No abstract

No abstract

The more frequent meteorological anomalies and climate changes push us to consider green sustainable energy as a chance to slow down such issues. Thus, we should introspect the correlations between indicators over time and understand the underneath of their meaning. Large volumes of data regarding energy are provided by Eurostat and other official data sources that require data analytics to extract valuable insights from energy indicators and indices to better understand the dynamics towards a green energy transition of the European Union State Members (EU-SM). In this paper, we analyze several energy indicators calculated for a 12-year time span with statistics and machine learning techniques, such as an unsupervised clustering algorithm with Self-Organizing Maps (SOM). Grouping the EU-SM by energy indicators from the beginning years to the end of the analyzed interval reveals differences and similarities in their efforts, shifted trends, influencing power and tendencies towards a green energy transition. The results of our analyses can be further used to assess the efficiency of stimuli for green energy generation and improve the policymakers' strategies.

No abstract

The smart metered electricity consumption data and high dimensional questionnaires provide useful information for designing the tariffs aimed at reducing electricity consumption and peak. The volume of data generated by smart meters for a sample of around four thousand residential consumers requires Not only Structured Query Language (NoSQL) solutions, data management and artificial neural network clustering algorithms, such as Self-Organizing Maps. In this paper, we propose a novel methodology that handles a large volume of data and extracts information from electricity consumption measured at 30 min and from complex questionnaires. Five three-level Time-of-Use tariffs are altered and investigated to minimize the consumers’ payment. Then, input data analysis revealed that the peak consumption is influenced by a segment of consumers that can be targeted to flatten the peak. Based on simulations, more than 23% of the peak consumption can be reduced by shifting it from peak to off-peak hours.