Abstract
Encouraged by the considerable cost reduction, small-scale solar power deployment has become a reality during the last decade. However, grid integration of small-scale photovoltaic (PV) solar systems still remains unresolved. High penetration of Renewable Energy Sources (RESs) results in technical challenges for grid operators. To address this, Virtual Power Plants (VPPs) have been defined and developed to manage distributed energy resources with the aim of facilitating the integration of RESs. This paper introduces a hybrid irradiance forecasting approach aimed at facilitating the integration of PV systems into a VPP, especially when a historical irradiance dataset is exiguous or non-existent. This approach is based on Artificial Neural Networks (ANNs) and a novel similar hour-based selection algorithm, has been tested for a real PV installation, and has been validated also considering irradiance measurements from an aggregation of ground-based meteorological stations, which emulate the nodes of a VPP. Under a reduced historical dataset, the results show that the proposed similar hour-based method produces the best forecasts with regard to those obtained by the ANN-based approach. This is particularly true for one-month and two-month datasets minimizing the mean error by 16.32% and 9.07% respectively. Finally, to demonstrate the potential of the proposed approach, a comparative analysis has been carried out between the hybrid method and the most used benchmarks in the literature, namely, the persistence method and the method based on similar days. It has been demonstrated conclusively that the proposed model yields promising results regardless the length of the historical dataset.
Highlights
The adoption of photovoltaic (PV) power generation is rising steeply worldwide [1]
The main contributions of this paper are summarized as follows: (i) the proposed forecasting approach is implemented in the context of a Virtual Power Plants (VPPs) considering the challenges it poses and drawing on its strengths; (ii) the input data for the forecasting algorithms comes from weather forecasts regularly published, free of charge, by the AEMET; (iii) the similar hour-based approach, which produces accurate irradiance forecasts for a reduced dataset, this usually being the case when a new node is integrated in the VPP; and (iv) the ensemble of Artificial Neural Networks (ANNs) and the similar hourbased approach which, through a dynamically weighted function that depends on the type of day, produces encouraging results
This section presents the results obtained from the implementation of the similar hour-based and hybrid global horizontal irradiance (GHI) forecasting strategies in two different scenarios: firstly, the approach is applied to an experimental setup that plays the role of a VPP node; and secondly, an aggregation of different PV installations in the shape of ground-based meteorological stations making up a VPP, is considered
Summary
The adoption of photovoltaic (PV) power generation is rising steeply worldwide [1]. There are several reasons behind its success: (a) the cost of photovoltaic power has plummeted since PV modules, storage systems and balance of system costs have been steadily dropping [2]. The main contributions of this paper are summarized as follows: (i) the proposed forecasting approach is implemented in the context of a VPP considering the challenges it poses and drawing on its strengths; (ii) the input data for the forecasting algorithms comes from weather forecasts regularly published, free of charge, by the AEMET; (iii) the similar hour-based approach, which produces accurate irradiance forecasts for a reduced dataset, this usually being the case when a new node is integrated in the VPP; and (iv) the ensemble of ANNs and the similar hourbased approach which, through a dynamically weighted function that depends on the type of day, produces encouraging results. This is one of the main contributions of this paper since, to the best of the author’s knowledge, this is the first time this approach has been used for irradiance forecasting This algorithm, in contrast to the similar day-based methods, uses extra-terrestrial radiation to filter the most important time instants for the prediction, and delivers outstanding results in the early stage of the VPP node. The algorithm iteratively repeats the similarity matching as long as the extra-terrestrial radiation (E) is greater than zero for the forecast hour (t)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
