Abstract: The turbine disk is a core part of aeroengine, of which the lightweight design is essential to improve the efficiency of aeroengine. The structural optimization of turbine disk under centrifugal load is made based on the bidirectional evolutionary structural optimization (BESO) algorithm topology optimization model which is programed by parametric design language ANSYS APDL (ANSYS parametric design language), combined with sensitivity filtering method to suppress the checkerboard phenomenon, as well as considering the influence of blades on the optimization result by applying the same load as blade weight. The result shows that the BESO algorithm can reduce the number of topology optimization iteration steps and significantly improve the optimization efficiency. Under the condition of 26% weight reduction, the total strain energy is reduced by 48%, the structural stiffness is increased and the maximum equivalent stress is reduced by 25%, and also the strain energy distribution is more uniform for the new structural topology model.