Optimization of zero-carbon-photovoltaic home energy system employing day-ahead solar irradiance prediction

Engineering notes

Engineering notes will be added by the Power for Solar editorial team.

Chinese explanation / 中文解读

中文解读待补充：本站会优先为光伏效率、钙钛矿太阳能电池、储能技术、太阳能热利用、BIPV、并网技术等高价值论文补充中文说明。

Original abstract

Rapidly increasing electricity demands pose significant environmental threats due to high CO 2 emissions resulting from the use of non-renewable energy sources. Zero-carbon energy solutions, which rely entirely on renewable energy systems, represent a promising approach to addressing this challenge. However, complete reliance on renewable sources such as solar and wind energy remains difficult due to their intermittent and volatile nature. Consequently, most existing renewable energy solutions still incorporate diesel-based systems alongside renewable sources. In this work, we consider a zero-carbon energy system that utilizes solar energy in combination with an energy storage system and optimized demand scheduling. The proposed energy management policy leverages day-ahead solar forecasting to efficiently plan energy allocation in advance, ensuring that users’ energy demands are met throughout the day. For accurate hourly day-ahead energy prediction, we propose a computationally efficient hybrid RF–LSTM feature augmentation model that utilizes only historical solar irradiance observations. The proposed model achieves mean absolute percentage error (MAPE), normalized root mean square error (nRMSE), and normalized mean absolute error (nMAE) values of 15.99, 5.1%, and 2.1%, respectively. To manage energy allocation, we formulate the scheduling problem as a multi-objective Mixed Binary Integer Linear Program (MBILP), aiming to minimize (i) the mismatch between user demand and allocated energy, (ii) energy wastage, and (iii) scheduling delays for delay-sensitive devices, subject to battery charging/discharging constraints and storage capacity limits. We evaluate the proposed scheme under varying battery capacities and solar panel sizes. The results demonstrate that our approach improves demand satisfaction by approximately 23%, while ensuring battery safety and meeting demand threshold requirements throughout the day.

Links and sources

Comments