GMDH-neural network approaches for packed bed thermal energy system based on PCM integrated with a porous medium and magnetic field

Abstract

Packed bed thermal energy systems (PBTES) are recognized as one of the innovative technologies in the field of energy storage. This study numerically investigates the effects of a porous medium, magnetic field, mechanical vibrations, and various configurations of concrete-phase change material (PCM) on the performance of PBTES. The results indicate that substituting PCM for concrete can increase the discharge-to-charge energy ratio by over 300 times. The presence of a porous medium in the PBTES system with PCM significantly enhances the charge energy ratio (by 3.81–4.14 times) compared to scenarios without a porous medium, due to its influence on the melting process of the PCM. The presence of a magnetic field, along with an increase in its intensity, positively affects the melting process and enhances charge energy, potentially increasing it by approximately 4.132–5.281 times compared to cases without a magnetic field. Mechanical vibrations also influence charge energy in the PBTES system, resulting in an improvement of 4.41–4.56 times compared to the no-vibration scenario, with optimal efficiency achieved at A = 1e-5 m and f = 0.1 Hz. Notably, the use of a porous medium, magnetic field, and forced vibrations reduces discharge energy by approximately 0.34–0.37, 0.36 to 0.47, and 0.39 to 0.41 times, respectively, compared to the baseline scenario. Utilizing the Group Method of Data Handling (GMDH) neural network model based on the available data in this study, the discharge energy to charge energy ratio has been estimated, and the model has accurately predicted the desired parameter with a high degree of precision.