Abstract:Purpose To improve the speed and accuracy of turbine blade film cooling design process, the most advanced deep learning models were introduced into this study to investigate the most suitable define for prediction work. This paper aims to create a generative surrogate model that can be applied on multi-objective optimization problems. Design/methodology/approach The latest backbone in the field of computer vision (Swin-Transformer, 2021) was introduced and improved as the surrogate function for prediction of the multi-physics field distribution (film cooling effectiveness, pressure, density and velocity). The basic samples were generated by Latin hypercube sampling method and the numerical method adopt for the calculation was validated experimentally at first. The training and testing samples were calculated at experimental conditions. At last, the surrogate model predicted results were verified by experiment in a linear cascade. Findings The results indicated that comparing with the Multi-Scale Pix2Pix Model, the Swin-Transformer U-Net model presented higher accuracy and computing speed on the prediction of contour results. The computation time for each step of the Swin-Transformer U-Net model is one-third of the original model, especially in the case of multi-physics field prediction. The correlation index reached more than 99.2% and the first-order error was lower than 0.3% for multi-physics field. The predictions of the data-driven surrogate model are consistent with the predictions of the computational fluid dynamics results, and both are very close to the experimental results. The application of the Swin-Transformer model on enlarging the different structure samples will reduce the cost of numerical calculations as well as experiments. Research limitations/implications The number of U-Net layers and sample scales has a proper relationship according to equation (8). Too many layers of U-Net will lead to unnecessary nonlinear variation, whereas too few layers will lead to insufficient feature extraction. In the case of Swin-Transformer U-Net model, incorrect number of U-Net layer will reduce the prediction accuracy. The multi-scale Pix2Pix model owns higher accuracy in predicting a single physical field, but the calculation speed is too slow. The Swin-Transformer model is fast in prediction and training (nearly three times faster than multi Pix2Pix model), but the predicted contours have more noise. The neural network predicted results and numerical calculations are consistent with the experimental distribution. Originality/value This paper creates a generative surrogate model that can be applied on multi-objective optimization problems. The generative adversarial networks using new backbone is chosen to adjust the output from single contour to multi-physics fields, which will generate more results simultaneously than traditional surrogate models and reduce the time-cost. And it is more applicable to multi-objective spatial optimization algorithms. The Swin-Transformer surrogate model is three times faster to computation speed than the Multi Pix2Pix model. In the prediction results of multi-physics fields, the prediction results of the Swin-Transformer model are more accurate.

A double-layer progressive architecture-based surrogate model for efficiency analysis of spiral shaft in shield machine

Surrogate Model Via Artificial Intelligence Method for Accelerating Screening Materials and Performance Prediction

Machine Learning Surrogate Model Optimized by Improved Sparrow Search Algorithm for Multi-Objective Optimization of Permanent Magnet Synchronous Motor Direct-Drive Pump

Accelerating Analysis for Structure Design via Deep Learning Surrogate Models

Adaptive Directed Support Vector Machine Method for the Reliability Evaluation of Aeroengine Structure

Prediction of multi-physics field distribution on gas turbine endwall using an optimized surrogate model with various deep learning frames

An Efficient Finite Element Analysis Method for Aero Engine Gear Reliability Combining Co-Simulation and Surrogate Model

Two-Stage Surrogate Modeling for Data-Driven Design Optimization with Application to Composite Microstructure Generation

Advanced Corrective Training Strategy for Surrogating Complex Hysteretic Behavior

Extremum hybrid intelligent-inspired models for accurate predicting mechanical performances of turbine blisk

Sheet metal forming optimization by using surrogate modeling techniques

Multisource information fusion for real-time prediction and multiobjective optimization of large-diameter slurry shield attitude

An adaptive surrogate model to structural reliability analysis using deep neural network

Research on Infilling Strategy in Surrogate-Aided Optimization for Axial Compressor Blades

Structural Dynamic Probabilistic Evaluation Using a Surrogate Model and Genetic Algorithm

A dynamic structure-adaptive symbolic approach for slewing bearings’ life prediction under variable working conditions

Generative adversarial surrogate modeling framework for aerospace engineering structural system reliability design

Research on Optimal Design of Ultra-High-Speed Motors Based on Multi-Physical Field Coupling Under Mechanical Boundary Constraints

Multi-Physics Simulation Based Design And Analysis Of A High Speed Aerostatic Spindle And Its Performance Assessment

Decomposed-coordinated Surrogate Modeling Strategy for Compound Function Approximation in a Turbine-Blisk Reliability Evaluation

A Novel Multi-Fidelity Surrogate for Efficient Turbine Design Optimization