Abstract
Micro grid represents an emergent paradigm to address the challenges of recent smart electrical grid visions, where several small-scale and distributed electrical units cooperate to achieve higher levels of energy self-sustainability, by reducing the main grid dependence. Nevertheless, the realization of this paradigm requires advanced intelligent approaches that are able to effectively manage the micro grid infrastructure and its elements. Multi-agent systems provide a suitable framework to support the development of such systems, where autonomous agents endowed with predictive data analysis capabilities take advantage of the large amount of data produced to predict the renewable energy production and consumption. In this context, this paper presents a predictive data analysis driven multi-agent system for the management of micro grids renewable energy production. The proposed approach was applied to an experimental case study, considering different predictive algorithms and data sources for the short and midterm forecasting of the production of wind and photovoltaic energy-based units.

Abstract

Abstract

Abstract
This paper aims to make a comparison between the logistics costs of buying Wood Pellets (WP) and Torrefied Biomass Pellets (TBP) produced in Portugal and exported to the major consumer markets of Northern Europe. The starting point is to determine the value of a shipload of WP and TBP delivered to a North European port and loaded in Aveiro, the main Portuguese WP expeditor port. Torrefaction implies higher energy and bulk density pellets, which contributes to increase the logistics costs associated with them. The loss of mass is greater than the loss of energy. These changes in bulk and energy densities are an advantage in terms of logistics: more tonnes per unit of volume and more energy per tonne will decrease the transportation cost per energy unit. The analysis carried out in this paper determines the energy in gigajoules (GJ) per tonne and all the comparisons are based on the cost per energy unit. This analysis is supported by real data collected in the Argus Biomass Markets report.

Abstract

Abstract
Spatial Data Infrastructures (SDI) are repositories of large volumes of data, documented through standardized metadata. Data mining is one of the main techniques used to extract knowledge from large amounts of data, because of its versatility. The purpose of this article is to use clustering techniques and data mining to extract relationships and knowledge from metadata in SDI. For this reason, knowledge discovery techniques, clustering, text mining and data mining algorithms were used. In order to demonstrate the effectiveness of the proposed method, a case study was implemented to evaluate the performance of data mining techniques in this type of database. The results showed that the data mining process and clustering techniques guided to the classification proposed method for extracting relations and knowledge from a group of metadata extracted from within the database.