Microgrid is becoming an essential part of the power grid regarding reliability, economy, and environment. Renewable energies are main sources of energy in microgrids. Long-term solar generation forecasting is an important issue in microgrid planning and design from an engineering point of view. Solar generation forecasting mainly depends on solar radiation forecasting. Long-term solar radiation forecasting can also be used for estimating the degradation-rate-influenced energy potentials of photovoltaic (PV) panel. In this paper, a comparative study of different deep learning approaches is carried out for forecasting one year ahead hourly and daily solar radiation. In the proposed method, state of the art deep learning and machine learning architectures like gated recurrent units (GRUs), long short term memory (LSTM), recurrent neural network (RNN), feed forward neural network (FFNN), and support vector regression (SVR) models are compared. The proposed method uses historical solar radiation data and clear sky global horizontal irradiance (GHI). Even though all the models performed well, GRU performed relatively better compared to the other models. The proposed models are also compared with traditional state of the art methods for long-term solar radiation forecasting, i.e., random forest regression (RFR). The proposed models outperformed the traditional method, hence proving their efficiency.

It has been mandated that 5% of the generation capacity of conventional fossil fuel power plants shall be used exclusively for frequency regulation (FR) purposes in South Korea. However, the rotational speed of generators cannot be controlled quickly, and thus the variation in the power generation for FR takes some time. Even during this short period of time, frequency fluctuations may occur, and the frequency may be out of range of its reference value. In order to overcome the limitations of the existing FR method, 374 MW (103 MWh) battery energy storage systems (BESSs) for FR have been installed and are in operation at 13 sites in South Korea. When designing the capacity of BESS for FR, three key factors, i.e., the deployment time, duration of delivery, and end of delivery, are considered. When these times can be reduced, the required capacity for BESS installation can be decreased, achieving the same operational effects with minimal investment in the facilities. However, because a BESS for FR (FR BESS) needs to be installed under a large capacity, providing a single output, a centralized control method is employed. The centralized control method has the advantage of being able to view and check the entire system at once, although in the case of FR BESS, a novel system design that can optimize the above three factors through a faster and more accurate control is required. Therefore, this paper proposes the implementation of a distributed autonomous control-based BESS for frequency regulation. For the proposed FR BESS, the central control system is responsible for the determination of external factors, e.g., power generation/demand forecasting; and the system is designed such that the optimal control method of renewable energy sources and BESS according to real-time frequency variations during practical operation is determined and operated using a distributed autonomous control method. Furthermore, this study was verified through the simulation that the proposed distributed autonomous control method conducts FR faster than an FR BESS with conventional centralized control, leading to an increase in the FR success rate, and a decrease in the deployment time required (e.g., 200 ms).