A three-stage axial turbine was redesigned by using an aerodynamic optimization design process of multistage axial turbine in multiple working conditions

which combines quasi-3Ddesign methods and multistage local optimization methods.Genetic algorithm and artificial neural network were employed to 3Dlocal optimization of various cascades.The flow field was computed through a three-dimensional viscosity Navier-Stokes equation.With optimization design

the performance of every cascade was optimized

and the overall efficiency increased by 1% under the reliable total flow mass

indicating that the total performance was improved to satisfy the design requirements.