시뮬레이션 기반의 최적화를 이용한 제조 공장의 디지털 트윈 프레임워크 개발 - 원유정제 공장의 실시간 운전 최적화 적용

Abstract

A digital twin framework with Simulation-based Optimization (SBO) for industrial plants is developed. The proposed framework is composed of a combination of DWSIM and Python. DWSIM is adopted for SBO simulation because it is open source process simulation software, used to model and simulate rigorously the processes such as refinery, petrochemical, chemical, power generation and utilities. Python is adopted for SBO optimization because it provides free optimization libraries for solving various forms of nonlinear optimization problems. The combination of Aspen HYSYS and MATLAB was previously proposed for SBO and used primarily for process synthesis and design because of its advanced simulation and optimization capabilities. This study applied the SBO framework with DWSIM and Python to Crude Unit real-time optimization (RTO). The result showed that the time requirement of RTO configuration, calculation efficiency and convergence stability are acceptable compared with RTO solutions already commercialized, confirming that the SBO framework with the combination of DWSIM and Python can be helpful for Digital Twin development in industrial plants.