[PPT] - A Review of Predictive Control Techniques for Matrix Converters - PowerPoint Presentation

SLIDE 1

A Review of Predictive Control Techniques for Matrix Converters - Part II

marcoriv@utalca.cl

M. Rivera , P. Wheeler, A. Olloqui

February 17, 2016

SLIDE 2

A Review of Predictive Control Techniques for Matrix Converters - Part II Outline

Outline

1 Predictive Control Strategies and Applications 2 Limitations and/or Weaknesses 3 Open Questions and Future Trends 4 Conclusions

SLIDE 3

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies applied to MCs

Model Predictive Control

Predictive Current Control

Basic strategy Reactive power minimization Filter resonance mitigation Imposed sinusoidal waveform Common-mode voltage reduction Reduction of switching losses

Predictive Torque Control

Basic strategy Basic strategy Reactive power minimization Reactive power minimization

Predictive Voltage Control Predictive Power Control

SLIDE 4

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

Predictive Current Control (PCC) ♣ Only one term in the cost function

g(k+1) = |i∗

a −ip a |+|i∗ b −ip b |+|i∗ c −ip c | (1)

♣ One drawback: variable switching

frequency

♣ Source currents are highly distorted

Predictive Model AC Source vs Filter vi

3 3 3 3 3 3 3

Cost Function Minimization Matrix Converter SAa...SCc ip

(k + 1)

io vo Load

27 Ro Lo i∗

SLIDE 5

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

a) b)

vsA isA i∗

a

ia

Experimental results using a three-phase variac as the AC-source; a) source voltage vsA [V] and current isA [A]; b) output current ia [A] and its reference i∗ a [A].

SLIDE 6

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

PCC with input reactive power minimization ♣ Two terms in the cost function

g(k + 1) = △io + γq△Q (2)

△io = |i∗ a − ip a | + |i∗ b − ip b | + |i∗ c − ip c | △Q = |vsαisβ − vsβ isα| ♣ Input currents in phase with their

respective voltages

♣ Weighting factor selection ♣ The method is very sensitive to source

voltage distortion and input filter resonances

Predictive Model Predictive Model AC Source Filter vi vi vs is io io vo

3 3 3 3 3 3 3 3 3 3 27 27

Cost Function Minimization Matrix Converter SAa...SCc Load Ro Lo i∗

ip
(k + 1)

Qp(k + 1) Q∗

SLIDE 7

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

a) b)

vsA isA i∗

a

ia

Experimental results using a three-phase variac as the AC-source; a) source voltage vsA [V] and current isA [A]; b) output current ia [A] and its reference i∗ a [A].

SLIDE 8

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

PCC with active damping ♣ The same cost function

g(k + 1) = △io + γq△Q (3)

△io = |i∗ a − ip a | + |i∗ b − ip b | + |i∗ c − ip c | △Q = |vsαisβ − vsβ isα| ♣ The main idea: to emulate a

virtual resistor

♣ The harmonic content is

passed from the input to the

utput side

Predictive Model Predictive Model AC Source is vs

27 27

Filter vi vi

3 3 3 3 3 3 3 3 3 3

Cost Function Minimization Matrix Converter SAa...SCc ip

(k + 1)

io io vo Load

PLL

θi abc abc dq dq vdq

i

Digital DC Blocker

vdq

ih 1 R

idq

ih

Qp(k + 1) Q∗ I ∗

dq

i∗

Ro

Lo

SLIDE 9

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

PCC with active damping ♣ The performance of the

system is improved

♣ The power losses are reduced ♣ The price: PLL - Digital dc

blocker implementation

♣ Sensitive to distortions in

source voltages

Predictive Model Predictive Model AC Source is vs

27 27

Filter vi vi

3 3 3 3 3 3 3 3 3 3

Cost Function Minimization Matrix Converter SAa...SCc ip

(k + 1)

io io vo Load

PLL

θi abc abc dq dq vdq

i

Digital DC Blocker

vdq

ih 1 R

idq

ih

Qp(k + 1) Q∗ I ∗

dq

i∗

Ro

Lo

SLIDE 10

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

a) b)

vsA isA i∗

a

ia

Experimental results using a three-phase variac as the AC-source; a) source voltage vsA [V] and current isA [A]; b) output current ia [A] and its reference i∗ a [A].

SLIDE 11

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

PCC with imposed

sinusoidal source currents g(k +1) = △io +γi△is (4) Is = −λVs ±

(λVs)2 − 4λRf RLI ∗2
/η

−2λRf (5) i∗

sA = Is sin(wst + θ)

i∗

sB = Is sin(wst − 2π/3 + θ)

i∗

sC = Is sin(wst + 2π/3 + θ)

(6)

Predictive Model Predictive Model Filter vi vs vs is io vo

3 3 3 3 3 3 3 3 3 3 27 27

Cost Function Minimization Matrix Converter SAa...SCc Load Ro Lo i∗

ip
(k + 1)

ip

s (k + 1)

Q∗ Vs Io δ φ

PLL

Eq. (6)
Eq. (5)

Is

SLIDE 12

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

PCC with imposed

sinusoidal source currents

♣ Sinusoidal input

current regardless the distortion on the input side

♣ Lower (THD) of both

input and load currents

♣ Reducing the resonance

f the input filter

allowing longer lifetime

f the converter

♣ The price: higher

switching frequency

Predictive Model Predictive Model Filter vi vs vs is io vo

3 3 3 3 3 3 3 3 3 3 27 27

Cost Function Minimization Matrix Converter SAa...SCc Load Ro Lo i∗

ip
(k + 1)

ip

s (k + 1)

Q∗ Vs Io δ φ

PLL

Eq. (6)
Eq. (5)

Is

SLIDE 13

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

a) b)

vsA isA i∗

a ia

Experimental results using a three-phase variac as the AC-source; a) source voltage vsA [V] and current isA [A]; b) output current ia and its reference i∗ a [A].

SLIDE 14

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

PCC with Common-Mode

Voltage (CMV) minimization and imposed sinusoidal input waveform

♣ Control of CMV

reduces overvoltage stress to the winding insulation and bearing deterioration

♣ Increasing the lifetime

f electric machines

Predictive Model Predictive Model Predictive Model Filter vi vi vs vs is io vo

3 3 3 3 3 3 3 3 3 3 3 27 27 27

Cost Function Minimization Matrix Converter SAa...SCc Load Ro Lo i∗

ip
(k + 1)

ip

s (k + 1)

Q∗ Vs Io δ φ

PLL

Eq. (6)
Eq. (5)

Is v p

cm(k + 1)

SLIDE 15

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

PCC with Common-Mode

Voltage (CMV) minimization and imposed sinusoidal input waveform

♣ Is possible to include

several control

bjectives in the cost

function g(k+1) = △io+γi△is+γv |vcm(k+1)| (7)

Predictive Model Predictive Model Predictive Model Filter vi vi vs vs is io vo

3 3 3 3 3 3 3 3 3 3 3 27 27 27

Cost Function Minimization Matrix Converter SAa...SCc Load Ro Lo i∗

ip
(k + 1)

ip

s (k + 1)

Q∗ Vs Io δ φ

PLL

Eq. (6)
Eq. (5)

Is v p

cm(k + 1)

SLIDE 16

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

0.21 0.22 0.23 0.24 0.25 0.26 0.27 0.28 0.29

10

10 0.21 0.22 0.23 0.24 0.25 0.26 0.27 0.28 0.29

10

10 0.21 0.22 0.23 0.24 0.25 0.26 0.27 0.28 0.29

100

100 200 300

vsA isA i∗

sA

vcm ia i∗

a

Without CMV reduction

0.21 0.22 0.23 0.24 0.25 0.26 0.27 0.28 0.29

10

10 0.21 0.22 0.23 0.24 0.25 0.26 0.27 0.28 0.29

10

10 0.21 0.22 0.23 0.24 0.25 0.26 0.27 0.28 0.29

100

100 200 300

vsA isA i∗

sA

vcm ia i∗

a

With CMV reduction

SLIDE 17

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

PCC for Torque and Flux

Control with input reactive power minimization

♣ Hybrid combination

between PCC and FOC

♣ No need for

modulation techniques

Predictive Model Predictive Model AC Source Filter vi vi vs is io io vo

3 3 3 3 3 3 3 3 27 27

Cost Function Minimization Matrix Converter SAa...SCc ωr,θr Load IM i∗

ip
(k + 1)

i∗dq

Qp(k + 1)

Q∗ T ∗ Ψ∗ ωr ωr ω∗

r

abc dq FOC

SLIDE 18

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

PCC for Torque and Flux

Control with input reactive power minimization

♣ Simple and

straightforward technique

♣ Phase control of the

input current (capacitive, unity, or inductive) g(k + 1) = △io + γq△Q (8)

Predictive Model Predictive Model AC Source Filter vi vi vs is io io vo

3 3 3 3 3 3 3 3 27 27

Cost Function Minimization Matrix Converter SAa...SCc ωr,θr Load IM i∗

ip
(k + 1)

i∗dq

Qp(k + 1)

Q∗ T ∗ Ψ∗ ωr ωr ω∗

r

abc dq FOC

SLIDE 19

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

PCC for Torque and Flux

Control with input reactive power minimization with increment of efficiency

♣ In the cost function is

included the number of commutations necessary to go from the present switching state to the state under evaluation g(k + 1) = △io + γq△q + γsw n (9)

Predictive Model Predictive Model AC Source Filter vi vi vs is io io vo

3 3 3 3 3 3 3 3 27 27

Cost Function Minimization Matrix Converter SAa...SCc ωr,θr Load IM i∗

ip
(k + 1)

i∗dq

Qp(k + 1)

Q∗ T ∗ Ψ∗ ωr ωr ω∗

r

abc dq FOC

SLIDE 20

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

PCC for Torque and Flux Control

with input reactive power minimization with switching losses reduction

♣ The idea is to predict the

switching losses that every switching state would produce, if applied during the next sampling time g(k+1) = △io+γq△q+γsl

18

i=1

△i(i)

c △v (i) ce

(10)

Predictive Model Predictive Model AC Source Filter vi vi vs is io io vo

3 3 3 3 3 3 3 3 27 27

Cost Function Minimization Matrix Converter SAa...SCc ωr,θr Load IM i∗

ip
(k + 1)

i∗dq

Qp(k + 1)

Q∗ T ∗ Ψ∗ ωr ωr ω∗

r

abc dq FOC

SLIDE 21

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

Predictive Torque and Flux Control

(PTC)

♣ Simple and effective torque

and flux control method of an induction machine

♣ It uses the mathematical

model of the induction machine

♣ A PI controller is used to

generate the torque reference g = △Te(k + 1) + γψ△ψ(k + 1) (11)

Predictive Model AC Source Filter vi vs io vo

3 3 3 3 3 27 27

Cost Function Minimization Matrix Converter SAa...SCc ωr,θr Load IM T ∗ T p(k + 1) Ψ∗ Ψp(k + 1) ωr ωr ω∗

r

Speed Control Flux Calc Ψr Ψs

SLIDE 22

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

PTC with input reactive power

minimization

♣ Possibility to control the

input currents and the output machine variables

♣ No need for modulation

techniques and linear controllers

♣ Fast dynamic response and a

decoupled control g = △Te(k+1)+γψ△ψ(k+1)+γq△qs(k+1) (12)

Predictive Model Predictive Model AC Source Filter vi vi vs is io io vo

3 3 3 3 3 3 3 3 3 27 27 27

Cost Function Minimization Matrix Converter SAa...SCc ωr,θr Load IM T ∗ T p(k + 1) Ψ∗ Ψp(k + 1) Qp(k + 1) Q∗ ωr ωr ω∗

r

Speed Control Flux Calc Ψr Ψs

SLIDE 23

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

vs v1 v1 v2 v2 vi vi is ii i1 i1 i2 i2 Rf Lf Cf 3 3 3 3 3 3 3 3 3 3 3 3 3 18 192 192 192 192 ψ∗

s

ψs1 ψr1 ψs2 ψr2 w ∗ w1 w1 w1 w2 w2 w2 Sr1...Sr6 Si11...Si16 Si21...Si26 T ∗

e1

T ∗

e2

T p1

e

T p2

e

ψp

s1

ψp

s2

Prediction Torque and Flux

Flux Flux Estimator 1 Estimator 2

Switching State Selector IMC IM1 IM2 Induction Machines Input Filter Voltage Source PI1 PI2 References Speed Controller + +

SLIDE 24

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

The cost function used in this scheme is given as: g = g1 + g2 + hp, (13) Errors between the references and predicted values on machine 1 are: g1 = λψ1|ψ∗

s1 − ψp s1| + λT1|T ∗ e1 − T p e1|

(14) Errors between the references and predicted values on machine 2 are: g2 = λψ2|ψ∗

s2 − ψp s2| + λT2|T ∗ e2 − T p e2|

(15)

SLIDE 25

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

100

100 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

50

50 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

20

20 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.5 1

ω∗

1

ωmec1 Te1 T ∗

e1

i1 ψs1

∗

ψs1 a) b) c) d) Time[s]

SLIDE 26

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

100

100 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

50

50 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

20

20 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.5 1

ω∗

2

ωmec2 Te2 T ∗

e2

i2 ψs2

∗

ψs2 a) b) c) d) Time[s]

SLIDE 27

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

0.34 0.35 0.36 0.37 0.38 0.39 0.4

20

20 0.34 0.35 0.36 0.37 0.38 0.39 0.4

20

20 0.34 0.35 0.36 0.37 0.38 0.39 0.4

20

20

vsA vsB vsC isA isB isC a) b) c) Time[s]

SLIDE 28

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

vs vs v1 v1 v2 v2 vi vi is is ii io i1 i1 i2 i2 Rf Lf Cf 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 6 18 192 192 192 192 192 q∗

s

qp

s

ψ∗

s

ψs1 ψr1 ψs2 ψr2 w ∗ w1 w1 w1 w2 w2 w2 Sr1...Sr6 Si11...Si16 Si21...Si26 T ∗

e1

T ∗

e2

T p1

e

T p2

e

ψp

s1

ψp

s2

Reactive Power Prediction Prediction Torque and Flux Flux Flux Estimator 1 Estimator 2 Switching State Selector IMC IM1 IM2 Induction Machines Input Filter Voltage Source PI1 PI2 References Speed Controller + +

SLIDE 29

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

The cost function is defined now as follows: g1 = λψ1|ψ∗

s1 − ψp s1| + λT1|T ∗ e1 − T p e1|,

(16) g2 = λψ2|ψ∗

s2 − ψp s2| + λT2|T ∗ e2 − T p e2|,

(17) g3 = λq|q∗

s − qp s |,

(18) g = g1 + g2 + g3 + hp. (19)

SLIDE 30

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

100

100 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

50

50 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

20

20 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.5 1

ω∗

1

ωmec1 Te1 T ∗

e1

i1 ψs1

∗

ψs1 a) b) c) d) Time[s]

SLIDE 31

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

100

100 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

50

50 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

20

20 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.5 1

ω∗

2

ωmec2 Te2 T ∗

e2

i2 ψs2

∗

ψs2 a) b) c) d) Time[s]

SLIDE 32

A Review of Predictive Control Techniques for Matrix Converters - Part II Predictive Control Strategies and Applications

Predictive Control Strategies

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

20

20 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

20

20 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

20

20 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

20

20

vsA vsB vsC isA isB isC qs a) b) c) d) Time[s]

SLIDE 33

A Review of Predictive Control Techniques for Matrix Converters - Part II Limitations and/or Weaknesses

Limitations and/or Weaknesses

Predictive Control ♣ Variable switching frequency ♣ High dependence on the predictive model quality ♣ High sampling frequency ♣ Sensitive to variation on the parameters ♣ High computational cost

SLIDE 34

A Review of Predictive Control Techniques for Matrix Converters - Part II Open Questions and Future Trends

Open Questions and Future Trends

There are still some open topics for research ♣ The Matrix Converter has not yet proved to be the most appropriate

solution for:

⋆ High-load dynamics ⋆ Single-phase operation capability ⋆ Extended ride-through capability ⋆ Unconstrained reactive power compensation ♣ An exhaustive comparison between predictive control and classical con-

trol techniques must be taking into consideration in order to demon- strate the simplicity and feasibility of the strategy

♣ EMI filters must be considered in future implementations and use in

industrial applications

SLIDE 35

A Review of Predictive Control Techniques for Matrix Converters - Part II Conclusions

Conclusions

Based on the review given by the authors, predictive control has a very high impact in the control of matrix converters, due its simplicity and intuitive approach, making this control strategy a real alternative in power electronics. Simplicity, fast dynamic response, and easy inclusion of constraints Feasibility of the proposed control scheme with a very fast dynamic response and very low THD

SLIDE 36

A Review of Predictive Control Techniques for Matrix Converters - Part II Conclusions

Acknowledgments

This publication was made possible by the Newton Picarte Project EPSRC: EP/N004043/1: New Configurations of Power Converters for Grid Interconnection Systems / CONICYT DPI20140007 and British Council through the Institutional Skills Development Newton Picarte Project ISCL 2015006.