Earticle

Embedded systems involve a coupling of hardware and software. Modeling of an embedded system is a key step in the requirements phase, when a product’s specifications are gathered and modeled. This paper focuses on the specification and modeling phase (the requirements phase) in systems that have been modeled using the FSM approach. The concept of state has been a central notion in this development. While FSM has many diverse and unique properties, the notion is not rich enough to capture complex systems, especially embedded systems with software and hardware parts. The concept of state raises misconceptions, especially in education. To provide an alternative to FSM for system description, this paper introduces a new model as the foundation for a new approach to modeling applications that need state-like methodology. The methodology is based on identifying “things that flow” (e.g., data, numbers, signals) and specifies their streams of flow in terms of no more than six generic stages. The resulting conceptual picture provides a map of diverse flow systems that trigger each other. In this paper, study cases that originally used FSM are re-diagrammed using the new methodology to compare the semantics reflected in each method. Such semantics can be used to understand the behavior of a given system in design and educational contexts. The results point to the viability of the flow-based method for capturing the specification and modeling of systems.

요즘 들어 컴퓨터 게임에서의 AI(Artificial Intelligence， 인공지능) 비중은 하드웨어의 발전에 따른 것도 있지만 게임에서의 재미요소로서 그 중요성과 더불어 점차 높아지고 있다. 본 논문에서는 이러한 시대적 흐름에 따라 실시간 전략게임에서의 AI로서 인공생명 알고리즘 중 FSM(Finite State Machine) 기법을 이용하여 유저가 조작하는 PC(Player Character)에게 생물학적 진화의 개념을 도입한 진화 알고리즘을 설계 구현하였다.