Luís José Pontes e Oliveira (ist153896)

Avaliada

Desenvolvimento de modelo de comando e controle de Energy Box no âmbito da tecnologia V2G

In recent years, an increasing amount of research has been done in smart electricity management solutions. This is mainly due to the expected increase in (i) the production of electricity by means of distributed generation, (ii) the availability of local electricity storage options (e.g., electric vehicle?s batteries) and (iii) controllable appliances (e.g., domotics). In this context, a wide range of decisions allowed to the consumers emerge, such as charge or discharge the electric vehicle?s battery, set the temperature for the air conditioner, or buy/sell electricity. To handle these electricity management options this work proposes the use of an Energy Box (EB), a 24/7 computer process with the purpose of managing the electrical energy usage, either at residences or small business locations. The EB requires as input the forecasted information of several parameters, such as the wind speed for the considered location. If it simply used the predictions for the nearest location, a considerable amount of error might be introduced into the model, so there was the need to develop a tool that predicts the wind speed values at the desired location. Other issues to consider are the particularities of each residence, that if not correctly calibrated would influence the model?s behavior, such as the thermal model for a building and its occupancy patterns. To address these issues it is proposed the creation of several neural networks, working as auxiliary tools, to create an effective and self-adaptive method with the purpose of improving the input quality for the EB.

{pt=Energy Box, Gestão da Energia, Previsão de velocidade de vento, Veículo Eléctrico, en=Energy Box, Electricity Management, Wind Speed Forecast, Electric Vehicle, Vehicle-to-Grid}

25-11-2010 09:00

Christos Ioakimidis

Centro de Estudos em Inovação, Tecnologia e Políticas de Desenvolvimento (IN)