Scribd
Upload
7 views27 pages

Control 4 SFG Mason

Uploaded by

John Valenzona
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
7 views27 pages

Control 4 SFG Mason

Uploaded by

John Valenzona
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 27

SFG and Mason’s Rule : A revision

Andersen Ang
2016-Nov-29

SFG and Mason’s Vu Pham


Review

SFG: Signal-Flow Graph

SFG is a directed graph

SFG is used to model signal flow in a system


SFG can be used to derive the transfer function
of the system by Mason’s Rule / Mason’s Gain
Formula.

SFG and Mason’s Vu 2


SFG terminologies
Node Forward path gain
Edge
Gain
Input / Sources
Output / Sinks
Path
Path gain
Forward
path
SFG and Mason’s Vu 3
L oop
o Loop Gain
o Non-touching loop
p

S
e
l
f

l
SFG and Mason’s Vu 4
SFG terminologies
Node: variables. e. g. 𝑋%, 𝑋' , 𝑋(, 𝑋)
Edge: directed branches. e. g. 𝑋% 𝑋'
Gain: transmission of that branch. e. g. 𝐴'%
Input/Sources: nodes with out-going branches only e. g. 𝑋
%

Output/Sinks: nodes with incoming branches only e. g. 𝑋)

SFG and Mason’s Rule Vu 4


SFG terminologies

e. g. 𝑋% 𝑋' 𝑋)
Path: successive branches without repeated nodes,

Forward path: path from input to output, e. g. 𝑋


Path gain: the gain of the path

% 𝑋' 𝑋( 𝑋) or
𝑋% 𝑋' 𝑋)
Forward path gain: gain of forward path

SFG and Mason’s Rule Vu 5


SFG terminologies

terminates on the same node, e. g. 𝑋' 𝑋( 𝑋'


Loop / Feedback path: a closed path which originates and

Self loop: loop with only one branch, e.g. 𝑋(𝑋(


Loop gain: gain along the loop
Non-touching loop: two loops are non-touching if they do
not share any nodes nor branches

SFG and Mason’s Rule Vu 6


Some SFG simplifications

Branch in parallel:

Branch in series:
Isolated loop:

SFG and Mason’s Vu 7


SFG Example 1

Two forward path: G1G2G3G4G5G7 & G1G2G3G4G6G7

Four loops: G2H1, G4H2, G4G5H3, G4G6H3

Non-touching loops: G2H1 & G4H2, G2H1 & G4G5H3, G2H1 & G4G6H3

SFG and Mason’s Vu 8


SFG Example 2

Three forward path:

A32A43A54A65A76

A72

A42A54A65A74

SFG and Mason’s Vu 9


SFG Example 2
Loops

A32A33, A43A34, A54A45, A65A56, A76A67, A77


A42A34A23, A65A76A57
* A65A76A67A56 is not a loop since X6 is repeated on the path!
A72A57A45 A34A23
A72A67A56A45 A34A23
SFG and Mason’s Vu 1
Mason’s Rule
• How to derive transfer function: block diagram (BD)
reduction or signal flow graph reduction.

• BD approach requires successive application of


fundamental relationships in order to derive transfer
function.

• SFG just applies one formula – Mason’s Rule

SFG and Mason’s Vu 1


Mason’s Rule
#789:;9< >;?@
𝑇𝐹(𝑠) = 1 3 𝑃55Δ
Δ
Equation
5A%

Pi = the i th forward-path gain.


∆ = Determinant of the system
∆i = Determinant of the ith forward path
∆ = 1 - (sum of all individual loop gains)
+ (sum of products of gains of all 2 loops that do not touch each other)
– (sum of products of gains of all possible three loops that do not touch
each other) + …
∆i = Δ for part of SFG that does not touch i-th forward path
Δi = 1 if no non-touching loops to the i-th path, or if taking out i-th
path breaks all the loops
SFG and Mason’s Vu 1
Example 1
X2 X5
X1X3X4X6

Input: X1

Output: X6

Two Forward paths: paths along X1X2X3X4X5X6

Two Forward path gains: 𝑃% = 𝐺%𝐺' 𝐺) and 𝑃' = 𝐺%𝐺( 𝐺)

Loops: X2X3X2 and two loops along X2X3X4X5X2

SFG and Mason’s Vu 1


Loops gains: 𝐿% = 𝐺% 𝐺) 𝐻% , 𝐿' = −𝐺% 𝐺' 𝐺) 𝐻' and 𝐿( =
−𝐺% 𝐺( 𝐺) 𝐻'

SFG and Mason’s Vu 1


Example 1
X2 X5
X1X3X4X6

Δ = 1 - (sum of all individual loop gains) + (sum of


products of gains of all 2 loops that do not touch each
other) – (sum of products of gains of all possible three
loops that do not touch each other)
Since no non-touching loop, thus
Δ = 1 - (sum of all individual loop gains)
Δ = 1 − (𝐿% + 𝐿' + 𝐿()
SFG and Mason’s Vu 1
Example 1
X2 X5
X1X3X4X6

∆i = Δ for part of SFG that does not touch i-th forward


path (Δi = 1 if no non-touching loops to the i-th path)
Since no non-touching loop, thus Δ1 = Δ2 = 1

SFG and Mason’s Vu 1


Example 1
X2 X5
X1X3X4X6

#789:;9< >;?@
1 3𝑃5Δ5
TF
Δ 5A%

1
= Δ 𝑃%Δ% + 𝑃'Δ%
1 𝑃% + 𝑃'
= 1 − (𝐿% + 𝐿' + 𝐿 ()

𝐺% 𝐺' 𝐺) + 𝐺% 𝐺( 𝐺)
= 1 − 𝐺%𝐺)𝐻% + 𝐺%𝐺'𝐺)𝐻' +𝐺%

SFG and Mason’s Vu 1


𝐺(𝐺)𝐻'

SFG and Mason’s Vu 1


Example 2
Input X1, Output X8
X5 X7
Two Forward paths:
X6
X1X2X3X4X8 and X1X5X6X7X8 X1 X8
X3
Two Forward paths gain:
𝑃 % = 𝐺% 𝐺I 𝐺J 𝐺K 𝑃 ' = 𝐺% 𝐺' 𝐺( 𝐺)
X2 X4

SFG and Mason’s Vu 1


Example 2
Loops: X2X3X2, X3X4X3,

X 5 X 6 X 5 , X 6X 7X 6 X5 X7
X6
X1 X8
Individual Loop gains: X3
𝐿%% = 𝐺I𝐻I 𝐿%' = 𝐺J 𝐻J , X2 X4

𝐿%( = 𝐺' 𝐻' 𝐿%) = 𝐺(𝐻(


Two non-touching loops:

𝐿'% = 𝐿%%𝐿%( 𝐿'' = 𝐿%%𝐿%) 𝐿'( = 𝐿%'𝐿%(


𝐿') = 𝐿%'𝐿%)

SFG and Mason’s Vu 2


Three non-touching loops: no

Δ = 1 𝐿%% + 𝐿%' + 𝐿%( + 𝐿+%)(𝐿'% + 𝐿'' + 𝐿'( + 𝐿'))


−

SFG and Mason’s Vu 2


Example 2
∆i = Δ for part of SFG that
does not touch i-th forward X 5 X7
X
path (Δi = 1 if no non- X1 X8
touching loops to the i-th X3
path.) X2 X4

Δ' = Δ for part of SFG that does


not touch 2nd forward path

Δ% = Δ for part of SFG that


does not touch 1st forward
path
SFG and Mason’s Rule Vu 19
Example 2

SFGs without the i-th forward path:


X8
X1

X1 X8

For these SFGs, compute their Δ as:


Δ = 1 - (sum of all individual loop gains) + (sum of products
of gains of all 2 loops that do not touch each other) – (sum of
products of gains of all possible three loops that do not touch
each other)
SFG and Mason’s Vu 2
Example 2
Both SFGs do not have two/more non-touching
loops

X8 X5 X7
X1

X2 X1 X8
X4

Δ' = 1 − 𝐿%% + 𝐿%' Δ% = 1 − (𝐿%( + 𝐿%))


= 1 − 𝐺I 𝐻 I − = 1 − 𝐺' 𝐻' − 𝐺(𝐻(
𝐺J 𝐻 J

SFG and Mason’s Vu 2


Example 2

#789:;9<
X5 X7
1 >;?@ X6

TF 3 𝑃5Δ5
= Δ
X1 X8
5A%
X3
X2 X4

𝑃 %Δ % + 𝑃 ' Δ '
= 1 −𝐿%% + 𝐿%' + 𝐿%( + 𝐿+%)(𝐿'% + 𝐿'' + 𝐿'( + 𝐿'))

𝐺%𝐺I𝐺J𝐺K 1 − 𝐺' 𝐻' − 𝐺(𝐻(+ 𝐺%𝐺'𝐺(𝐺)(1 − 𝐺I𝐻I − 𝐺J𝐻J)


=
1 −𝐺'𝐻' + 𝐺(𝐻( + 𝐺I𝐻I + 𝐺J𝐻J+ (𝐺'𝐻'𝐺I𝐻I + 𝐺'𝐻'𝐺J𝐻J + 𝐺(𝐻(𝐺I𝐻I + 𝐺(𝐻(𝐺J

END
SFG and Mason’s Vu 2

You might also like