Perbandingan Performa Turn-Based Game Menggunakan Algoritma Genetika dan Logika Fuzzy

Abstract

There are a lot of games with said genre that have been published for public and most of the games have a vs. Artificial Intelligence mode or AI for short. But the variations of the AI for Turn-Based Strategy genre are abundant. That is why hopefully this research will be able to help game developers who would like to make a Turn-Based Strategy game but have no idea which type of AI suits this kind of genre the most.
The program, which is a game, that is made containts two AIs which are Genetic Algorithm and Fuzzy Logic. There are also 3 type of game modes that user can choose from, which are player vs. GA AI, player vs. Fuzzy AI, and the last one is GA AI vs. Fuzzy AI. The two AIs will be used to determine what type of unit the AI should make, who to attack and retreat from.
From the results of the testcase which are a fight between the two AIs, genetic algorithm is more than able to make different kinds of unit, but it’s also capable of producing units that are suitable for the situation. Fuzzy logic consumes more time than genetic algorithm because of the huge amount of rules and genetic algorithm having some kind of threshold to get out of its iteration before it’s supposed to end that results in not-so time consuming process for each process. But on the other hand, if the fuzzy rules and its membership functions to be optimized, it’s almost guaranteed to produce better results.