Practical large-scale design optimizations remain a challenge because of the high cost of ML training. We show how cutting-edge ML approaches can benefit ASO and address challenging demands, such as interactive design optimization. In addition to providing a comprehensive summary of the research, we comment on the practicality and effectiveness of the developed methods. Then, we review ML applications to ASO addressing three aspects: compact geometric design space, fast aerodynamic analysis, and efficient optimization architecture. Next, we introduce ML fundamentals and detail ML algorithms that have been successful in ASO. We first introduce conventional ASO and current challenges. We review the applications of ML in ASO to date and provide a perspective on the state-of-the-art and future directions. Machine learning (ML) has been increasingly used to aid aerodynamic shape optimization (ASO), thanks to the availability of aerodynamic data and continued developments in deep learning. In addition, the predicted flow and thermal fields will provide the physical interpretation of the aerodynamic and heat transfer performances. Consequently, the present approaches will be useful to design the NACA section-based shape giving higher aerodynamic and heat transfer performances by quickly predicting the force and heat transfer coefficients. In order to physically interpret the heat transfer performance, more quantitative and qualitative information are needed owing to the lack of the correlation and the resolution of the thermal fields. The contours of the velocity components and the pressure coefficients reasonably explained the variation of the aerodynamic coefficients according to the geometric parameter of the NACA section. The predictions mostly matched well with the true data. These two models were trained and tested by the dataset extracted from the computational fluid dynamics (CFD) simulations. The established ED model predicts the velocity, pressure and thermal fields to explain the performances of the aerodynamics and heat transfer. The established CNN predicts the aerodynamic coefficients and the Nusselt number. The present study established two different models based on the convolutional neural network (CNN) and the encoder–decoder (ED) to predict the characteristics of the flow and heat transfer around the NACA sections.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |