Multi-objective aerodynamic optimization and flow mechanism investigation of centrifugal compressor impeller based on ANN and NSGA- III
Proceedings of the Institution of Mechanical Engineers, Part A: Journal of Power and Energy · 01.07.2026
DOI: 10.1177/09576509261464337 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the Institution of Mechanical Engineers, Part A: Journal of Power and Energy, online erschienen am 01.07.2026. Verfasst von Yuqian Zhang und 6 weiteren Autor:innen. Die Arbeit nennt 48 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Open Foundation of State Key Laboratory of High-end Compressor and System Technology, The Science and Technology Key Plan Project of Zhejiang Province.
Abstract (Originalsprache)
Semi-open centrifugal impellers are crucial parts of industrial compressors, as their aerodynamic performance directly influences the energy efficiency and operational stability of the system. The current optimization research faces major challenges: substantial computational costs for high-dimensional multi-objective optimization, susceptibility to local optima, and insufficient elucidation of the physical mechanisms connecting blade angle modifications to flow loss reduction, all of which severely hinder enhancements in compressor performance. The main contribution of this work is the comprehensive revealing of the physical mechanism by which blade profile modification suppresses tip clearance-induced flow losses, as well as the efficient 9-dimensional blade angle parameterization system. The geometric parameters of the centrifugal impeller are sampled using Latin hypercube sampling (LHS), and computational fluid dynamics (CFD) is employed to execute high-precision full-flow passage numerical simulations for each sample. To significantly reduce the computational costs and overcome the local optimum defect of traditional optimization algorithms, the design space is approximated via artificial neural networks (ANNs) and optimized using the non-dominated sorting genetic algorithm-III (NSGA-III). The optimization results reveal that the polytropic efficiency and pressure ratio of the optimized impeller are increased by 0.33% and 1.56% at design conditions, respectively. Furthermore, the aerodynamic performance of the compressor has been enhanced across the whole range of operations. At the peak efficiency point (1.65 kg/s), the polytropic efficiency and pressure ratio have been enhanced by 1.68% and 3.08%, respectively. Finally, this study highlights that semi-open centrifugal impeller blades can be angled appropriately in the reverse direction of rotation to enhance flow field uniformity and mitigate flow losses, indicating a comprehensive physical understanding of flow characteristics and the aerodynamic performance of the centrifugal compressor.
Förderung
- National Natural Science Foundation of China
- Open Foundation of State Key Laboratory of High-end Compressor and System Technology
- The Science and Technology Key Plan Project of Zhejiang Province
Zitieren
APA
Zhang, Y., Mao, J., Chen, J., Shu, Y., Wang, Z., Yang, H., & Wei, Y. (2026). Multi-objective aerodynamic optimization and flow mechanism investigation of centrifugal compressor impeller based on ANN and NSGA- III. Proceedings of the Institution of Mechanical Engineers, Part A: Journal of Power and Energy. https://doi.org/10.1177/09576509261464337
Vancouver
Zhang Y, Mao J, Chen J, Shu Y, Wang Z, Yang H, et al. Multi-objective aerodynamic optimization and flow mechanism investigation of centrifugal compressor impeller based on ANN and NSGA- III. Proceedings of the Institution of Mechanical Engineers, Part A: Journal of Power and Energy. 2026. doi:10.1177/09576509261464337
BibTeX
@article{Zhang2026,
author = {Zhang, Yuqian and Mao, Jiabin and Chen, Jinxi and Shu, Yue and Wang, Zhengdao and Yang, Hui and Wei, Yikun},
title = {Multi-objective aerodynamic optimization and flow mechanism investigation of centrifugal compressor impeller based on ANN and NSGA- III},
journal = {Proceedings of the Institution of Mechanical Engineers, Part A: Journal of Power and Energy},
year = {2026},
doi = {10.1177/09576509261464337}
}
EUVIMED