Nonlinear Systems
Nonlinear Systems
Part 4.
Nonlinear Control
Nonlinear Control Systems
Introduction
Describing function method
Phase plane method
3
Introduction
What is a nonlinearity?
What is a nonlinear control system?
Typical nonlinearities.
Characteristics of a nonlinear system
Analysis method for nonlinear systems
4
y(t )
kx(t) x(t) a
y(t) b
ka sign x(t) x(t) a a k x(t )
5
6
ìï0 x(t) £ a a
k
x(t )
y(t) = í
ïîk [ x(t) - asin g x(t)] x(t) > a k 0 a
Typical Nonlinearities
Relay
y(t )
b x(t) ma
0 ma x(t) a x(t) 0 b
y(t ) b x(t) a a ma x(t )
b x(t) ma 0 ma a
0 a x(t) ma x(t) 0
b
b x(t) a
9
10
11
12
H(s)
nonlinear linear
If : x(t) = X sin w t a sinusoidal input,
y(t) may not be a sinusoidal but a periodic function, which
can be expressed as a Fourier series:
¥
y(t ) = A0 + å (An cos nw t + Bn sin nw t )
n =1
¥
= A0 + å Yn sin(nw t + fn )
n =1
13
Nonlinear Control Systems
A1
Y1 A12 B12
0
A
1 arctg 1 2
y( t ) sintd (t )
1
B1 B1
0
14
15
x(t) = X sin w t
Solve y(t) and obtain the fundamental component of y(t)
k 2k é a a a 2ù
N(X) = ê arcsin + 1- ( ) ú
a p ë X X X û
saturation
k
2k é p D D D 2ù
N(X) = ê - arcsin - 1- ( ) ú
dead zone p ë2 X X X û
16
17
M 4M h 2 4M
N(X) = 1- ( ) - j
h
pX X p X2
ép 2h ù
+ arcsin(1 - )
k ê2 X ú
k
N(X) = ê ú
h
pê 2h h h ú
êë +2(1 - X ) X (1 - X ) úû
backlash
hysteresis 4kh h
+j ( - 1)
pX X
18
compare wi th
1 linear system
* G ( j ) doesn' t circle the
N(X ) *G(jw ) doesn't circle the point
curve, the nonlinear system is stable; ( - 1, 0) , the system is stable;
1
* G ( j ) circles the curve,
N(X ) *G(jw ) circles the point ( - 1,0),
the nonlinear system is unstable; the system is unstable;
1
* G ( j ) intersects with the *G(jw ) intersects with the point
N(X )
curve, there will be a self - oscillatio n ( - 1,0), the system is in the
in the nonlinear system . critical stability.
21
Graphical explanation:
Im X increasing Im G ( j ) Im
1 direction
N(X)
Re Re Re
G ( j )
G ( j ) 1 1
N(X ) N(X)
G ( j ) G ( j )
does not circle -1/N(X) G ( j ) circles -1/N(X) Intersects with -1/N(X)
(stable) (unstable) (self-oscillation)
22
X-axis: x (t )
Y-axix: x (t )
x x ( X 2 , X 2 )
Phase plane
X 1 X 2 t x graph
0
( X 3 , X 3 )
Increasing t
x ( X1 , X 1 ) direction
X3
X2
X1 t Always to be
clockwise
27
unstable
focus
stable
nodes
unstable
nodes
28
centers
saddle
points
29
0
0.5 x 0.5
We can make: x x1 x x 2 1 1
2
2
Then, we have:
x
dx2
x2 f ( x1 , x2 )
dx1
(x 0 , x 0 )
30
We can analyze the stability directly from the phase plane: the
phase locus is convergent or divergent.
We can analyze the self-oscillation directly from the phase plane:
the phase locus converges upon a limit circle.
We can transform the phase locus into the time response curve
x(t) to analyze the transient response of the system.
We can analyze the steady state error and overshoot directly from
the phase locus.
The phase plane method is only used for analyzing or designing
1st-order or 2nd-order nonlinear systems.