Key points are not available for this paper at this time.
The hydraulic performance of a high-speed submersible axial flow pump is investigated to reduce its energy consumption. A more efficient and stable optimization method that combines parametric design, computational fluid dynamics, and a computer algorithm is proposed. The main aim is to broaden the high-efficiency operating zone, so the average efficiency under multiple conditions is optimized while considering rotor–stator matching. The design-of-experiments method and a radial-basis-function neural network are combined to form the optimization platform, and automatic optimization of the pump design is realized through repeated execution of design and simulation. The flow loss mechanism inside the pump is studied in depth via the entropy generation rate, and regression analysis shows that the pump efficiency is influenced mainly by the blade angles. After optimization, the target efficiency is increased by 8.34%, and the flow field distribution shows that the channel vortex and hydraulic loss are controlled effectively. Finally, the results are validated by experiment. The proposed optimization approach has advantages in saving manpower and obtaining globally optimal solutions.
Lu et al. (Thu,) studied this question.