Multi-Layer RNN-based Short-term Photovoltaic Power Forecasting using IoT Dataset

Photovoltaic power fluctuation in daytime is one of critical problems for the efficient and stable operation of the smart grid. To respond the PV power fluctuation resulting from weather change, the short-term PV power forecasting algorithm using multi-layer RNN is proposed in this paper. It consists of multiple RNN layers using power and meteorological data which are collected by on-site IoT (Internet of Things) sensors. Experimental results showed that the accuracies of the short-term PV power prediction of 5 minutes and 1 hour later using 3 RNN layers with 12 time-step were 98.02% and 96.58% based on the normalized RMSE, respectively. These experimental results confirmed that the proposed short-term prediction algorithm using multi-layer RNN model was applicable to respond the short-term PV power fluctuation.