Multi-objective Optimization Design of Cyclone Pump Impeller Based on Agent Model and Genetic Algorithm

To determine the optimal design configuration for enhancing the performance of cyclone pumps, this study employs a multi-parameter objective optimization approach based on three critical variables: impeller blade thickness, blade width, and outlet deflection angle. A systematic methodology was used, combining Latin hypercube sampling with computational fluid dynamics simulations to construct a comprehensive training dataset. This dataset was subsequently utilized to develop a Gaussian process regression (GPR) prediction model. The non-dominated sorting genetic algorithm II (NSGA-II) was applied for multi-objective optimization, yielding a Pareto-optimal solution set, from which the optimal configuration was determined using the technique for order preference by similarity to ideal solution (TOPSIS) method. Comparative analyses of pre- and post-optimization models were conducted through multi-case numerical simulations, focusing on pressure distribution contours, velocity vector diagrams, and turbulence kinetic energy fields. The results reveal significant improvements in internal flow dynamics following optimization. The optimized pump configuration exhibits superior hydraulic performance across multiple operating conditions, particularly under the rated condition, where the head increased by ۴.۹۳ m and efficiency improved by ۲.۸۳ percentage points relative to the baseline model. Detailed flow field analysis indicates that the optimized design provides several critical enhancements: reduced pressure gradient magnitude, improved flow stream uniformity, decreased spatial extent of turbulent dissipation regions, and lower energy losses. These modifications collectively account for the enhanced hydraulic performance and demonstrate the effectiveness of the proposed optimization framework for cyclone pump design.

کلیدواژه ها:

Cyclone pump ، Numerical simulation ، Genetic Algorithm ، Agent model ، Multi-objective optimization