Comparative Study of Optimal Long Short Term Memory Networks (LSTMs) for One Day Ahead Solar Irradiance Hourly Forecast

Energy sustenance is one the key challenges India is facing in the contemporary time. Rise in global warming and the increasing need for dependency on clean energy has motivated researchers to develop novel techniques for harnessing maximum energy from renewable sources such as solar irradiance. However, one major issue which is impeding the large scale optimal implementation of solar farms is the uncertainty associated with solar irradiance. Although several statistical forecasting methods have helped in this regard, they could not contribute to efficient utilization of solar energy. In this work, LSTMs, are implemented for modelling the time series data of solar irradiance. LSTMs are deep neural networks which are proven to be extremely efficient in modelling nonlinear large range time series data with long term dependencies. However, LSTM networks are modelled using several heuristically governed parameters whose dubious estimation renders them as an ineffective tool for time series regression. A novel NSGA II driven multi-objective evolutionary optimization framework is proposed for optimal design of LSTM networks for emulating the real world solar irradiance data. The optimally trained LSTMs were then used to forecast a 1 day-ahead hourly prediction. LSTMs were also compared with state-of-the-art system identification tools – Wavelet networks and feedforward neural networks through NARX modelling. LSTM based models were found to better among the three with an R^2 of 0.9