Dynamic equilibrium through reinforcement learning

Reinforcement Learning is an area of Machine Learning that deals with how an agent should take actions in an environment such as to maximize the notion of accumulated reward. This type of learning is inspired by the way humans learn and has led to the creation of various algorithms for reinforcement learning. These algorithms focus on the way in which an agent’s behaviour can be improved, assuming independence as to their surroundings. The current work studies the application of reinforcement learning methods to solve the inverted pendulum problem. The importance of the variability of the environment (factors that are external to the agent) on the execution of reinforcement learning agents is studied by using a model that seeks to obtain equilibrium (stability) through dynamism – a Cart-Pole system or inverted pendulum. We sought to improve the behaviour of the autonomous agents by changing the information passed to them, while maintaining the agent’s internal parameters constant (learning rate, discount factors, decay rate, etc.), instead of the classical approach of tuning the agent’s internal parameters. The influence of changes on the state set and the action set on an agent’s capability to solve the Cart-pole problem was studied. We have studied typical behaviour of reinforcement learning agents applied to the classic BOXES model and a new form of characterizing the environment was proposed using the notion of convergence towards a reference value. We demonstrate the gain in performance of this new method applied to a Q-Learning agent.

Execução de um projecto solar térmico para uma instalação de A.Q.S. da Escola Básica Guilherme Stephens

O presente documento tem como principal objectivo efectuar o projecto de dimensionamento de um sistema de águas quentes sanitárias para uma escola. Numa primeira fase foi elaborado uma pesquisa sobre o contexto energético, a nível mundial, europeu e nacional, bem como o seu contexto jurídico a nível europeu e nacional, e uma explicação superficial sobre os fundamentos da energia solar, onde se foca a importância da radiação solar e os vários tipos de sistemas solares térmicos, bem como os seus constituintes. Segue-se a abordagem ao caso de estudo onde foram efectuados inicialmente inquéritos como forma de determinar os consumos de água quente utilizada nessa escola. Continuou-se o estudo efectuando-se a variação de duas características do sistema solar: o tamanho dos depósitos e o tipo de colectores solares a aplicar. Após as simulações efectuadas para a determinação das soluções a aplicar ao sistemasolar e apresentadas ao longo do presente documento, foram efectuadas análises económicas como forma de se verificar a viabilidade do sistema a aplicar. Por último foram elaboradas conclusões sobre o sistema a aplicar e apresentados alguns cenários financeiros do mesmo.

Assessment of fungal contamination in waste sorting and incineration: case study in Portugal

Organic waste is a rich substrate for microbial growth, and because of that, workers from waste industry are at higher risk of exposure to bioaerosols. This study aimed to assess fungal contamination in two plants handling solid waste management. Air samples from the two plants were collected through an impaction method. Surface samples were also collected by swabbing surfaces of the same indoor sites. All collected samples were incubated at 27◦C for 5 to 7 d. After lab processing and incubation of collected samples, quantitative and qualitative results were obtained with identification of the isolated fungal species. Air samples were also subjected to molecular methods by real-time polymerase chain reaction (RT PCR) using an impinger method to measure DNA of Aspergillus flavus complex and Stachybotrys chartarum. Assessment of particulate matter (PM) was also conducted with portable direct-reading equipment. Particles concentration measurement was performed at five different sizes (PM0.5; PM1; PM2.5; PM5; PM10). With respect to the waste sorting plant, three species more frequently isolated in air and surfaces were A. niger (73.9%; 66.1%), A. fumigatus (16%; 13.8%), and A. flavus (8.7%; 14.2%). In the incineration plant, the most prevalent species detected in air samples were Penicillium sp. (62.9%), A. fumigatus (18%), and A. flavus (6%), while the most frequently isolated in surface samples were Penicillium sp. (57.5%), A. fumigatus (22.3%) and A. niger (12.8%). Stachybotrys chartarum and other toxinogenic strains from A. flavus complex were not detected. The most common PM sizes obtained were the PM10 and PM5 (inhalable fraction). Since waste is the main internal fungal source in the analyzed settings, preventive and protective measures need to be maintained to avoid worker exposure to fungi and their metabolites.

A gestão dos custos da qualidade: Modelo A.B.Q.C. aplicado à Mercauto

Mestrado em Controlo e Gestão dos Negócios

O webmarketing e o perfil de consumidor online: estudo de caso da Delta Q

Mestrado em Controlo de Gestão e dos Negócios