A Modular Approach for Surrogate Modeling of Flowsheets

Flowsheet simulation, an important building block in chemical process development and design, generally requires the solution of a nonlinear system equations. Such simulation can fail if either no exists for chosen specifications or initial values solver are unfavorably. In this work, to overcome such convergence failures, AI-based surrogate models trained unit operations flowsheet then interconnected. For pressure swing distillation with recycle, it is shown that interconnection yields more accurate results compared model whole at once, adequate starting points be obtained from interconnected surrogates.