Professional Documents
Culture Documents
Continous Time Linear Quadratic Regulator
Continous Time Linear Quadratic Regulator
Continous Time Linear Quadratic Regulator
Lecture 4
Continuous time linear quadratic regulator
41
Continuous-time LQR problem
T is time horizon
Q = QT 0, Qf = QTf 0, R = RT > 0 are state cost, final state
cost, and input cost matrices
subject to x(t) = z, x = Ax + Bu
lets take u(t) = w Rm, a constant, over the time interval [t, t + h],
where h > 0 is small
z T Ptz z T Ptz+
T T T T
+h z Qz + w T
Rw + (Az + Bw ) Ptz + z Pt(Az + Bw ) + z Ptz
Pt = AT Pt + PtA PtBR1B T Pt + Q
Pt = AT Pt + PtA PtBR1B T Pt + Q, PT = Qf
(backward in time)
AT P + P A P BR1B T P + Q = 0
N 1
hX 1
J T T
x(kh) Qx(kh) + u(kh) Ru(kh) + x(N h)T Qf x(N h)
2 2
k=0
Kk = (R + B T Pk+1B)1B T Pk+1A
P = Q + AT P + P A P BR1B T P
Kk = (R + B T Pk+1B)1B T Pk+1A
= (hR + h2B T Pk+1B)1hB T Pk+1(I + hA)
R1B T Pkh
as h 0
1 T
x( )T Qx( ) + u( )T Ru( ) d + 21 x(T )T Qf x(T )
R
minimize J = 2 0
subject to x(t) = Ax(t) + Bu(t), t [0, T ]
(note: you need distribution theory to really make sense of the derivatives here . . . )
Z T Z T
( )T x( ) d = (T )T x(T ) (0)T x(0) ( )T x( ) d
0 0
T
d x(t) A BR B1
x(t)
=
dt (t) Q AT (t)
Pt = AT Pt + PtA PtBR1B T Pt + Q, PT = Qf
d
= (P x) = P x + P x
dt
= (Q + AT P + P A P BR1B T P )x + P (Ax BR1B T )
= Qx AT P x + P BR1B T P x P BR1B T P x
= Qx AT
P = AT P + P A P BR1B T P + Q
T
d x(t) 1
A BR B x(t)
=
dt (t) Q AT (t)
T
d X(t) 1
A BR B X(t)
=
dt Y (t) Q AT Y (t)
Z = AT Z + ZA ZBR1B T Z + Q
d
(F (t)G(t)) = F (t)G(t) + F (t)G(t)
dt
d 1
F (t) = F (t)1F (t)F (t)1
dt
subject to x(0) = z, x = Ax + Bu
HJ equation is ARE
Q + AT P + P A P BR1B T P = 0
x = Ax + Bu = (A + BK)x
T T
A BR B 1
I A BR B P 1
A + BK
= =
Q AT P Q AT P Q AT P
and so
T T
I 0 A BR B 1
I 0 A + BK BR B1
=
P I Q AT P I 0 (A + BK)T
note that 1
I 0 I 0
=
P I P I
T
T 1
0 A + BK
BR B 1
T 0
0 TT (A + BK)T
0 0 T T
1 T T
T BR B T
1
=
0
and so
T T
H =
PT PT
T
thus, the n columns of are the eigenvectors of H associated with
PT
the stable eigenvalues 1, . . . , n
(this is very close to the method used in practice, which does not require
A + BK to be diagonalizable)