V3, V4, V5, V6, and V7. The same transmutation can be applied to the desired end product electromotive force to acquire the coveted mention electromotive force vector Vref in the d-q plane. The aim of infinite vector PWM technique is to come close the mention electromotive force vector Vref utilizing the eight exchanging forms. One simple method of estimate is to bring forth the mean end product of the inverter in a little clip period T, to be same as that of Vref in the same clip period.

2.5 DETERMINATION OF Switching States

Let us see a province say V0.This province at which all of the upper devices are in the ON place. So in this province all of the gate pulsations are given to the upper devices of the inverter circuit, i.e. , S1, S3, S5. So the end point end product electromotive force will be zero. The moving ridge signifier will be at the zero degree. This can be illustrated as shown in the below.2.5.

So from the above ure all the three upper switches are in the ON province and the lower 1s are in the OFF province. Similarly all the other provinces can be represented in the diagrammatical mode as shown.2.6

State 1: ( 1 0 0 )

, ,

State 2: ( 1 1 0 )

, ,

State 3: ( 0 1 0 )

, ,

State 4: ( 0 1 1 )

, ,

State 5: ( 0 0 1 )

, ,

State 6: ( 1 0 1 )

, ,

State 7: ( 1 1 1 )

, ,

State 8: ( 0 0 0 )

, ,

Representing all the above mentioned eight electromotive force vectors in infinite, the SVPWM for two degree inverter is built. The infinite vector locations for a two-level inverter signifier the vertices of a regular hexagon, organizing 6 sectors as shown in. 2.7. For the provinces ( 000 ) and ( 111 ) the motor stages are short-circuited and hence are non connected to the beginning. These provinces are called the Zero provinces or void provinces during which there is no power flow from the beginning to the motor. Hence, by commanding the continuance of these zero province intervals, one can command the magnitude of the output-voltage. It is deserving observing that in six-state manner of operation, such intervals of zero province exchanging do non be. Consequently, commanding the input DC nexus electromotive force controls the magnitude of the end product electromotive force in an inverter operating in a square moving ridge manner. The remainder of the vectors 1 through 6 are called the active vectors.

2.6 Calculation of exchanging times

Switch overing times of the SVPWM based inverter can be calculated by utilizing volt-sec relation. .2.8 represents the computation based on the voltage-sec relation of the mention vector Vsr. The volt-seconds produced by the vectors V1, V2 and V7 orV0 along vitamin D and Q axes are the same as those produced by the mention vector Vsr.

2.7 Optimized Switching Sequence

In order to minimise the figure of exchanging in the inverter, the following optimized exchanging form is selected [ 15 ] .

Table 2.1: Switch Form



( ON sequence )


( OFF sequence )
























Upper switches

( S1, S3, S5 )

Lower switches

( S4, S6, S2 )


S1 = T1 + T2 + T0 /2

S3 = T2 + T0 /2

S5 = T0 /2

S4 = T0 /2

S6 = T1 + T0 /2

S2 = T1 + T2 + T0 /2


S1 = T1 + T0 /2

S3 = T1 + T2 + T0 /2

S5 = T0 /2

S4 = T2 + T0 /2

S6 = T0 /2

S2 = T1 + T2 + T0 /2


S1 = T0 /2

S3 = T1 + T2 + T0 /2

S5 = T2 + T0 /2

S4 = T1 + T2 + T0 /2

S6 = T0 /2

S2 = T1 + T0 /2


S1 = T0 /2

S3 = T1 + T0 /2

S5 = T1 + T2 + T0 /2

S4 = T1 + T2 + T0 /2

S6 = T2 + T0 /2

S2 = T0 /2


S1 = T2 + T0 /2

S3 = T0 /2

S5 = T1 + T2 + T0 /2

S4 = T1 + T0 /2

S6 = T1 + T2 + T0 /2

S2 = T0 /2


S1 = T1 + T2 + T0 /2

S3 = T0 /2

S5 = T1 + T0 /2

S4 = T0 /2

S6 = T1 + T2 + T0 /2

S2 = T2 + T0 /2

Space Vector PWM is considered a better technique of PWM execution owing to its associated advantages mentioned below

* better cardinal end product electromotive force

* better harmonic public presentation

* Easier execution in Digital Signal Processor and Microcontrollers.

The two-level inverters are holding certain drawbacks.

* These are non suited for high power degrees.

* High DC nexus electromotive force requires series connexion of devices.

* Difficult in dynamic electromotive force during exchanging.

Multi-level topology has been applied in several state of affairss, such as high electromotive force AC thrust, FACTS, SVC and so on. Multi-level topology has advantages over traditional two-level topology as followers:

* The electromotive force blocked by the power device is decreased enormously,

* Multilevel inverters produce low harmonic deformation for Ac currents even when operated at moderate shift frequence.

* The switch losingss are lower than two-level inverters.


Multilevel inverter engineering has emerged late as a really of import option in the country of high-power medium-voltage energy control. And the Main characteristics of multi degree inverter are

* Ability to cut down the electromotive force emphasis on each power device due to the use of multiple degrees on the DC coach

· Important when a high DC side electromotive force is imposed by an application ( e.g. grip systems )

· Even at low exchanging frequences, smaller deformation in the multilevel inverter AC side wave form can be achieved ( with stepped transition technique )

Jae Hyeong Seo at al. , [ ] proposed new simplified space-vector pulsation breadth transition ( SVPWM ) method for three-level inverter based on the simplification of the space-vector diagram of a three-level inverter into that of a two-level inverter the staying processs necessary for the three-level SVPWM are done like conventional two-level inverter. They developed three-level IGBT inverter system was applied to the steel doing mill of Pohang Steel Corporation ( POSCO ) .

H. Pinheiro at al. , [ ] described a incorporate attack of the infinite vector transitions for electromotive forces beginning inverters and applied to single-phase full-bridge, three-phase three-wire, three-phase four-wire, three-phase four-leg and three-phase three-level inverters. Switch overing vectors, separation and boundary planes in the inverter end product infinite every bit good as decomposition matrices and possible shift sequences are derived.

A. Koochaki at al. , [ ] proposed a individual stage application of infinite vector pulse breadth transition for shunt active power filters. In conventional SVPWM, all of the stage ‘s currents are controlled together, but in this method, they controlled each of stage currents independently from the mensural currents of other stages and they demonstrated that proposed method has good public presentation in coevals of all types of compensation currents by active power filters.

Zeliang Shu at al. , [ ] developed a compact algorithm of infinite vector pulse breadth transition for three-phase inverter. The conventional SVPWM is decomposed in to fast integer operations wholly by utilizing an intermediate vector, which will decently antagonize the excess computations of the staying processs. This has examined merely in two degree inverter applications.

Subrata K. Mondal at al. , [ ] proposed the infinite vector PWM algorithm for a three-level voltage-fed inverter has been extended to over transition scope thy proposed that the over transition scheme easy blends with the under transition algorithm so that the inverter can run swimmingly from low velocity to the extended velocity scope.

A important job with neutral-point-clamped three-level inverters is the fluctuation in the impersonal point electromotive force. Rangarajan at al. , [ ] developed capacitance electromotive force reconciliation technique for carrier-based three-level PWM. It was shown that the method is applicable to both dipolar and unipolar manners, and that the inverter outputs characteristic three-level wave forms even in electromotive force rectification manner.

Prasad N. Enjeti at al. , [ ] utilized the shift map attack to deduce relevant analytical looks for input/output variables. They proposed optimum power control schemes for an NPC inverter using programmed PWM forms. With the proposed PWM form, the frequence of the first important harmonic constituent at the inverter end product is at least 3 p.u. for a switching frequence of 1 p.u.

R. Sommer at al. , [ ] presented a new scope of medium electromotive force motor thrusts with a three-level impersonal point clamped inverter utilizing high electromotive force IGBTs. Their control is based on field-oriented vector control and an optimized PWM modulator, so that exchanging losingss and current harmonics are minimized and efficiency is optimized.

Satoshi Ogasawara at al. , [ ] analyzed the impersonal point possible fluctuation of the NPC electromotive force beginning PWM inverter for ac motor thrusts and inactive volt-ampere compensations, with the focal point on the current fluxing out of or in to the impersonal point of District of Columbia nexus.

By and large, the reconciliation of the DC-link electromotive force degrades at really low runing frequences of the inverter. Most proposed methods of impersonal point equilibrating techniques consequence in an addition of the shift losingss of the inverter. To work out this job, Lazhar Ben-Brahim [ ] proposed a new energy salvaging PWM method which consequences in a important decrease of the fluctuation of the impersonal point electromotive force of NPC inverters. That method incorporated techniques to the built-in minimal ON-OFF pulsation breadth restriction in a GTO without increasing the shift losingss of the devices.

Imbalance of the neutral-point potency in three-level impersonal point-clamped convertor may look in some operating conditions. Jan-Hwan Kim at al. , [ ] analyzes the electromotive force deformation caused by the instability of the impersonal point potency in a three-level neutral-point-clamped convertor, and besides depict the PWM method to bring forth the low-frequency harmonics-free end product electromotive forces even under the unbalanced status. In the proposed method they used the three electromotive force vectors near to the mention vector, which has advantages in footings of the high-frequency harmonics mention vector are used.

Ramon C. Portillo at al. , [ ] described an analytical scheme to pattern a consecutive three-level convertor. And different incorporated to the overall theoretical account. This theoretical account pays particular attending to the imbalance in the capacitances electromotive force of three-level convertors, including the kineticss of the capacitances electromotive force.

Zeliang Shu at al. , [ ] developed a compact algorithm of infinite vector pulse breadth transition for three-phase inverter. The conventional SVPWM is decomposed in to fast integer operations wholly by utilizing an intermediate vector, which will decently antagonize the excess computations of the staying processs. This has examined merely in two degree inverter applications.

S. Brovanov at al. , [ ] described a SVPWM provender frontward technique that taken in to account the DC electromotive force imbalance. They showed that responsibility rhythm of infinite vector have taken restrictions that depends on the value of the imbalance DC electromotive force.

There are three chief topologies for multi degree inverter




Panagiotis Panagis at al. , [ ] performed a comparing between bing province of the art multilevel inverter topologies. The topologies examined are the Neutral Point Clamp Multilevel inverter ( NPCMLI ) or Diode-Clamped Multilevel Inverter ( DCMLI ) , the Flying Capacitor Multilevel Inverter ( FCMLI ) and the Cascaded Cell Multilevel Inverter ( CCMLI ) . They compared of these inverters is based on the standards of end product electromotive force quality ( Peak value of the cardinal and dominant harmonic constituents and THD ) , power circuitry complexness, and execution cost.


Diode-clamped multilevel inverter ( DCMI ) is an Extension of impersonal point clamped, and it is based on construct of utilizing rectifying tubes to restrict power devices voltage emphasis Structure and basic operating rule Consists of series connected capacitances that divide DC coach electromotive force into a set of capacitance electromotive forces

* A DCMI with n figure of degrees typically comprises ( n-1 ) capacitances on the DC coach.

* Voltage across each capacitance is VDC/ ( n -1 )

( N nodes on DC coach, n degrees of end product stage electromotive force, ( 2n -1 ) degrees of end product line electromotive force )

* Output stage electromotive force can presume any electromotive force degree by choosing any of the nodes.

* DCMI is considered as a type of multiplexer that attaches the end product to one of the available nodes.

* Consists of chief power devices in series with their several chief rectifying tubes connected in analogue and clamping rectifying tubes.

* Main diodes behavior merely when most upper or lower node is selected.

* Although chief rectifying tubes have same electromotive force evaluation as chief power devices, much lower current evaluation is allowable.

* In each stage leg, the forward electromotive force across each chief power device is clamped by the connexion of rectifying tubes between the chief power devices and the nodes.

* Number of power devices in ON province for any choice of node is ever equal to ( n -1 ) .

For illustration three-phase six-level diode-clamped inverter is shown in ure 1. Each of the three stages of the inverter portions a common District of Columbia coach, which has been sub divided by five capacitances into six degrees. The electromotive force across each capacitance is Vdc, and the electromotive force emphasis across each shift device is limited to Vdc through the clamping rectifying tubes. Table 31.1 lists the end product electromotive force degrees possible for one stage of the inverter with the negative District of Columbia rail electromotive force V0 as a mention. State status 1 means the switch is on, and 0 means the switch is away. Each stage has five complementary switch braces such that turning on one of the switches of the brace require that the other complementary switch be turned off. The complementary switch braces for stage leg ‘a ‘ are ( Sa1, Sa’1 ) , ( Sa2, Sa’2 ) , ( Sa3, Sa’3 ) , ( Sa4, Sa’4 ) , and ( Sa5, Sa’5 ) . Table 1 besides shows that in a diode-clamped inverter, the switches that are on for peculiar stage legs are ever next and in series. For a six-level inverter, a set of five switches is on at any given clip.

Voltage Vao

Switch overing provinces











V5 = 5Vdc











V4 = 4Vdc











V3 = 3Vdc











V2 = 2Vdc











V1 = 1Vdc











Vo = 0











Table 1. Diode-clamped six-level inverter electromotive force degrees and matching exchanging provinces

· General characteristics

* For three-phase DCMI, the capacitances need to filtrate merely the high-order harmonics of the clamping rectifying tubes currents, low-order constituents per se cancel each other.

* For DCMI using measure transition scheme, if n is sufficiently high, filters may non be required at all due to the significantly low harmonic content.

* Ifeach clamping rectifying tube has same electromotive force evaluation as power devices, for n-level DCMI.

* Number of clamping diodes/phase = ( n-1 ) ten ( n-2 ) .

* Each power device blocks merely a capacitance electromotive force.

· Advantages:

* All of the stages portion a common District of Columbia coach, which minimizes the electrical capacity demands of the convertor. For this ground, a consecutive topology is non merely possible but besides practical for utilizations such as a high-potential consecutive inter-connection or an adjustable velocity thrust.

* The capacitances can be pre-charged as a group.

* Efficiency is high for cardinal frequence shift.

· Disadvantages:

* Real power flow is hard for a individual inverter because the intermediate District of Columbia degrees will be given to soak or dispatch without precise monitoring and control.

* The figure of clamping rectifying tubes required is quadratically related to the figure of degrees, which can be cumbersome for units with a high figure of degrees.

After general description of all celebrated solutions, the focal point is paid on Diode Clamped Multilevel Inverters ( DCMI ) and Flying Capacitor Multilevel Inverters ( FCMI ) . The comparing of topological construction differences and control schemes were presented by Oleg Sivkov at al. , [ ] . The particular focal point is paid to reconciliation of electromotive forces on capacitances. Switch overing provinces and their passages of three-level inverter allow equilibrating the capacitance electromotive forces in both types of inverters. The advantages and disadvantages of DCMI and FCMI are compared for the same end product power.


Flying capacitance multi degree inverter is Capable of work outing capacitance electromotive force imbalance job and inordinate rectifying tube count demand in rectifying tube capacitance multi flat inverter and it is besides known as winging capacitance multilevel inverter ( capacitances are arranged to drift with regard to Earth ) . Structure and basic runing rule of winging capacitance multi degree inverter are

* Employs separate capacitances pre-charged to [ ( n-1 ) / ( n-1 ) x VDC ] , [ ( n-2 ) / ( n-1 ) x VDC ] … { [ n- ( n-1 ) ] / [ n-1 ] x VDC }

* Size of electromotive force increase between two capacitances defines size of electromotive force stairss in winging capacitance multi degree inverter end product electromotive force wave form

* n-level winging capacitance multi degree inverter has n degrees end product stage electromotive force and ( 2n-1 ) degrees end product line electromotive force

* Output electromotive force produced by exchanging the right combinations of power devices to let adding or subtracting of the capacitance electromotive forces

For illustration the three stage six-level winging capacitance multi degree inverter is shown in ure 2, the construction of this inverter is similar to that of the diode-clamped inverter except that alternatively of utilizing clamping rectifying tubes, the inverter uses capacitances in their topographic point.

This topology has a ladder construction of dc side capacitances, where the electromotive force on each capacitance differs from that of the following capacitance.

§ General characteristics:

* With measure transition scheme, with sufficiently high N, harmonic content can be low plenty to avoid the demand for filters.

* Advantage of interior electromotive force degrees redundancies – allows discriminatory charging or discharging of single capacitances, facilitates use of capacitance electromotive forces so that their proper values are maintained.

* Active and reactive power flow can be controlled.

* Additional circuit required for initial charging of capacitances.

§ Advantages:

* Phase redundancies are available for equilibrating the electromotive force degrees of the capacitances.

* Real and reactive power flow can be controlled.

* The big figure of capacitances enables the inverter to sit through short continuance outages and deep electromotive force droops.

§ Disadvantages:

* Control is complicated to track the electromotive force degrees for all of the capacitances. Besides, pre-charging all of the capacitances to the same electromotive force degree and startup are complex.

* Switch overing use and efficiency are hapless for existent power transmittal. The big Numberss of capacitances are both more expensive and bulky than clamping rectifying tubes in multilevel diode-clamped convertors. Packaging is besides more hard in inverters with a high figure of degrees.


It referred to Modular structured multilevel inverter ( MSMI ) or series connected H-bridge inverters, Structure and basic runing rule of cascade H-Bridge multi degree inverter Consists of ( n-1 ) /2 or h figure of single-phase H-bridge inverters ( MSMI faculties ) i.e.. ,

MSMI end product stage electromotive force is

V0 = Vm1 + Vm2+ … … .. +Vmh

Vm1: end product electromotive force of faculty 1

Vm2: end product electromotive force of faculty 2

Vmh: end product electromotive force of faculty H

A single-phase construction of an m-level cascaded inverter is illustrated in ure 3. Each separate District of Columbia beginning ( SDCS ) is connected to a single-phase full-bridge, or H-bridge, inverter. Each inverter degree can bring forth three different electromotive force end products, +Vdc, 0, and -Vdc by linking the District of Columbia beginning to the Ac end product by different combinations of the four switches, S1, S2, S3, and S4. To obtain +Vdc, switches S1 and S4 are turned on, whereas -Vdc can be obtained by turning on switches S2 and S3. By turning on S1 and S2 or S3 and S4, the end product electromotive force is 0. The ac end products of each of the different full-bridge inverter degrees are connected in series such that the synthesized electromotive force wave form is the amount of the inverter end products. The figure of end product stage electromotive force degrees m in a cascade inverter is defined by m = 2s+1, where s is the figure of separate District of Columbia beginnings.

§ Advantages:

· The figure of possible end product electromotive force degrees is more than twice the figure of dc beginnings ( thousand = 2s + 1 ) .

· The series of H-bridges makes for modularized layout and packaging. This will enable the fabrication procedure to be done more rapidly and cheaply.

§ Disadvantages:

· Separate District of Columbia beginnings are required for each of the H-bridges. This will restrict its application to merchandises that already have multiple SDCSs readily available.

M.G. Hosseini Aghdam at al. , [ ] proposed Homotopy algorithm to work out the non additive transcendent equations which are formed to happen switching angles of the devices in a cascaded H-bridge multi-level inverter with unequal DC beginnings, in order to extinguish some selected harmonics from the end product electromotive force. The proposed algorithm is really effectual, efficient and dependable in happening solutions to high-order nonlinear equations and solves the nonlinear transcendent equations with a much simpler preparation. Computer simulations based on a seven-level cascaded H-bridge inverter have been provided for the confirmation.

4. Other Multilevel Inverter Structures:

Besides the three basic multilevel inverter topologies antecedently discussed, other multilevel convertor topologies have been proposed ; nevertheless, most of these are “ intercrossed ” circuits that are combinations of two of the basic multilevel topologies or little fluctuations to them. Additionally, the combination of multilevel power convertors can be designed to fit with a specific application based on the basic topologies.

A. Generalized Multilevel Topology

Existing multilevel convertors such as diode-clamped and capacitor-clamped multilevel convertors can be derived from the generalised convertor topology called P2 topology proposed by Peng as illustrated in ure 4. The generalised multilevel convertor topology can equilibrate each electromotive force degree by itself irrespective of burden features, active or reactive power transition and without any aid from other circuits at any figure of degrees automatically. Therefore, the topology provides a complete multilevel topology that embraces the bing multilevel convertors in rule.

B. Mixed-Level Hybrid Multilevel Converter

To cut down the figure of separate DC beginnings for high-voltage, high-power applications with multilevel convertors, diode-clamped or capacitor-clamped convertors could be used to replace the full-bridge cell in a cascaded convertor. An illustration is shown in ure 5. The nine-level cascade convertor incorporates a three-level diode-clamped convertor as the cell. The original cascaded H-bridge multilevel convertor requires four separate DC beginnings for one stage leg and 12 for a three-phase convertor. If a five-level convertor replaces the full-bridge cell, the electromotive force degree is efficaciously doubled for each cell. Therefore, to accomplish the same nine electromotive force degrees for each stage, merely two separate DC beginnings are needed for one stage leg and six for a three-phase convertor. The conuration has mixed-level intercrossed multilevel units because it embeds multilevel cells as the edifice block of the cascade convertor. The advantage of the topology is it needs less separate DC beginnings. The disadvantage for the topology is its control will be complicated due to its intercrossed construction.

C. Soft-Switched Multilevel Converter

Some soft-switching methods can be implemented for different multilevel convertors to cut down the shift loss and to increase efficiency. For the cascaded convertor, because each convertor cell is a bi-level circuit, the execution of soft shift is non at all different from that of conventional bi-level convertors. For capacitor-clamped or diode-clamped convertors, soft-switching circuits have been proposed with different circuit combinations. One of soft-switching circuits is a zero-voltage-switching type which includes subsidiary resonant commutated pole ( ARCP ) , coupled inductance with zero-voltage passage ( ZVT ) , and their combinations as shown in ure 6.

D. Back-to-Back Diode-Clamped Converter

Two multilevel convertors can be connected in a consecutive agreement and so the combination can be connected to the electrical system in a series-parallel agreement as shown in ure 7. Both the current demanded from the public-service corporation and the electromotive force delivered to the burden can be controlled at the same clip. This series-parallel active power filter has been referred to as a cosmopolitan power conditioner when used on electrical distribution systems and as a cosmopolitan power flow accountant when applied at the transmittal degree. Previously, Lai and Peng proposed the consecutive diode-clamped topology shown in ure 8.for usage as a high-potential District of Columbia inter connexion between two asynchronous Acs systems or as a rectifier/inverter for an adjustable velocity thrust for high-voltage motors. The diode-clamped inverter has been chosen over the other two basic multilevel circuit topologies for usage in a cosmopolitan power conditioner for the undermentioned grounds:

§ All six stages ( three on each inverter ) can portion a common District of Columbia nexus. Conversely, the cascade inverter requires that each District of Columbia degree be separate, and this is non contributing to a consecutive agreement.

§ The multilevel flying-capacitor convertor besides portions a common District of Columbia nexus ; nevertheless, each stage leg requires several extra subsidiary capacitances. These excess capacitances would add well to the cost and the size of the conditioner.

Because a diode-clamped convertor moving as a cosmopolitan power conditioner will be expected to counterbalance for harmonics and/or operate in low amplitude transition index parts, a more sophisticated, higher-frequency switch control than the cardinal frequence exchanging method will be needed. For this ground, multilevel infinite vector and carrier-based PWM attacks are compared in the following subdivision, every bit good as fresh carrier-based PWM methodological analysiss.

Academy award Lopez at al. , [ ] presented multilevel multiphase infinite vector PWM algorithm which provides a sorted shift vector sequence that minimizes the figure of exchanging. This SVPWM algorithm proves suited for real-time execution due to its low computational complexness.

Nicolau Pereira Filho at al. , [ ] proposed an unreal nervous web based infinite vector PWM for a five-level voltage-fed inverter. The attack uses two ANNs to implement the SVPWM algorithm. One ANN was used for triangle designation, coevals of the corresponding weight matrices for the 2nd ANN, and the coefficient matrices for the PWM moving ridges. The 2nd ANN was used for computation of the responsibility rhythms of the nearest three vectors. The ANN-based SVM execution, peculiarly with ASIC bit, is well simpler than the traditional DSP-based solution.

P.Purkait at al. , [ ] presented a new manner of implementing the infinite vector transition algorithm for cut downing the impersonal point current in the multilevel inverter. From the FFT analysis of line electromotive force and impersonal current it is concluded that 1. Low frequence harmonic content of the impersonal current is nothing. 2. Impersonal point current has a nothing District of Columbia mean value. But with this method, the line electromotive force contains somewhat larger harmonics with regard to nearest three vector transition.

R. S. Kanchan at al. , [ ] presented a electromotive force transition strategy of the SVPWM for multilevel inverters. The focus of the in-between inverter exchanging vectors of the SVPWM is achieved by the add-on of an offset clip signal to the inverter gating signals, derived from the sampled amplitudes of the mention stage electromotive forces. This SVPWM strategy covers the full transition scope, including the over transition part and it does non necessitate any sector designation, as it required in conventional SVPWM strategies.

S. Ali Khajehoddin at al. , [ ] proposed a simple and accurate current flow theoretical account for m-level diode-clamped multilevel convertors. Mugwump of the transition scheme, the theoretical account predicts the new provinces of the convertor based on the mensural values of end product currents, dc-link electromotive forces, and the current shift provinces. It presents a new apprehension of electromotive force sharing handiness among the District of Columbia nexus capacitances, and simplifies the anticipation of the convertor provinces and public presentation.

Using the exchanging map construct, a general, simple, and comprehensive current flow theoretical account for five-level diode-clamped multilevel convertors is derived by S. A. Khajehoddin at al. , [ ] .Which presents a new apprehension of electromotive force sharing handiness among the District of Columbia nexus capacitances. An immediate decision is that the ordinary sinusoidal pulsation breadth transition ( SPWM ) fails to supply a electromotive force equilibrating solution and therefore is inappropriate for exchanging diode-clamped multilevel convertors. Furthermore, an optimized electromotive force equilibrating scheme is proposed. The proposed strategy requires really simple computations to accurately foretell the capacitances electromotive forces which are being used by the quadratic parametric quantity to minimise the divergences of the DC-link electromotive forces.

Giuseppe Carrara at al. , [ ] analyzed the multilevel transitions procedure with a mathematical method which provided the analytical look of the end product stage electromotive forces of the inverter and validated the consequence in over transition operation. They highlighted the betterments in the harmonic contacts due to the increased figure N of degrees.

Zeliang Shu at al. , [ ] developed a compact algorithm of infinite vector pulse breadth transition for three-phase inverter. The conventional SVPWM is decomposed in to fast integer operations wholly by utilizing an intermediate vector, which will decently antagonize the excess computations of the staying processs. This has examined merely in two degree inverter applications.

In this thesis the chief attending is given to Neutral Point Clamped Inverter.


This Thesis consists of six chapters. Chapter 1 gives an debut to multilevel inverters, basicss of SVPWM, organisation of the thesis and chief aim of the thesis. Chapter 2 outlines the SVPWM applied for a two degree inverter, determines the exchanging provinces, exchanging clip continuance and shift sequences for a two degree inverter. In Chapter 3 algorithm for the execution of SVPWM utilizing fractal attack is discussed in item.

The Chapter 4 explains SVPWM for a multilevel inverter utilizing a fractal attack and is implemented for three degree and five degree inverters. In Chapter 5 the experimental probes of the algorithm is implemented in MATLAB/SIMULINK package. Chapter 6 discusses the consequences of the MATLAB /SIMULINK theoretical account. Decisions are given in Chapter7.