A software tool for genetic optimization of multilayer mirrors

bigskymanΤεχνίτη Νοημοσύνη και Ρομποτική

24 Οκτ 2013 (πριν από 4 χρόνια και 16 μέρες)

112 εμφανίσεις

CM



1

Abstract


This paper presents
a software tool that was
developed to be used for a very fast optimization of
soft x
-
ray
and extreme
-

ultra
-

violet (EUV) multilayer mirrors.
The tool
it is based on u
sing genetic algorithms
and
provi
des an efficient
and accurate way of obtaining the absolute maximum
reflectance of a non
-
transparent system.
The user has the
possibility to use a variety of genetic operators and to tune the
various parameters of a standard genetic algorithm.


Keywords


m
ultilayers, optical coatings, optical properties,
genetic algorithms,
computer simulation

I.

M
ULTILAYER
M
IRRORS

Multilayer mirrors for the soft X
-
ray/extreme ultraviolet
(EUV) spectral domain are key components for the
development of various new optical appli
cations, in various
fields, such as: astronomy (in particular for solar
investigations), microscopy, lasers,
d
eep UV
lithography

for
microelectronic applications

[1
-
3]. These multilayer

mirrors

should provide not only an enhanced reflectance at normal
inci
dence, but also should be characterised by other useful
properties such as: high spectral and spatial resolution, filter
action and polarising power, low optical aberrations. The
achievement of such good properties represents both a
technological and a des
ign problem
.
In these circumstances
the "computer aided design" experiments are vital and, due to
the complexity of the problem (many parameters and
restrictions which must be accomplished), new design methods

and optimisation techniques
, s
uch as "Genetic
Algorithms”
must be used

[4], [5]
.

This paper presents an original
software
tool

(based on the
use of genetic algorithms), which was specifically developed
for the design and optimization of multilayer mirrors.

II.

S
OFTWARE TOOL

The software tool that was dev
eloped is based on using
the power of
genetic algorithms for the design and optimization

of multilayer mirrors. Two optimization criteria can be taken
into account

with variable weights
: mirror’s reflectance and
optical thickness. The user can
select

the n
umber of layers and
the optic
al constants for each material, as well as choosing
from
4 cro
ssover and 2 mutation operators; in addition, a
ll the
parameters can be
also
tuned manually in order to achieve
better performances.
A screenshot is presented in Fig
. 1.




Fig. 1.
Software optimization tool

III.

C
ONCLUSION

The design of complex optical systems (such as multilayer
mirrors) represents a challenge both from a theoretical point of
view and the computational one. The complexity of the
problems can increase
dramatically especially when not only
an optical analysis is needed, but also the thermal
and

mechanical effects must be included.
The
p
resent paper
proposes a software tool that can be used in genetic
algorithms based computer
-
aided design of multilayer m
irrors.
The obtained results demonstrate the power of genetic
algorithms as optimization tools and provide a complete and
powerful software tool for the genetic optimization of
multilayer mirrors.

A
CKNOWLEDGEMENT

We acknowledge the support of Alexandru Poi
ana.

IV.

R
EFERENCES

[1]

]
R. B. Hoover (ed.), "X
-

Ray / EUV Optics for Astronomy and
Microscopy", Proceed. SPIE 1160 (1989
).

[2]

R
.

G. Ohl, T
.
T. Saha, S
.
D. Friedman, R
.

H. Barkhouser, and

H. W
.
Moos
,
“Imaging performance of telescope mirrors for

far
-
ultraviolet ast
ronomy”, Appl. Optics vol. 30 (25), pp. 4513


4523, 2000.

[3]

K. Jain, M. Zemel, and M. Klosner, “
L
arge Area, High Resolution
Lithography and Photoablation Systems for Microelectronics and
Optoelectronics Fabrica
tion”, Proc. IEEE, vol. 90 (10)
, pp. 1681


16
89, 2002.

[4]

I. Dumitrache, C. Buiu,
Introduction to Genetic Algorithms
,
Politehnica Press, 1995

[5]

O. Buiu,
C. Buiu, “
Optimizing multi
-
layer mirrors by using genetic
algorithms

, Proc
.

of CSCS12, Vol. II, pp. 398
-
401, Bucharest,
Romania,
1999

A software tool for
genetic optimization of multilayer mirrors




Octavian Buiu, Cătălin Buiu

University of Liverpool, Department of Electrical Engineering and Electronics

, “Politehnica” University of
Bucharest

Brownlow Hill, L69 3GJ, Liverpool, United Ki
ngdom


obuiu@liv.ac.uk