Percorrer por autor "Paulo Novais"
A mostrar 1 - 2 de 2
Resultados por página
Opções de ordenação
- Consumption Performance, of a Multi Agent System Used to Achieve Comfort PreferencesPublication . Oliveira, Pedro Filipe; Paulo Novais; Matos, PauloIn this work, a multi-agent system was developed to manage comfort preferences, in an autonomous and completely automatic and non-invasive way for the user. This system was based on the use of an architecture supported by low cost hardware, namely Raspberrys, to support the different actuators present in the different spaces. To carry out its validation, a methodology was created to analyze consumption performance, and the results obtained in two scenarios are demonstrated here, a domestic housing scenario, and a professional environment scenario. The results obtained were quite positive for the prototype developed, and validate the option of using low cost hardware.
- Smart Environment: Using a Multi-agent System to Manage Users and Spaces Preferences ConflictsPublication . Oliveira, Pedro Filipe; Paulo Novais; Matos, PauloIn this new technological era of the Internet of Things (IoT), people search for comfort and everything that can simplify their daily lives. This paper proposes a smart environment to help achieve this purpose. The problem or challenge, which aims to overcome this, can be defined in general terms as a contribution to creating intelligent environments capable of adapting to the user’s comfort, needs, and preferences in an automatic, transparent, and non-invasive way, whether these environments are for domestic, professional, or public use. We propose the use of a multi-agent system to achieve the management of the user’s preferences, as well as any preference conflicts that may naturally exist. In this way, the space can adapt smartly to the present user’s preferences in a transparent and non-invasive way. This solution also supports the user’s mobility between different spaces and at different times, making it a fully ubiquitous solution, and consequently, an Ambient Intelligence (AmI) solution. We have developed a complete specification of an architecture that supports the proposed solution. The multi-agent system is fully developed, tested, and validated with the presented results.