한국형 임무공학 방법론의 수학적 정식화와 다중 기관 에이전트 시뮬레이션 개념 연구

Abstract

This paper proposes the Korean Mission Engineering Methodology (K-MEM), a mission-level decision framework tailored to Korea's defense environment. K-MEM adapts mission engineering, which treats the mission as the system of interest, to Korea's defense context, including compressed geography, North Korean nuclear, missile, and long-range artillery threats, combined and joint operations, the Korean three-axis system, Defense Innovation 4.0, rapid acquisition, R&D, test and evaluation, training, mid-term planning, and budget cycles. It formulates mission analysis as a constrained stochastic decision problem that integrates effectiveness, timeliness, resilience, combined interoperability, governance quality, industrial capability evidence, and PPBEES/R&D/acquisition pathway selection. K-MEM also models key institutional agents to show how common mission analysis outputs may be evaluated differently across organizations. A Python proof-of-concept prototype using synthetic normalized inputs illustrates how K-MEM structures alternatives, constraints, uncertainty, and institutional preferences; it is not intended as empirical validation. The simulation distinguishes alternative-level expected mission-effectiveness scores from institutional consensus scores, clarifying whether a technically preferred alternative is also institutionally acceptable. K-MEM does not replace existing planning, requirements, acquisition, R&D, test, evaluation, or training systems, but connects their outputs through a mission artifact graph to support iterative learning based on mission effectiveness.