Variasi perilaku NPC di combat defense menggunakan metode Hierarchical Finite State Machine (HFSM)

Fauzi, Rahmat (2015) Variasi perilaku NPC di combat defense menggunakan metode Hierarchical Finite State Machine (HFSM). Masters thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 2213205027-Dissertation.pdf]
Preview
Text
2213205027-Dissertation.pdf - Accepted Version

Download (4MB) | Preview

Abstract

Game RTS merupakan game yang paling diburu untuk dimainkan.
Pada Game RTS, Pada umumnya NPC Bangunan Pertahanan (Tower)
memiliki perilaku menyerang terdekat (Attack nearest) tanpa melihat
parameter musuh. Hal ini menyebabkan NPC Tower lebih mudah ditebak
oleh lawan dan mudah dikalahkan jika diserang oleh musuh secara
berkelompok. Penelitian ini membuat simulasi NPC Tower dengan
menggunakan metode Hierarchical Finite State Machine. Dalam penelitian
ini, NPC Tower mendeteksi musuh dengan melihat empat parameter yakni
Kesehatan NPC Tower, Kesehatan Musuh, Jenis Musuh, dan Jarak Tower
terhadap musuh. NPC Tower menyerang musuh yang paling berbahaya
sesuai parameter Degree of Danger. Kemudian digunakan logika
pengambilan keputusan Sistem rule based. Adapun Output perilaku NPC
Tower adalah perilaku Menyerang Agresif, menyerang biasa, dan
menyerang dengan skill khusus.
Dari hasil pengujian terhadap 3 NPC Tower, NPC Tower Kamandaka
dengan metode HFSM unggul 8,92 % dari Tower Kamandaka dengan
metode FSM. Untuk NPC Tower Gayatri didapatkan hasil yang berimbang
baik menggunakan HFSM maupun FSM. Sedangkan NPC Adikara dengan
metode HFSM unggul 4,62 % dari Tower Adikara dengan metode FSM.

================================================================================================================

RTS game is one of the famous game in the world. Defensive
Building in RTS game usually only have one behavior which is attacking the
nearest enemy. It make the NPC tower very vulnerable and easy to destroy
when attacked by a group of enemy. Therefore in this research the behavior
of NPC tower created using Hierarchical Finite State Machine to make NPC
Tower more powerful. In this research NPC tower decide the target enemy
by looking four parameters, which is Tower Health, Enemy Health, the type
of the enemy, and the distance between tower and enemy. NPC Tower will
attack the most dangerous enemy in group. The most dangerous enemy
detected using degree of danger parameter, the highest one is the most
dangerous. After tower get the target, the behavior of tower will decided
using Rule Based System. The Output behavior are aggressive attack, normal
attack, and attack with special skill.
The experiment with 3 NPC tower called kamandaka show that
HFSM is better 8.92 % from FSM, when HFSM and FSM used in NPC
Tower Gayatri, both methods show the same result. The last experiment
using NPC tower Adikara show that HFSM is better 4.62 % than FSM.

Item Type: Thesis (Masters)
Additional Information: RTE 794.815 26 Fau v
Uncontrolled Keywords: Agen; NPC; Combat Defense; artificial intelligent; HFSM
Subjects: G Geography. Anthropology. Recreation > GV Recreation Leisure > GV1469.2 Computer games
Divisions: Faculty of Electrical Technology > Electrical Engineering > 20101-(S2) Master Thesis
Depositing User: - Taufiq Rahmanu
Date Deposited: 27 Jun 2019 03:53
Last Modified: 07 Jan 2020 08:38
URI: http://repository.its.ac.id/id/eprint/63313

Actions (login required)

View Item View Item