Qineng Wang

Qineng Wang

Auto-generated profile, based on publicly available information

Qineng Wang

Institute of Turbomachinery, Xi'an Jiaotong University 1 , Xi'an, Shaanxi 710049

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
21 December 2023 in Journal of Turbomachinery
Verified
Journal Article
Closed access
PDF file
QWQineng Wang
Qineng Wang
LSLiming Song
Liming Song
ZGZhendong Guo
Zhendong Guo
JLJun Li
Jun Li
ZFZhenping Feng
Zhenping Feng

1 December 2024 in Physics of Fluids
Verified
Journal Article
Closed access
PDF file
QWQineng Wang
Qineng Wang
LSLiming Song
Liming Song
TLTianyuan Liu
Tianyuan Liu
ZGZhendong Guo
Zhendong Guo
In this study, a physics-enhanced neural operator framework is proposed to enhance the generalization prediction ability of the cooling layout of a turbine end wall with variable number of film holes. Specifically, inspired by the film cooling superposition principle, the superposition-based deep neural operator (SDNO) network is proposed, which divides the turbine end wall's temperature field prediction into two stages. In the first stage, the cooling layout of a turbine end wall is divided into several sub-parts, and a transformer-based neural operator network, namely Calculate Net, is designed to predict the temperature field of each sub-part. Then, in the second stage, another neural operator network, i.e., Superposition Net, is designed to combine all the temperature fields of each sub-part and obtain the final superposed field of full cooling layout. Additionally, instead of directly taking the film cooling layout as binary pixel data, a signed distance function which is sensitive to the variable locations of cooling holes is designed to preprocess the input layout information. Furthermore, the proposed end wall film cooling prediction model is trained with samples varying the number of film holes from 1 to 5 at different locations. Then, the trained prediction shows excellent generalization prediction ability, which can accurately predict the film effectiveness of the cooling layout with 10–20 film cooling holes that are unseen in the training samples. In the meantime, the proposed SDNO network also shows remarkable better prediction accuracy. With the above, the effectiveness of the SDNO has been well demonstrated.

27 January 2026 in Aerospace Science and Technology
Verified
Journal Article
Closed access
QWQineng Wang
Qineng Wang
XZXinrui Zhou
Xinrui Zhou
YGYang Gao
Yang Gao
XZXiaofeng Zhang
Xiaofeng Zhang
HXHairun Xie
Hairun Xie
YZYonghe Zhang
Yonghe Zhang

16 July 2021
Verified
Preprint
Open access
QWQineng Wang
Qineng Wang
LSLiming Song
Liming Song
ZGZhendong Guo
Zhendong Guo
JLJun Li
Jun Li
This paper draws motivations from the fact that current efforts of engineering optimization process were mostly carried out from scratch; i.e., tabula rasa-style design. In contrast however, the humans routinely take advantage of the knowledge that drawn from past experiences whenever a new task is met. Such transfer learning process by leveraging knowledge from completed tasks can

9 February 2026 in Aerospace Science and Technology
Verified
Journal Article
Closed access
XZXinrui Zhou
Xinrui Zhou
HXHairun Xie
Hairun Xie
BLBokai Li
Bokai Li
JWJing Wang
Jing Wang
QWQineng Wang
Qineng Wang
YZYonghe Zhang
Yonghe Zhang

24 August 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
QWQineng Wang
Qineng Wang
ZLZhaocheng Li
Zhaocheng Li
WYWenxin Yang
Wenxin Yang
GJGuohao Jing
Guohao Jing
HLHuijie Liu
Huijie Liu
HXHairun Xie
Hairun Xie
Large language models (LLMs) still struggle to participate reliably in search and decision making for complex engineering optimization problems with strong physical coupling, stringent constraints, and expensive simulations. To address this challenge, this paper proposes LASR (Large Language Model Agentic Semantic Routing), which integrates LLM-based operator selection into heuristic optimization and shifts high-level decisions from purely numerical adaptation to physics-aware semantic routing. LASR extracts compact state evidence, including feasibility constraints, resource consumption, and physical response characteristics, and uses pre-trained physical priors to prioritize operators. We further define an operator library with deterministic transformations in the mathematical design space and interpretable engineering semantics, enabling semantic decisions to be executed numerically while keeping physical operations within valid bounds. Operator execution feedback and diversity preservation close the optimization loop. On a multi-objective, multi-constraint satellite thermal design task, we construct 10C, 15C, and 20C benchmarks using DeepSeek-V4-Flash as endpoint. With an identical PDE simulation budget and a common evaluation protocol, LASR consistently improves over NSGA-II, UCB1, and Credit AOS. It increases Hypervolume (HV) by at least 20.8% and reduces Inverted Generational Distance (IGD) in the objective space by at least 34.2%, demonstrating stronger Pareto trade-off generation under limited simulations. Ablation and sensitivity analyses clarify the core mechanisms and endpoint dependence of LASR. Overall, LASR provides a practical high-level decision interface for high-cost PDE-constrained engineering multi-objective optimization.