A physics and data co-driven surrogate modeling approach for temperature field prediction on irregular geometric domain

In the whole aircraft structural optimization loop, thermal analysis plays a very important role. But it faces severe computational burden when directly applying traditional numerical tools, especially each involves repetitive parameter modification and analysis. Recently, with fast development of deep learning, several Convolutional Neural Network (CNN) surrogate models have been introduced to overcome this obstacle. However, for temperature field prediction on irregular geometric domains (TFP-IGD), CNN can hardly be competent since most them stem from processing regular images. To alleviate difficulty, we propose novel physics data co-driven modeling method. First, after adapting Bezier curve in parameterization, body-fitted coordinate mapping is generate transforms between physical plane plane. Second, physics-driven partial differential equation (PDE) residuals as loss function utilized meshing (meshing surrogate); then, present data-driven model based multi-level reduced-order method, aiming learn solutions above (thermal surrogate). Finally, combining grid position information provided by scalar (combined model), reach an end-to-end parameters domain. Numerical results demonstrate that our method significantly improve accuracy smaller dataset while reducing training time compared other methods.