Real-Time Strategy Generation System Using Case-Based Reasoning

摘要

Case-based reasoning (CBR), a problem-solving method that effectively draws the proper solutions to new issues by appropriately altering the solutions based on similar past problem-solving cases when new issues are presented, is widely used in daily life as it is very similar to humans' problem-solving process. This research was conducted to design a system in which the most proper strategy will be used in the current battle situation based on similar past cases, by applying CBR to the combat strategy system. In such system, when the squad members gather to conduct a military operation on a squad scale, the squad members will act according to the designated plan, to achieve the final goal. This process should consist of military tactics for actions such as attack, ambush, and strategic maneuvers. In other words, new strategies are necessary according to the attendant circumstances, such as the location of the shelters and enemies, and such strategies can be generated based on the learning acquired by the squad or platoon commanders from military manuals, and on their accumulated experiences. In this paper, a combat strategy generation system using CBR is proposed, in which the most appropriate similar cases stored in a case database (DB) was retrieved and applied after proper modification to the current situation. Simulated is followed by implementing the proposed system.