An electrification revolution in the Chinese building energy field has been promoted by China’s carbon peak and carbon neutrality goals. An accurate electricity load prediction was essential to resolve the conflict between substations which was caused by the current increase in energy demand, on both the generation and consumption sides. This review provided an in-depth study of prediction models for residential building electricity load and inspected various output types, prediction methods and driving factors. The prediction types were divided into three categories: (i) time scale, (ii) geographical scale and (iii) regional scale. Predictive model building methods were classified as classical, algorithms based on Machine Learning (ML) or Deep Learning (DL) and hybrid methods. Driving factors included single and multiple features. By summarizing the driving factors, the influence of improving the prediction accuracy according to the characteristics of output types on selecting the driving factors correctly was discussed. The review provided a key perspective for future studies in electricity load prediction by analyzing the regional variations in electricity load characteristics. It suggested that the regional electricity load prediction model for residential buildings based on diverse driving factors in each region was established to offer valuable solutions for future residential planning and energy distribution.
Against the backdrop of the ever-expanding practical applications of magnetic nanofluids, the self-driven flow and heat transfer characteristics of water-based Fe3O4 magnetic nanofluids were experimentally investigated under a uniform magnetic field in the closed-loop pipeline system in this work. Specifically, Fe3O4 nanoparticles were synthesized using the co-precipitation method, and stable magnetic nanofluids with concentrations ranging from 0.025 wt% to 0.150 wt% were prepared using sodium citrate as a dispersant. In the presence of a magnetic field, a closed-loop system that integrates heating and cooling branches was established. Furthermore, the effects of magnetic field strength, temperature difference between the heating and cooling sections, magnetic nanofluid concentration, and pipeline length on the self-circulation flow velocity were discussed, leading to insights into the heat transfer characteristics of the magnetic nanofluid. The results showed that the circulation flow velocity increases with the increase in magnetic field strength, magnetic nanofluid concentration, and temperature difference, while it decreases with the increase in pipeline length. Correspondingly, the heat transfer coefficient between the pipeline wall and the fluid increased significantly with the increase in circulation flow velocity. The priority of factors on the thermomagnetic effect is ranked as magnetic field strength > pipeline length > temperature difference > magnetic nanofluid concentration.