Solar flat-plate collectors (FPCs) have many applications in low-temperature thermal systems, but they are limited due to the intermittent nature of solar radiation, the small thermal storage capacity and the poor thermal conductivity of traditional phase change materials (PCMs).
This paper presents a novel Graphene-Augmented PCM with Echo State Network-Optimized Solar Collector (GPCM-ESN-SC), which combines the graphene-enhanced PCM with fractal fin structure as inspired by nature, Synchrosqueezed Wavelet Transform (SWT)-based thermal signal analysis, and the Echo State Network (ESN) for intelligent real-time thermal prediction and control.
The uniform heat transfer and fast thermal charging/discharging characteristics of the graphene-enhanced PCM and fractal fin configuration enable high-resolution transient thermal features to be extracted by SWT, which allows for accurate prediction of the transient thermal features based on ESN.
The proposed framework enables to accurately predict the transient thermal response and also offers computational efficient thermal-energy management for future integration of adaptive control strategies.
The improvements are representative of the engineering potential of combining advanced thermal materials, hierarchical heat transfer structures, and lightweight artificial intelligence to create high-performance next generation solar thermal energy storage systems.
Solar flat-plate collectors (FPCs) have many applications in low-temperature thermal systems, but they are limited due to the intermittent nature of solar radiation, the small thermal storage capacity and the poor thermal conductivity of traditional phase change materials (PCMs). This paper presents a novel Graphene-Augmented PCM with Echo State Network-Optimized Solar Collector (GPCM-ESN-SC), which combines the graphene-enhanced PCM with fractal fin structure as inspired by nature, Synchrosqueezed Wavelet Transform (SWT)-based thermal signal analysis, and the Echo State Network (ESN) for intelligent real-time thermal prediction and control. The uniform heat transfer and fast thermal charging/discharging characteristics of the graphene-enhanced PCM and fractal fin configuration enable high-resolution transient thermal features to be extracted by SWT, which allows for accurate prediction of the transient thermal features based on ESN. The simulation results show that it can improve the thermal efficiency by about 5–12%, reduce the prediction error with RMSE of 1.72 ± 0.06 °C and MAE of 0.98 ± 0.08 °C, enhance the heat-storage capacity under the condition of no solar energy, and make the heat output more stable than the traditional flat-plate collector with PCM. The proposed framework enables to accurately predict the transient thermal response and also offers computational efficient thermal-energy management for future integration of adaptive control strategies. The improvements are representative of the engineering potential of combining advanced thermal materials, hierarchical heat transfer structures, and lightweight artificial intelligence to create high-performance next generation solar thermal energy storage systems.
