Neural Architecture Search for Surrogate Model in Heat Source Layout optimization

In practical engineering, e.g. satellite layout design, heat source layout optimization (HSLO) is an effective technique to control the maximum temperature and improve the heat management of the whole system. Recently, deep learning surrogate assisted HSLO has been proposed, which introduces a deep learning surrogate model to learn the mapping from layout to its corresponding temperature field, so as to substitute the simulation during optimization to decrease the computational cost largely. However, the neural network surrogate for the certain task is often manually designed to be complex and requires rich debugging experience, which is challenging for the designers in the engineering field. To address the challenge, considering reducing the total parameter numbers and ensuring the similar accuracy as well as, a neural architecture search (NAS) method combined with Feature Pyramid Network (FPN) framework is developed to realize the purpose of automatically searching for a small deep learning surrogate model for HSLO. One typical two-dimensional heat conduction optimization problem is utilized to demonstrate the effectiveness of the proposed method. With the similar accuracy, NAS finds models with 80% fewer parameters, 64% fewer FLOPs and 36% faster inference time than the original FPN.