시스템다이내믹스 시뮬레이션에서의 동적 최적화에 관한 연구

Abstract

For the optimization of system dynamics simulation, only one optimal policy has been calculated through simulation by applying static optimization. This method does not reflect the reality that the policy changes depending on the point of time, so it may not be the optimal value at a specific point in time. Therefore, this study proposes a method to calculate the optimal policy for each time point by using dynamic optimization in the system dynamics simulation. For the simulation, a fixed order quantity inventory management model was built using the system dynamics software (Vensim). Dynamic optimization was limited in implementation in Vensim, so the optimal reorder point (ROP) was calculated at each time point using data analysis software (Python). This study is meaningful in that it presented a method to apply dynamic optimization in system dynamics simulation by linking Vensim and Python. It is expected that it will be used as a tool to support real-time policy making in public institutions and the private sector in the future.