Blade optimization of mixed-flow pump by using inverse design method and neural network

The ordinary parameterization programs of three-dimensional blade contain a great deal of design variables, therefore an indirect parameterization method was proposed, where the angular momentum was treated as design variables, the blade was calculated by inverse design method, and neural networks were adopted to construct the response relation between the design variable and the objective function. The sample data used to train neural networks were schemed according to the design of experiment theory, the application of two kinds of neural networks-back propagation network and radial basic function network were investigated, and a new optimization method was proposed. Compared with the ordinary optimization programs, fewer variables were required in this method based on the three-dimensional viscous computational fluid dynamics (CFD) analysis and the calculation time was shortened obviously. An optimized blade in a mixed-flow pump, where the head and the efficiency were selected as the objective functions, confirms the validity of this newly proposed method.