IMPLEMENTASI ALGORITMA Q-LEARNING PADA FINITE STATE MACHINE DALAM GAME ARPG “AFTER THE WAR”

Game semakin berkembang pesat dan menjadi salah satu hiburan yang sangat popular diberbagai kalangan, terutama oleh golongan muda. Game tidak hanya sebagai sarana hiburan, tetapi juga dapat meningkatkan kreativitas, logika berfikir, dan kemampuan pengambilan keputusan.
Dalam sebuah game terdapat elemen-elemen penting seperti karakter, item, dan objek-objek lain didalamnya. Karakter dalam game dapat dibagi menjadi dua, player character (PC) dan non-player character (NPC). NPC memiliki peran yang beragam seperti support partner, enemies, dan allied. Untuk membuat NPC yang cerdas, diperlukan pengembangan algoritma kecerdasan buatan dan sistem Finite State Machine (FSM) sebagai landasan desain perilaku NPC. Integrasi kecerdasan buatan (AI) dengan FSM dapat meningkatkan respons dan kecerdasan NPC dalam menghadapi situasi permainan yang kompleks.
Berdasarkan penjelasan diatas, penulis melakukan penelitian tentang penerapan algoritma Q-Learning ke dalam FSM untuk menciptakan NPC yang responsif dan dinamis. Dengan melalui beberapa tahapan penelitian mulai dari studi literetur, perancangan pembuatan game, implementasi, sampai pada pengujian, dihasilkan kesimpulan bahwa algoritma Q-Learning bisa diimplementasikan ke dalam FSM pada NPC.
====================================================================================
The gaming industry is rapidly growing and has become a widely popular form of entertainment across various demographics, particularly among the younger generation. Games serve not only as a source of amusement but also as a means to enhance creativity, logical thinking, and decision-making skills.
In a game, essential elements include characters, items, and various objects. Characters in a game can be divided into two categories: player character (PC) and non-player character (NPC). NPCs play diverse roles such as support partners, enemies, and allies. Developing intelligent NPCs requires the integration of artificial intelligence (AI) algorithms and Finite State Machine (FSM) systems as the foundation for NPC behavior design. The integration of AI with FSM can enhance the responsiveness and intelligence of NPCs when facing complex in-game situations.
Based on the aforementioned explanation, the author conducted research on the implementation of the Q-Learning algorithm into FSM to create responsive and dynamic NPCs. The research involved several stages, including literature review, game design, implementation, and testing. The conclusion drawn is that the Q-Learning algorithm can be successfully implemented into the FSM of NPCs.