Integrated planning, scheduling, and dynamic optimization for continuous processes

Integration of planning, scheduling, and dynamic optimization significantly improves the overall performance of a production process, compared to the traditional sequential method that solves each sub-problem one by one. The integrated model can be formulated as a mixed-integer dynamic optimization (MIDO) problem which can be then transformed into a mixed-integer nonlinear program (MINLP). However, widely-used simultaneous methods, which solve the integrated problem by a general-purpose MINLP solver, encounter computational complexity. They are difficult to apply to large-scale problems. To address this difficulty, we propose a novel efficient method to solve the integrated problem for a multi-product reactor. The method decomposes the dynamic optimization problems from the planning and scheduling problem by discretizing transition times and transition costs. Then the integrated problem is transformed into a mixed-integer linear program, which is much easier to solve than the large-scale MINLP. In the case studies, the proposed method can reduce the computational time by more than three orders of magnitudes in comparison with the simultaneous method.