On the basis of the three-dimensional design platform of the mixed-flow pump impellers, an optimization design system was developed in this paper by improving the genetic algorithm with application of both strategies of keeping the optimal individual and employing the niche. This system took the highest efficiency of the impeller as the optimization objective and employed P, a0, Δθh and Δθt, which could directly affect the shape and the position of the blade, as optimization parameters. In addition, loss model was used to obtain fast and accurate prediction of the impeller efficiency. The optimization results illustrated that this system had advantages such as high accuracy and fine convergence, thus to effectively improve the design of the mixed-flow pump impellers. Numerical simulation was applied to determine the internal flow fields of the impeller obtained by optimization design, and to analyze both the relative velocity and the pressure distributions. The test results demonstrated that the mixed flow pump had the highest efficiency of 87.2%, the wide and flat high efficiency operation zone, the relatively wide range of blade angle adjustment, fine cavitation performance and satisfied stability.
