Investigation of Heat Source Layout Optimization by Using Deep Learning Surrogate Models

Abstract

The optimization of heat source layout (HSLO) is able to facilitate superior heat dissipation, thereby addressing the complexities associated with thermal management. However, HSLO is characterized by numerous degrees of freedom and complex interrelations between components. Conventional optimization methodologies often exhibit limitations such as high computational demands and diminished efficiency, particularly with large-scale predicaments. This study introduces the application of deep learning surrogate models grounded in backpropagation neural (BP) networks to optimize heat source layouts. These models afford rapid and precise evaluations, diminishing computational loads and enhancing the efficiency of HSLO. The suggested framework integrates coarse and fine search modules to traverse the layout space and pinpoint optimal configurations competently. Parametric examinations are taken to explore the impact of refinement grades and conductive ratios, which dominates the optimization problem. The pattern changes of the conductive channel have been presented. Moreover, the critical conductive ratio has been found, below which the conductive material can not contribute to heat dissipation. The outcomes elucidate the fundamental processes of HSLO, providing valuable insights for pioneering thermal management strategies.