Deep Learning for the Modeling and Inverse Design of Radiative Heat Transfer

Abstract

Deep learning is having a tremendous impact in many areas of computer science and engineering. Motivated by this success, deep neural networks are attracting increasing attention in many other disciplines, including the physical sciences. In this work, we show that artificial neural networks can be successfully used in the theoretical modeling and analysis of a variety of radiative-heat-transfer phenomena and devices. By using a set of custom-designed numerical methods able to efficiently generate the required training data sets, we demonstrate this approach in the context of three very different problems, namely (i) near-field radiative heat transfer between multilayer systems that form hyperbolic metamaterials, (ii) passive radiate cooling in photonic crystal slab structures, and (iii) thermal emission of subwavelength objects. Despite their fundamental differences in nature, in all three cases we show that simple neural-network architectures trained with data sets of moderate size can be used as fast and accurate surrogates for doing numerical simulations, as well as engines for solving inverse design and optimization in the context of radiative heat transfer. Overall, our work shows that deep learning and artificial neural networks provide a valuable and versatile toolkit for advancing the field of thermal radiation.

Heat-shedding with photonic structures: radiative cooling and its potential

“…Some researchers have tried to solve this problem using machine learning. 197,198 However, the demand for huge data sets and high initial costs are a barrier to entry.…”

Machine Learning for Harnessing Thermal Energy: From Materials Discovery to System Optimization

“…Some works also deal with BTEs to tailor to other systems (entropy closure of the momentum system, fluids, etc.) and properties. − Moreover, the framework of using ML to solve BTEs can be readily extended to other macroscopic PDEs in thermal transport (i.e., heat conduction equation, NS equation, radiative transfer equation) and replace the current slow trial-and-error finite element methods as well. ML has been shown to efficiently provide accurate results in contrast to conventional methods.…”

Designing radiative cooling metamaterials for passive thermal management by particle swarm optimization

“…[41][42][43][44][45][46] The machine learning is utilized to inversely design photonic crystal structures [47][48][49] and radiative coolers. [41,[50][51][52][53][54][55][56] Moreover the multilayer photonic device with microstructures has been demonstrated as highperforming radiative cooling devices. [7,21,51,[57][58][59] Therefore, the machine learning technology has the wide application prospects of designing the high-performing radiative cooling metamaterials.…”