Numerical analysis of time discretization of optimal control problems
Numerical analysis of time discretization of optimal control problems
Numerical analysis of time discretization of optimal control problems
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
Symplectic schemes<br />
<br />
0 I<br />
Consider the 2n×2n matrix J := .<br />
−I 0<br />
Given H smooth: Rn ×Rn → R, the associated Hamiltonian system is<br />
can be written as (note that J −1 = −J)<br />
˙p = −Hq(p,q); ˙q = Hp(p,q) (4)<br />
d<br />
dt<br />
p<br />
q<br />
<br />
= J −1 DH(p,q), (5)<br />
and the variational equation (linearization) may be written as<br />
d<br />
dt<br />
Zp<br />
Zq<br />
<br />
= J −1 D 2 H(p,q)<br />
Zp<br />
Zq<br />
<br />
. (6)<br />
47