Implementasi Algoritma Collision Detection dan Markov Chain untuk Menentukan Behaviour NPC dan Karakter Player pada Game Higeia

Implementation of Collision Detection and Markov Chain Algorithms to Determine NPC Behavior and Player Characters in Higeia Games

Adventure Game menjadi salah satu genre game terbanyak yang dimainkan sejak video game dikembangkan. Di dalam adventure game player dilatih untuk berfikir menguraikan tempat secara visual, menyelesaikan permasalahan, dan juga merangkai kesimpulan dari peristiwa berdasarkan percakapan maupun benda-benda yang terdapat di dalam scene game. Penelitian ini bertujuan untuk membuat sebuah game dalam bentuk adventure game bertema kesehatan menggunakan Unity dengan nama game “Higeia”. Game ini dibuat menggunakan gabungan metode dari Markov Chain dan Collision Detection dimana metode Markov Chain digunakan sebagai klasifikasi perilaku penyerangan karakter NPC sedangkan metode Collision Detection digunakan untuk memberikan keputusan perilaku ketika karakter player menabrak obstacle di dalam gameplay. Dimana perilaku NPC akan dibagi menjadi tiga perilaku yaitu perilaku patrol, kejar dan serang. Variabel yang digunakan ialah kecepatan (Kc), jarak (Jr), dan nyawa (Ny) dari kondisi karakter player.  Dari hasil pengujian menggunakan perhitungan metode markov chain dengan masukan 15 data nilai awal kondisi player didapatkan hasil perilaku NPC 40% NPC melakukan patrol, 13,33% NPC melakukan kejar dan 46,67% NPC melakukan aksi tembak dengan event collision detection yang berfungsi sesuai dengan yang diharapkan.

Kata Kunci— Game, Collision Detection, Markov Chain, NPC, Artificial Intelligence

Adventure Game has become one of the most widely played game genres since video games were developed. In the adventure game players are trained to think visually describe places, solve problems, and also draw conclusions from events based on conversations and objects contained in the game scene. This study aims to create a game in the form of a health-themed adventure game using Unity with the game name "Higeia". This game is made using a combination of methods from Markov Chain and Collision Detection where the Markov Chain method is used as a classification of the attack behavior of NPC characters while the Collision Detection method is used to provide behavioral decisions when the player character hits an obstacle in the gameplay. Where the behavior of the NPC will be divided into three behaviors, namely patrol, chase and attack behavior. The variables used are speed (Kc), distance (Jr), and life (Ny) from the condition of the player character. From the test results using the Markov chain method calculation with the input of 15 data for the initial value of the player condition, the results of NPC behavior are 40% NPCs patrolling, 13.33% NPCs chasing and 46.67% NPCs shooting with collision detection events that function in accordance with which are expected.

Keywords— Game, Collision Detection, Markov Chain, NPC, Artificial Intelligence