Prediction Based Reliability Assessment of Solar Integrated System

In this paper, the utilization of prediction methods to project the reliability of solar-integrated power systems has been proposed. Here, the performance and reliability of solar energy systems are aimed at being assessed by employing predictive analysis techniques. The evaluation of system reliability, which relies mostly on weather conditions and energy production, needs to be investigating in any models. The reliability assessment is more difficult as solar energy sources are represented by multistate generating models. The effect of solar power generation variability is considered in the current work. For reliability assessment, a time series-based non-linear autoregressive with exogenous input artificial neural network (NARX-ANN) model is employed to forecast solar irradiance and solar power generation.