A machine learning surrogate modeling benchmark for temperature field reconstruction of heat source systems

The temperature field reconstruction of heat source systems (TFR-HSS) with limited monitoring sensors in thermal management plays an important role in the real-time health detection systems of electronic equipment in engineering. However, prior methods with common interpolations usually cannot provide accurate reconstruction performance as required. In addition, no public dataset exists for the wide research of reconstruction methods to further boost reconstruction performance and engineering applications. To overcome this problem, this work develops a machine learning surrogate modeling benchmark for the TFR-HSS task. First, the TFR-HSS task is mathematically modeled from a real-world engineering problem, and four types of computational modelings are constructed to transform the problem into discrete mapping forms. Then, this work proposes a set of machine learning surrogate modeling methods, including general machine learning methods and deep learning methods, to advance the state-of-the-art methods over temperature field reconstruction. More importantly, this work develops a novel benchmark dataset, namely the temperature field reconstruction dataset (TFRD), to evaluate these machine learning surrogate modeling methods for the TFR-HSS task. Finally, a performance analysis of typical methods is given on the TFRD, which can serve as the baseline results on this benchmark.