Block Diagrams & Signal Flow Graphs Lectures 5 & 6: M.R. Azimi, Professor
Block Diagrams & Signal Flow Graphs Lectures 5 & 6: M.R. Azimi, Professor
Fall 2016
Then
r1 N1 θ1 T1
= = = =⇒ T1 θ1 = T2 θ2
r2 N2 θ2 T2
where N : Number of teeth on the gear; T1 (t): Applied torque, T2 (t): Delivered
torque; and θ(t): angular displacement
M.R. Azimi Control Systems
Rotational Mechanical Systems Block Diagrams Signal Flow Graph Method
Gear Trains-Cont.
Example:
2
J d dt
θ2 (t)
2 + D dθdt
2 (t)
+ Kθ2 = T2 (t) =⇒ (Js2 + Ds + K)Θ2 (s) = T2 (s)
Solve for Θ2 (s) and then find Θ1 (s) using the gear train equation.
3 Multiplier
Using superposition
C1 (s) = G11 (s)R1 (s) + G12 (s)R2 (s) + · · · + G1M (s)RM (s)
C2 (s) = G21 (s)R1 (s) + G22 (s)R2 (s) + · · · + G2M (s)RM (s)
..
.
CN (s) = GN 1 (s)R1 (s) + GN 2 (s)R2 (s) + · · · + GN M (s)RM (s)
To find Gij (s) set all inputs except Rj to zero and look at ith output i.e.
Ci (s)
Gij (s) = R j (s)
when all other Rk (s) = 0, k 6= j.
RM (s)
Output vector:
C1 (s)
C(s) = ...
CN (s)
Then
C(s) = Ĝ(s)R(s)
Notation: An underbar represents a vector and hat indicates a matrix.
M.R. Azimi Control Systems
Rotational Mechanical Systems Block Diagrams Signal Flow Graph Method
Though the block diagram approach is commonly used for simple systems, it
quickly gets complicated when there are multiple loops and subsystems or in
MIMO cases. Thus, we need a more streamlined and systematic approach for
such systems.
Signal Flow Graph (SFG): Pictorial representation of a system of equations, in
which:
variables → nodes of SFG
relationship between variables → branches of SFG
coefficients → gains of branches in SFG.
Example: F = M a
Key Definitions:
1 Input Node: Node with only outgoing branches;
2 Output Node: Node with incoming branches.
Note: Any non-input node can be made an output node by adding a
branch with gain= 1.
3 Path: Collection of branches linked together in same direction.
4 Forward Path: Path from input node to output node where node is visited
more than once.
5 Gain of Forward Path: Product of all gains of branches in the forward path.
6 Loop: Path that originates and terminates at the same node. No other
node is visited more than once.
7 Loop Gain: Product of branch gains in a loop.
8 Non-Touching: Two parts of a SFG are non-touching if they do not share
at least one node.
Example:
Input Node: x1
Output Node: All nodes besides x1 .
Forward Path: Assume x5 as output node, then Path 1 = x1 , x2 , x3 , x4 , x5 ;
Path 2 = x1 , x2 , x4 , x5
Gain of Forward Path: Path 1: M1 = abcd; Path 2: M2 = af d
Loop: x3 → x4 → x5
Loop Gain: P1 = −ce
Algebra of SFG:
1 Output variable of a node = weighted sum (by the gains of branches) of all
incoming branches.
x2 = ax1 , x3 = bx2 − ex4 , x4 = cx3 + f x2 , x5 = dx4
2 Parallel branches
Note: all branches must be in same direction (otherwise they form a loop).
3 Series branches
Steps:
1 Arrange SFG (from Block Diagram) and identify input and output nodes.
2 List all forward paths and gains Mk s.
3 List all loops and gains, Pm1 , Pm2 , . . . and form ∆.
4 Determine state of path k with loops and form ∆k s.
5 Apply Mason Gain Formula.
Example 1:
Consider a standard closed-loop system as shown.
R - input
C - output
M1 = G
P11 = −GH
∆ = 1 − P11 = 1 + GH
∆1 = 1 since path 1 is touching with the loop.
M 1 ∆1 G(s)
M (s) = ∆ = 1+G(s)H(s)
Example 2:
Step 1:
Input: R1 , output: C1
C1 (s)
G11 (s) = R 1 (s)
R2 (s)=0
Step 2:
M1 = G1
Step 3:
Loop gains: P11 = −G1 G2 G3 G4
Thus ∆ = 1 − P11 = 1 + G1 G2 G3 G4
Step 4:
∆1 = 1 since forward path is touching with the loop.
Step 5:
M1 ∆1 G1
G11 (s) = ∆ = 1+G1 G2 G3 G4
Step 1:
Input: R2 , output: C1
C1 (s)
G12 (s) = R 2 (s)
R1 =0
Step 2:
M12 = G1 G3 G4
Step 3:
∆ does not change.
Step 4:
∆1 = 1
Step 5:
M 1 ∆1 G1 G3 G4
G12 (s) = ∆ = 1+G1 G2 G3 G4
Input: R1 , output:
C2
C2 (s) −G1 G2 G3
G21 (s) = R1 (s) = 1+G1 G2 G3 G4
R2 (s)=0
Input: R2 , output:
C2
C2 (s) G3
G22 (s) = R2 (s) = 1+G1 G2 G3 G4
R1 (s)=0