New Time Horizon Based Classification of PV Power Generation Forecasting Techniques

Among the common renewable energy sources (RES), photovoltaic (PV) energy holds an increasing share in the global energy market. The depending on various environ mental aspects, weather conditions and non-linear characteristics create the main challenge in PV based installations specially on large farms level. Load variations and future expectations play a critical role in calculating the reserve stand-by units for vally-filling; a major challenge for grid-control authorities. Consequently, PV power generation forecasting is unavoidable for robust, reliable and cost-effective grid control from both operation and maintenance aspects. Despite the available review articles, recent forecasting techniques are not well classified or aggregated with solid benchmark assessments. This paper presents a novel criteria for classifying all the developed PV power generation forecasting techniques based on time horizon classification. In addition, inputs, outputs, forecasting methodology and performance metric are included in the proposed criteria. Detailed classification and comparison tables are concluded with detailed discussion and survey outcomes. The presented paper acts as a robust road-map for researchers concerned with this current topic as it accommodates all the available techniques with novel classification and comparison.