Process Transition Based on Dynamic Optimization with the Case of a Throughput-Fluctuating Ethylene Column

For continuous chemical processs, security and economy are threatened by the process transition. Considering the limitations of conventional control, dynamic optimization is a potential strategy for large-range process transition. To avoid dimension problems caused by full discretization, the dynamic optimization problem is transformed into the mixed integer nonlinear programming problem by control vector parameterization. In detail, the control variables are discretized as real parameters, whereas the time grid for discretization is determined by integer parameters. The strategy is verified by the case study of a throughput-fluctuating ethylene column. By simulation, two scenes are discussed. For small-range transition, conventional strategy is adequate. For large-range transition, the dynamic optimization is employed. Compared with the experiential operation, the equal time-interval dynamic optimization decreases the deviation and shortens the duration of process transition. Furthermore, the unbalancing of fluctuation is eliminated by the varied time-interval approach.