An optimal design method for complex structural components based on intelligent inference

Abstract

A optimal design method based on intelligent inference is proposed aiming at the existing deficiency such as large design space, low solving efficiency and poor quality caused by many characteristic parameters of complex structural components. By numerical simulation technology and DOE (Design of Experiment), the components are simulated under multi-conditions. The decision table of multidisciplinary simulation data is obtained through the discretization of continuous simulation data by clustering analysis. Then the design parameter characteristics are backtracked to guide the multidisciplinary design optimization of the components by using data mining technology. Finally, this method is verified through a case study on a cylinder head of inline4 engine.