A novel design of experiment algorithm using improved evolutionary multi-objective optimization strategy

Abstract

The paper aims to propose an intelligent design of experiment (DOE) algorithm using an improved evolutionary multi-objective optimization approach. Adaptive evolutionary strategies are embedded in the algorithm to support the design of simulation test schemes with multiple factors whose levels are same or different. Comparative results with several existing DOE algorithms show better sampling capacity and fine sampling efficiency of the proposed algorithm. Application effects of a complex flight simulator indicate the algorithm a wide technological prospect of serving well certain complex systems.