Multi-objective optimization of hydrocyclones using meta-heuristic algorithms and preference-informed decision-making

Abstract

Previous hydrocyclone optimizations often overlooked crucial objectives interactions, thereby weakening the overall system performance. This study presents a framework that integrates meta-heuristic algorithms with preference-informed decision-making to simultaneously optimize key performance objectives. Meta-heuristics identify comprehensive Pareto-optimal sets, while preference-informed decision-making evaluates each solution's overall separation performance according to specific separation preferences. Supported by computational fluid dynamics, the study quantifies the trade-off between optimal overall separation performance and pressure drop, enabling the attainment of optimal overall separation performance at any pressure drop within the Pareto-optimal set. Among the evaluated algorithms, the strength Pareto evolutionary algorithm 2 (SPEA2) stands out for its exceptional diversity and convergence. With this framework, the study circumvents excessive compromises on neglected but crucial objectives, especially highlighting the significant adverse effects of overlooking water split or cut size. Overall, this study provides an integrated approach for developing energy-efficient hydrocyclones that maximize overall separation performance tailored to specific separation requirements.