Surrogate-assisted evolutionary multi-objective optimisation applied to a pressure swing adsorption system

The complexity of chemical plant systems (CPS) makes optimising their design and operation challenging tasks. This also results in analytical numerical simulation models these having high computational costs. Research demonstrates the benefits using machine learning as surrogates or substitutes for computationally expensive during CPS optimisation. paper presents our study, extending recent research into operation. study explored original surrogate-assisted genetic algorithms (SA-GA) more complex variants optimisation problem. include additional parallel feedback components. proposes a novel multivariate extension, NSGA (SA-NSGA), to univariate SA-GA algorithm. evaluated SA-NSGA extension on popular pressure swing adsorption (PSA) system. outlines extensive experimentation, comparing various meta-heuristic techniques numerous surrogates. both PSA case show suitability combined with surrogate an framework single multi-objective scenarios. analysis further confirms that combining algorithm substitute long-running yields significant efficiency improvements, 1.7–1.84 times speedup increased examples 2.7 discussion successfully concludes evolutionary can be scaled increasingly