Optimization of thermal storage performance of cascaded multi-PCMs and carbon foam energy storage system based on GPR-PSO algorithm

Xueming Yang,Yi Li,Yongfu Ma,Jie Cui,Jianfei Xie
DOI: https://doi.org/10.1016/j.est.2024.110626
IF: 9.4
2024-01-27
Journal of Energy Storage
Abstract:The study is focused on the heat transfer enhancement method for a latent heat thermal energy storage (LHTES) system with the three-stage axially cascaded multi-phase change materials (PCMs) and composite carbon foam. It was found that the comprehensive storage density evaluation (CSDE) of the LHTES system was enhanced up to 11.41 % under a volume distribution of 1:2:3, compared with the three-stage axially cascaded PCMs using a uniform volume distribution. Furthermore, the addition of carbon foam with a porosity of 0.9 to the multi-PCMs at each stage led to an increase of CSDE by 100.01 %. Meanwhile, comparison of results between different case studies showed that a positive gradient arrangement of carbon foam porosity could increase the CSDE compared with the negative or no gradient arrangement. In addition, a hybrid machine learning (ML) approach using Gaussian process regression (GPR) and particle swarm optimization (PSO) algorithms was applied to optimize the carbon foam porosities unfolding a nonlinear relationship between the average porosity and the thermal performance of LHTES. The optimal results show that CSDE was enlarged by 17.46 % and the best thermal performance was observed if a porosity gradient of 0.7-0.809-0.844 was arranged for the carbon foam, compared to a uniform porosity of 0.9.
energy & fuels
What problem does this paper attempt to address?