Surrogates Aided Constrained Optimization Design for a Crude Oil Distillation Unit

Crude oil distillation is one of the most economical separation processes in the petroleum refinery industry. In this work, a novel optimization-based design methodology, termed as Surrogates Aided Constrained Optimization Design (SACOD), is proposed for time-consuming objectives and constraints in the evolutionary optimization design of Crude oil Distillation Units (CDUs). By constructing global and local surrogate column models of the CDUs to maximize economic profits, a constraint-handling optimizer is used for infilling operations and optimization search. The framework of SACOD consists of Surrogate Models Construction, Model Management, and Particle Swarm Optimization (PSO) Mixed-Integer Constrained Evolutionary (PSO-MICE) search. The optimization results of the CDUs indicate that SACOD outperforms the single surrogate column model based approaches. Besides, the results of the proposed method are more consistent with that of the rigorous CDU model optimization approach, whereas the evaluation numbers of the time-consuming rigorous models can be significantly reduced.