Multi-objective Optimization Design of High-Speed Centrifugal Pump Impeller Based on NSGA-III Algorithm

To improve the hydraulic performance of high-speed centrifugal pumps, a multi-objective optimization of the impeller was performed by the optimal objective of maximizing the hydraulic efficiency, minimizing the net positive suction head (NPSHr), and maximizing the head. Based on the hydraulic loss model and sensitivity analysis results, the impeller inlet diameter, impeller outlet width, and blade outlet angle were selected as the design variables. A sample library was established for different geometric parameter models of the impeller, and support vector regression (SVR) was used to fit the external characteristics of the test samples, establishing an approximate function between the impeller parameters and the external characteristics. The non-dominated sorting genetic algorithm based on reference points (NSGA-III) was used to perform multi-objective optimization on the approximate model, obtaining the optimal combination of design variables. Compared with the initial pump, the optimized pump's head has increased by 6.73%, efficiency has increased by 3.85%, and cavitation allowance has decreased by 3.4%, and the internal flow field of the pump is significantly improved.