An integrated blade optimization approach based on parallel ANN and GA with hierarchical fair competition dynamic-niche

摘要

This work presented an available multi-point blade optimization procedure for better aerodynamic performances. Based on the proposed Parallel ANN and GA with hierarchical fair competition dynamic-niche (GA-HFCDN), an integrated approach for the blade optimization design was put forward by combining Bezier parameterization with FINE/TURBO solver. In the optimization design, parallel ANN was employed to build a more proper approximate model. GA-HFCDN was proposed to obtain the global optimization solution more efficiently for the blade design. Two research cases, including plane cascades blade optimization and the optimization and experimental study of a low specific speed centrifugal blower blade, were performed by using the above approach. The conclusions showed the rationality and validity of the optimization approach.