#TITLE_ALTERNATIVE#

• Main Author: WAYAN SUASNAWA (NIM 23206311), I
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/10710
• Institution: Institut Teknologi Bandung

Interactive application technology has developed so rapidly in line with the progress of information technology. Game technology is an interactive application technology that has been lustrously developed recently, both in the forms of game online in the Internet, local networks as well as the games that can be run only in single computer. Generally the game has an agent as a character played. This thesis covered the agent behavior for RTS game Sultan Agung including the agent controlled by human player as well as computer player in a single player mode.<p> <br /> <br /> <br /> In general the agent in the RTS game Sultan Agung has been classified into three types: worker, soldier, building and resource producer. The three agent types have different capacities and objectives. The worker has an objective to gather resource, the soldier acted as a military force to defeat the enemy, as well as the building and resource producer acted as a supporting unit.<p> <br /> <br /> <br /> Agent design that controlled by a human player uses a finite state machine (FSM) method where the agent has been modeled in several limited operating states. In a time it is only one active state. The state change occurs when there is an order from player or the changes on the system and game environment. The agent design that controlled by computer consist of a series of rules to describe what is done by agent in a situation to form the behavior equal to the agent controlled by human player.<p> <br /> <br /> <br /> The agent movement is implemented on the map in the forms of cells in the game environment using a A* pathfinding technique. The use of heuristic method in A* pathfinding enables this algorithm can encounter a shortest path between two positions on the map without having to explore all areas.<p> <br /> <br /> <br /> The outcomes obtained indicate that the agent has had a behavior suitable to the design. The computer controlled agent can do the thing that can be done by agent controlled by player. The rules for the computer controlled agent is indeed still impressed that it can be predicted by player. The possibility of development using Artificial Neural Network (ANN) method can become on alternative to make the game to be more dynamic and interesting to play.