Bayesian optimization speeds up industrial process design with fewer simulations
Reduced-Space Multi-Fidelity Bayesian Optimization of Process Simulation Models
Machine Learning
Summary
Improving complex industrial processes often requires running many slow and expensive simulations, which can take a lot of time and computing power. To tackle this, the authors created a new method that learns which parts of the process matter most and uses cheaper, simpler models to guide the search for the best design. This approach balances detailed and approximate simulations to find efficient solutions faster. They tested it on two real-world industrial examples and showed it reduces the need for costly simulations while still finding good designs.
What this means in practice
- •For process engineers: Optimize complex industrial process designs faster by using fewer costly high-fidelity simulations combined with cheaper approximations.
- •For chemical plant operators: Improve plant performance predictions by incorporating multi-fidelity simulation data, reducing downtime for expensive test runs.
Authors
Niki Triantafyllou, Andrea Bernardi, Maria M. Papathanasiou
Abstract
Optimizing industrial process flowsheets is often computationally prohibitive due to the high cost of rigorous simulations and the curse of dimensionality inherent in complex design spaces. To address these challenges, we present a reduced-space multi-fidelity Bayesian optimization (RS-MFBO) framework designed for high-dimensional, expensive black-box functions. The approach integrates Global Sensitivity Analysis (GSA) for dimensionality reduction with a fidelity-augmented Gaussian process that captures correlations between low-cost approximations and expensive high-fidelity evaluations. A cost-aware acquisition strategy, augmented with cooldown and promotion mechanisms, adaptively guides the allocation of samples across fidelities. The framework is validated on two distinct industrial process simulators: a plasmid DNA bioprocess in SuperPro Designer and a green fuel synthesis plant in Aspen HYSYS. Results across diverse economic and physical objectives demonstrate that the proposed method substantially reduces the number of high-fidelity simulator evaluations while maintaining competitive optimization performance compared to single-fidelity baselines. These results highlight RS-MFBO as a scalable, simulator-agnostic approach for cost-constrained black-box optimization.