Genetic Selection of Neutron Star Structure Matching the X-Ray Observations - PowerPoint PPT Presentation

1 / 32
About This Presentation
Title:

Genetic Selection of Neutron Star Structure Matching the X-Ray Observations

Description:

For orientation of optimization process need fitness function ... Life time (Death) parametr set to 5. uniform crossing. Gene type 3/3/5, shade zone. Gray code ... – PowerPoint PPT presentation

Number of Views:30
Avg rating:3.0/5.0
Slides: 33
Provided by: cermk
Category:

less

Transcript and Presenter's Notes

Title: Genetic Selection of Neutron Star Structure Matching the X-Ray Observations


1
Genetic Selectionof Neutron Star Structure
Matching the X-Ray Observations
WMSCI 2007
  • Speaker
  • Petr Cermak
  • The Institute of Computer Science
  • Silesian University in Opava,
  • Czech Republic

2
Zdenek STUCHLÍK, Petr CERMÁK, Gabriel TÖRÖK,
Martin URBANEC, Pavel BAKALA Institute of
Physics, Faculty of Philosophy and
Science, Silesian University in Opava, CZECH
REPUBLIC Institute of Computer Science, Faculty
of Philosophy and Science, Silesian University in
Opava, CZECH REPUBLIC
Figs on this page nasa.gov
3
Presentation Topics
WMSCI 2007
  • Motivation
  • Neutron Star Structure modeling
  • Matching X-Ray Observation, QPOs
  • Genetic Algorithms
  • Advanced Genetic Algorithms
  • Parallelization of GA using MPI
  • GA Implementation
  • Experimental Results
  • Conclusion

4
Neutron star structure
WMSCI 2007
atmosphere
surface
outer crust
inner crust
core
outer core
inner core
5
Model of neutron star - assumptions
WMSCI 2007
  • Matter is described as perfect fluid
  • Perfect fluid is described by pressure and energy
    density
  • Relation between pressure and energy density is
    given by EOS
  • Axially symetric and stationary spacetime
  • Given by Hartle-Thorne metric

6
Model of NS Structure equations(perturbative
scheme)
WMSCI 2007
Schwarzschild metric
TOV eq. general relativistic equation of
hydrostatic equilibrium
where m(r) is mass inside the sphere of radius r
7
Model of NS EOS
WMSCI 2007
8
Binary systems neutron stars
WMSCI 2007
JET
gigant star
acrretion disc
neutron star
Figs nasa.gov
9
Simulation of binary system behavior
WMSCI 2007
OK, lets take starship Enterprise move with Warp
9 near to binary system
radio
X-ray
and visible
Figs, Videos nasa.gov
10
X-ray observation
WMSCI 2007
Light curve
I
t
Power density spectra (PDS)
Power
Frequency
Figs on this page nasa.gov
11
Power density spectra QPOs
WMSCI 2007
power
low-frequency QPOs
hi-frequency QPOs or kHz QPOs
frequency
12
Observed frequency relations
WMSCI 2007
13
General belief dominating in the astrophysical
community links the kHz QPOs to the orbital
motion near the inner edge of an accretion disc.
WMSCI 2007
Figs on this page nasa.gov
14
Orbital motion in strong gravidity
WMSCI 2007
15
Testing frequency relations 3 models
WMSCI 2007
The three investigated models relate the observed
frequencies to
1. The precession frequency relation involved in
the relativistic precession model
Related the kHz QPOs to the Keplerian and
periastron precession at an inner-disc orbit.
2.
correspond to the so called vertical precession
resonance ( Bursa 2005 )
3. Identify the lower QPO frequency with the
total precession frequency
Introduced in Stuchlik et al. 2007, STB, 2007
Notice that in the limit of Swarzschild spacetime
all the three relation coincide.
16
Neutron star structure modelingusing X-Ray
observations - relations
WMSCI 2007
The fundamental frequencies itself depend on the
parameters (M,j,q) of the central neutron star,
Outer description
which are directly connected to the neutron star
structure
Inner description
Therefore comparison of our models with the
observed frequency relations
Matching
17
Neutron star structure modelingusing X-Ray
observations
WMSCI 2007
WMSCI 2007
18
Optimization - GA
WMSCI 2007
19
Features of GA
WMSCI 2007
  • Wide range limits of utilization than classical
    optimization methods.
  • Operate with global structure chromosome
  • For orientation of optimization process need
    fitness function
  • Addition information may decline time markedly
  • GA uses statistical trans. rules for control of
    searching procedures

20
Genetic Selection of Neutron Star Structure
WMSCI 2007
WMSCI 2007
21
Genetic algorithm Mutation Crossover
WMSCI 2007
22
Genetic algorithm Sexual reproduction Death
WMSCI 2007
23
Genetic algorithm Shade zone
WMSCI 2007
Gen type 3/3/5
24
Genetic algorithm Note to Fitness
WMSCI 2007
If parameters exceed intervals by given min max
values ? Fitness 1e200
25
Population diversity x Rate of convergence x
WMSCI 2007
26
Parallelization of GA using MPI
WMSCI 2007
27
Specific for this experiment
WMSCI 2007
  • Parameters of GA
  • nEOS(4 bits), ?c(12 bits), ?(14 bits)
  • 301 989 888 combination
  • Sexual reproduction
  • Life time (Death) parametr set to 5
  • uniform crossing
  • Gene type 3/3/5, shade zone
  • Gray code
  • Implementation
  • Code writen in c (NET, gcc)
  • MPI-2

28
dependency of on number of members
ingeneration
WMSCI 2007
29
Fitness function and correspondingparameters
with preseted EOS for source 4U1636-53
WMSCI 2007
30
Fitness function and correspondingparameters for
all tested EOSfor source 4U1636-53
WMSCI 2007
31
Computing times
WMSCI 2007
32
Future
WMSCI 2007
  • Include q parameters into computation
  • Include others parameters describing magnetic
    fields (min additional 3 parameters)
  • Portal for astrophysical online computing
  • Solving problem with random generator on
    different platform and SW
  • Distribution parallel over all available platform.
Write a Comment
User Comments (0)
About PowerShow.com