Optimization Strategy for an Axial-Flow Compressor Using a Region-Segmentation Combining Surrogate Model

Axial-flow compressors work against varying inlet boundary layers in real working conditions and are therefore required to perform well and robustly. This paper presents a surrogate-based optimization procedure applied to a transonic compressor to improve its efficiency and reduce the sensitivity of efficiency variation to uncertain inlet boundary layer thicknesses while maintaining the total pressure ratio. The aerodynamic optimization of compressors involves high-fidelity computational models that would cost high amounts of computational time. To implement the optimization, a region-segmentation combining surrogate model is used that is based on combinational use of the region-segmentation idea and combining surrogate modeling method to further improve prediction accuracy and reduce computational cost. Based on the region-segmentation combining surrogate model, an optimization procedure is constructed and applied to a transonic compressor. The computational results of the benchmark function and compressor optimization indicate the validity of the region-segmentation combining surrogate model in improving the prediction accuracy and computational efficiency. The optimization procedure also presents the ability to improve the compressor efficiency and make the compressor perform well and robustly at uncertain inlet boundary layer thicknesses while maintaining the total pressure ratio. The achieved aerodynamic benefits of the compressor have demonstrated the feasibility and effectiveness of the optimization strategy.