Professional Documents
Culture Documents
2 - Από Τον Λογισμό Των Μεταβολών Στην Θεωρία Βέλτιστου Ελέγχου
2 - Από Τον Λογισμό Των Μεταβολών Στην Θεωρία Βέλτιστου Ελέγχου
54006
email : karampet@math.auth.gr
. ,
,
.
,
.
Abtsract. Optimal Control Theory has its roots in the development of a
major field of Mathematics: the Calculus of Variations. The main scope of
this work is twofold: a) to trace the development of the calculus of
variations, and b) to refer to the primary target of the optimal control theory.
1.1
.
,
.
(70-19 ..) ,
9 .. ,
,
.
, ,
, ,
.
,
,
, , ,
. (circumdare), ,
(circus),
.
( 1)
( , Cod. Vat. lat. 3225).
:
,
,
.
:
0 x a , y 0 ( 0, 0 ) ( a, 0 )
, y = 0, 0 x a
.
(0,0)
(a,0)
:
J ( y ( x ) ) = y ( x ) dx
a
L ( y ( x )) =
1 + ( y ' ( x ) ) dx =
2
, 19
. J ( y ( x ) )
C 1 [ a, b ] ( )
.
(290-350..)
.
, .
,
(365.. -300.., (13 )), (287..212.. , )
(200..-140.., ), , ,
(180 ..- 120.., (14 )),
(160.. -90.., ) ...
5
( )
( ).
180.. ( ) .
5 :
1. ,
.
2. ,
.
3. .
:
4. ,
.
5.
,
.
6.
, .
7. ,
.
8.
, .
9. ,
.
( 150 . . 300 ..).
,
.
,
.
(
) Fermat (1601-1655).
Fermat
Bernoulli.
:
A ( 0, ya ) B ( xb , yb )
y ( x ) = 0, 0 x xb .
D ( xd , 0 ) ADB
,.
A (0, y a )
B( xb , yb )
xb
D ( xd , 0 )
:
J ( y ( x )) =
xb
1 + ( y ' ( x ) ) dx
2
y ( 0 ) = ya , y ( xd ) = 0, y ( xb ) = yb
y ( x )
x = xd .
A, D D, B , D
sin (1 ) = sin (2 ) .
A (0, y a )
B( xb , yb )
D ( xd , 0 )
2
xb
C
B
Galileo
. Galileo
. 1669 J. Jungius (1587-1657),
.
Jacob Bernoulli 1690,
1691 Acta eruditorum,
Huygens (1629-1695), Leibniz Johann Bernoulli (
Jacob).
(catenary
) ( y ( x ) = ( e ax + e ax ) / ( 2a ) a
). Huygens
(catenary)
Leibniz 1690.
Pierre de Fermat (1601-1665),
(1662). Fermat
.
Gallileo. Fermat
Snell (1580-1626..) ( ).
Ibn Sahl (9401000.) 984...
Thomas Harriot (1560-1621),
Willebrord von Roijen Snel (1580-1626) .
Descartes Snell
. ,
. Huygens 1703,
Fermat,
.
1 sin (1 ) = 2 sin ( 2 )
1
2 .
Fermat :
n , li vi
li
.
i =1 vi
c
i vi , . i = c / vi ( ),
n
, t =
t =
1 n
ili .
c i =1
n
t S = i li .
i =1
S = d
.
.. d = 0 .
Snell
.
f ( x ) .
Newton Principia
,
,
AMB ,
.
:
.
c/2
.
Jacob Bernoulli ( 1690 Acta Eruditorum), Joseph Louis Lagrange
Leonhard Euler.
x
y*(x)
M
B(b,yf)
y(x)
y * ( x ) C1
, .
Johann Bernoulli (1667-1748)
Bernoulli
.
Bernoulli (brachistochrone
problem) (brachistos))
(chronos).
1
ds
2
mu ( t ) = mgy ( x ) u ( t ) = 2 gy ( x ) =
2
dt
u(t) m ,
g=9.81. T s ,
:
T
T = dt =
T
dt
1
1
ds =
1 + ( y ( x)) 2 dx =
0
ds
2 gy ( x)
2g
(1 + y ( x ) )1/ 2
2
y ( x)
1/ 2
dx
y * ( x ) C1 :
1
T ( y ( x )) =
2g
(1 + y ( x ) )1/ 2
2
y ( x)
1/ 2
dx
(cycloid).
: Gottfried Wilhelm Leibnitz
(1646-1716), Newton, Bernoulli Jacob
Bernoulli (1654-1705), Ehrenfried Walter
von Tschirnhaus (1651 1708) De LHospital (1661-1704).
Groningen
Bernoulli
1695 1705.
Johann Bernoulli ,
Snell ( Fermat).
1
2 2
3
3
4
Fermat Johann Bernoulli
.
sin sin r
=
v
vr
dx
sin =
ds
1 dx 1 dxr
=
v ds vr dsr
1 dxr
vr dsr
dx
= cv
ds
ds 2 = dx 2 + dy 2
2
dx 2
dx 2
2
dx
2 2
=
= c2v2
cv
c
v
( )
2
2
2
ds
dx + dy
ds
v = 2 gy
dy 2
1
1
1
2 2
2
=c v
= 2c gy y 1 + = k 2 , k =
2
2
dx
dy
c 2g
dy
1+ 2
1+
dx
dx
.
Jacob Bernoulli
.
,
,
.
A
: )
,
, ) ds / dt y .
.
Jacob Brook Taylor (1715) Johann
Bernoulli.
Leonhard Euler (1707-1783) Johann Bernoulli,
1744, Methodus Inveniendi Lineas Curvas Maximi
Minimive Proprietate Gaudentes sive Solutio Problematis Isoperimetrici
Latissimo Sensu Accepti,
.
J ( y ( x ) ) := F x, y ( x ) , y (1) ( x ) ,..., y ( n ) ( x ) dx
a
:
n
F
F d F d 2 F
n d
(1) + 2 ( 2) ... + ( 1)
=0
n
dx y ( n )
y dx y dx y
Euler-Poisson
2n. Euler
.. Euler.
. Euler
b
J := F x, y ( x ) , y (1) ( x ) , Q dx
a
dQ
= L x, y ( x ) , y (1) ( x )
dx
Lagrange. Euler
,
. Euler 2
On the motion of bodies in a
non-resisting medium, determined by the method of maxima and minima,
,
,
Maupertuis.
Lagrange, Poisson Jacobi,
Hamilton .
Joseph-Louis Lagrange (1736-1813) Euler
, .
Euler Lagrange,
Lagrange. Lagrange
Euler,
. Mecanique Analytique
. :
b
y ( x ) , z ( x ) ,... / /
. Lagrange
y ( x ) , z ( x ) ,...
.
, :
b
()
()
( x, y ( x ) , z ( x ) ,..., y ( x ) , z ( x ) ,...) dx = c , i = 1, 2,.., m
1
Mayer.
Lagrange
x ( t ) , y ( t ) ,...
.
Leonhard Euler(1707-1783)
.
. Jacobi 1838,
Hamilton
Hamilton-Jacobi
Richard Ernest Bellman
(1920-1984) 100 ( 1953).
Karl Wilhelm Theodor Weierstrass (1815-1897)
Oskar Bolza (1857-1942), Gilbert A. Bliss (1876-1951),
(1873-1950) McShane (1904-1989).
,
Morse (Marston Morse (1892-1977)),
(2 Field
Jesse Douglas 1936 Enrico Bombieri ),
( Hamilton,
Richard
Feynman (1918-1988)), .
2
:
J
A,
( ) J.
[ a, b] C[1a ,b]
y ( x ) : [ a, b ] .
J ( y ( x ) ) = f ( x, y ( x ) , y ' ( x ) ) dx
b
1
[ a ,b ]
y ( x ) ( )
J ( y ( x ) ) .
,
.
,
() .
() ,
, :
x ( t ) = a ( x(t ), u (t ), t )
(1)
y (t ) = c ( x(t ), u (t ), t )
y1 (t )
x1 (t )
u1 (t )
y2 (t )
x2 (t )
u2 (t )
x (t ) =
, u (t ) =
,y ( t ) =
xn (t )
um (t )
y p (t )
, .
:
J = h ( x(t f ), t f ) + g ( x ( t ) , u ( t ) , t ) dt
tf
t0
(2)
u * ( t ) , u * ( t )
u ,
(1),
x * ( t ) , x * ( t )
x ,
(2). u * ( t ) (optimal
control), x * ( t )
(optimal trajectory).
,
.
.
J * = h ( x* (t f ), t f ) + g ( x* ( t ) , u* ( t ) , t ) dt h ( x(t f ), t f ) + g ( x ( t ) , u ( t ) , t ) dt
tf
tf
t0
t0
u (t ) u , x ( t ) x
(2),
, ,
.
,
(),
().
.
1 .
() :
d (t ) = a(t ) + b(t )
d ( t ) t , a ( t ) , b ( t )
.
x1 (t ) = d (t ) x1 (t ) = d (t )
x2 (t ) = d (t ) = x1 (t )
:
x1 ( t ) 0 1 x1 ( t ) 0 0 u1 ( t )
=
+
x2 ( t ) 0 0 x2 ( t ) 1 1 u2 ( t )
x1 ( t ) d ( t ) u1 ( t ) a ( t )
=
;
=
x2 ( t ) d ( t ) u2 ( t ) b ( t )
d (t ) = a (t ) + b(t ) x2 (t ) = a (t ) + b(t ) .
( )
t f e ,
:
x1 ( t0 ) d ( t0 ) 0 x1 ( t f ) d ( t f ) e
=
=
=
= ;
x
t
d
t
0
(
)
(
)
2 0 0 x2 ( t f ) d ( t f ) 0
e t f
x ( t ) :
0 x1 ( t ) e,
0 x2 ( t )
,
, M 1 m / sec 2
M 2 m / sec2 . ,
:
0 u1 ( t ) M 1 , M 2 u2 ( t ) 0
..
G .
x2 ( t ) = d ( t ) u1 ( t ) = a ( t ) ,
:
tf
k1 x2 ( t ) + k2u1 ( t ) dt G
:
t0
tf
J = t f t0 = dt
t0
u * ( t ) x * ( t )
,
tf
J = dt .
t0
.
M,
:
b*(t)
a*(t)
M
t0
(1/2)(t0+tf)
tf
-M
t0
(1/2)(t0+tf)
tf
x2*(t)
x1*(t)
t0
(1/2)(t0+tf)
tf
t0
(1/2)(t0+tf) t
f
u1 ( t ) + u2 ( t ) = e (1) ( t ) x1 ( t ) = e, x2 ( t ) = e ( t ) ,
( ),
.
.
(time optimal control problem)
u * ( t ) u
x ( t0 )
x ( t f
) .
:
tf
J = t f t0 = dt
t0
x ( t f
r ( t f ) .
:
n
J = xi ( t f ) ri ( t f ) = x ( t f ) r ( t f )
i =1
2
J = x ( t f ) r ( t f ) H x ( t f ) r ( t f )
H .
T
.
x ( t )
t f ,
r ( t f ) .
(tracking problem)
u * ( t ) u
x ( t )
r ( t ) . :
T
f
f
2
J = [ x(t ) r (t ) ] Q [ x(t ) r (t ) ] dt = x(t ) r (t ) Q dt
t0
t0
Q .
t
.
x ( t )
r ( t ) .
(regulator problem)
.
u * ( t ) u
x ( t )
.
:
tf
2
2
J = x(t f ) + x(t ) Q dt
H
t0
Q ,
.
(minimum control
effort problem)
u * ( t ) u
x ( t0 ) x ( t f )
. :
tf
tf
J = uT ( t ) Ru ( t ) dt = u ( t ) R dt
t0
t0
R .
2
. (, ,
) .
,
u * ( t ) u
x ( t )
r ( t ) , .
:
tf
2
2
J = x(t ) r (t ) Q + u (t ) R dt
t0
Q,R .
.
x ( t )
r ( t ) , .
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
., 1994, , . .
., 2007,
,
.
Anderson, B., & Moore, J., 1971, Linear optimal control. Englewood
CliHs, NJ: Prentice-Hall.
M. Athans and P. Falb, 1966, Optimal Control : An Introduction to
the Theory and its Applications, McGraw Hill Book Company, New
York, NY.
Bryson Jr., A., 1996, Optimal Control1950 to 1985. IEEE Control
Systems Magazine, 6, 2633.
S.Cuomo, 2000,
, .
Goldstine, H., 1981, A history of the calculus of variations from the
17th to the 19th century. New York: Springer.
R.E. Kalman, 1960, Contributions to the Theory of Optimal Control,
Bol. Soc. Mat. Mexicana, Vol.5, pp.102-119.
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]