Multiobjective optimization of miniature proportional valve parameters for balancing flow and response characteristics: A non-dominated sorting genetic algorithm II approach

Miniature proportional valves enable precise and continuous airflow regulation, making them well suited for precision fluid-control applications. However, their performance depends strongly on the coordinated design of multiple coupled structural parameters. Traditional single-objective optimization methods often struggle to balance conflicting performance requirements, such as dynamic response, flow capacity, and flow linearity. To address this issue, this study proposes a multi-objective parameter co-optimization framework to improve the overall performance of a miniature proportional valve. First, an electromagnetic-mechanical-fluid coupled simulation model was established to construct a four-objective evaluation system for valve performance. Although the optimization was formulated with four objectives, the pairwise Pareto plots in the main text are presented only as representative two-dimensional projections for visualisation and interpretation purposes. Specifically, electromagnetic force, opening time, mass flow rate, and flow linearity were used to characterize actuation demand, dynamic response, flow capacity, and proportional control quality, respectively. Then, six key structural parameters were identified as design variables through parameter correlation analysis and optimized collaboratively within the proposed framework. Based on the optimized results, the final design specifications were determined, and the same parameter set was used in both the updated numerical simulation and prototype fabrication. Experimental results showed that the controllable current interval expanded from 30–46 mA to 32–105 mA, corresponding to an increase in effective regulation width from 16 mA to 73 mA. In addition, the flow-linearity index decreased from 18.4% to 3.5%, and the response time of the optimized valve was 2.5 ms. The updated simulation reproduced the same broadening trend of the current-flow characteristic, further supporting the validity of the proposed optimization framework.