Surrogate-based optimization with improved support vector regression for non-circular vent hole on aero-engine turbine disk

Abstract

This paper investigates a structural optimization for the non-circular vent hole on an aero-engine turbine disk. A novel optimization approach, namely SO-ISVR (the surrogate-based optimization using an improved support vector regression), for the vent hole design is developed. ISVR (the improved support vector regression) is proposed to extract some ignored valuable information from the existing training data by using the least squares method, so as to improve the performance of SVR (support vector regression). To validate the advantages of SO-ISVR, another two optimization approaches are also developed. They are SO-SVR (the surrogate-based optimization using SVR) and FEMO (the finite element method based optimization). The results show that ISVR is suitable and valuable in engineering optimization. Compared to the initial scheme, the maximum von Mises stress of the optimal scheme obtained by SO-ISVR is reduced from 1189.488 MPa to 948.530 MPa. Further comparative study for SO-ISVR, SO-SVR, and FEMO demonstrates that SO-ISVR possesses some advantages in both computational efficiency and optimization effect.