STUDI PERSAMAAN KEADAAN BINTANG NEUTRON DENGAN PENDEKATAN DEEP LEARNING
Neutron star is one of the most compact star in universe, just second to the black hole. High density in neutron star especially in its inner core make it possible for matter to take form of quantum particle, like quarks or gluon, with many possible configuration of state. This makes neutron star a...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Subjects: | |
| Online Access: | https://digilib.itb.ac.id/gdl/view/66978 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | Neutron star is one of the most compact star in universe, just second to the black hole. High density in neutron star especially in its inner core make it possible for matter to take form of quantum particle, like quarks or gluon, with many possible configuration of state. This makes neutron star a valuable natural laboratorium not just for stellar astrophysics but in general for particle physics too. However, describing structure of neutron star is not an easy task.
Structure of neutron star can be described using equation of state (EoS) which has form of P=P(?), with P=pressure and ?=density. So far, EoS from various theoretical model agrees with each other for region with density less than saturation density ?0 = 2.7×1017 kg/m3 but doesn’t agree otherwise. For this reason, another approach that doesn’t need initial assumption would be a good alternative for building EoS. One alternative is using deep learning that would be focus in this study.
In this study, we build EoS using neutron star mass and radius data and its error. Training and validation data is synthesized using theoretical model FPS which consist of mass and radius data for input layer and parameter cs=dP/d? for output layer. For prediction data, we will use real data from observation of 14 neutron star in the form of 2D probability of mass and radius.
Result of this study show the same trend for cs value compared to main reference (Yuki, F, dkk 2019), that conduct the same study earlier using theoretecal model Sly4, but with steeper gradient. This reasult yields EoS that doesn’t agree with EoS from main refrence especially for region with higher density. EoS as well as mass-radius curve show trend that agrees with Sly4 (F. Douchin and P. Haensel, 2001) dan ENG (L. Engvik, etc, 1996) theoretical model and our result predicts more massive and compact neutron star. |
|---|