Instituto Superior Técnico

About: Instituto Superior Técnico is a based out in . It is known for research contribution in the topics: Catalysis & Finite element method. The organization has 10085 authors who have published 30226 publications receiving 667524 citations. The organization is also known as: IST & Instituto Superior Tecnico.

Microalgae biomass production using wastewater: Treatment and costs

Abstract: This work is part of a LIFE project to treat urban wastewater from Aguas da Figueira (AdF, Figueira da Foz, PT) using a vertical tubular photobioreactor (PBR) prototype (150 L), to be scaled up and integrated in a waste water treatment plant (WWTP). The PBR was inoculated with three different microalgae: Chlorella vulgaris ( Cv ), Scenedesmus obliquus ( Sc ) and Consortium C ( ConsC ), isolated from the effluent. The study intends to find the best microalga in terms of wastewater remediation, biomass productivity and quality, for further uses, such as biofuel, biofertilizer and bioplastic production. The experiments achieved volumetric productivities of 0.1 g/L·d ( Cv ), 0.4 g/L·d ( Sc ) and 0.9 g/L·d ( ConsC ). The maximum removals attained by Cv , Sc and ConsC were: 84, 95 and 98% for total nitrogen; 95, 92 and 100% for phosphorus; and 36, 63 and 64% for COD, respectively. The treated water had values that are in accordance with environmental legislation (Directive 98/15/CE). Electrocoagulation was tested and resulted in an energy saving of 89%, compared with centrifugation alone. For drying the biomass, a solar dryer was used. Costs of overall processes versus conventional technologies are discussed and compared with other facilities and target values.

Large underground, liquid based detectors for astro-particle physics in Europe: scientific case and prospects

Abstract: This document reports on a series of experimental and theoretical studies conducted to assess the astro-particle physics potential of three future large scale particle detectors proposed in Europe as next generation underground observatories. The proposed apparatuses employ three different and, to some extent, complementary detection techniques: GLACIER (liquid argon TPC), LENA (liquid scintillator) and MEMPHYS (water Cherenkov), based on the use of large mass of liquids as active detection media. The results of these studies are presented along with a critical discussion of the performance attainable by the three proposed approaches coupled to existing or planned underground laboratories, in relation to open and outstanding physics issues such as the search for matter instability, the detection of astrophysical neutrinos and geo-neutrinos and to the possible use of these detectors in future high intensity neutrino beams.

Partitioning in Aqueous Two-Phase Systems: Fundamentals, Applications and Trends

Abstract: Over the last years, aqueous two-phase systems (ATPS) regained an increasing interest due to their potential in the downstream processing of biomolecules. After many years with only a few articles published, a lot of effort and work has been put into studying these systems for the partitioning of a range of compounds including proteins, organic low-molecular weight molecules or metal ions. Although several research and review articles appeared, a background review on ATPS partitioning fundamentals is needed. In this article, partitioning theories and main effects of several important factors for partitioning, such as molecular weight of the polymer, effect of added salts, pH, electrical charges, and temperature on phase diagrams, tie-line lengths, interfacial tension and settling time of the two aqueous phases are extensively reviewed. The trend in ATPS research is given compiling the recent 2008–2013 research articles published in the field.

Active Learning for Reward Estimation in Inverse Reinforcement Learning

Abstract: Inverse reinforcement learning addresses the general problem of recovering a reward function from samples of a policy provided by an expert/demonstrator. In this paper, we introduce active learning for inverse reinforcement learning. We propose an algorithm that allows the agent to query the demonstrator for samples at specific states, instead of relying only on samples provided at "arbitrary" states. The purpose of our algorithm is to estimate the reward function with similar accuracy as other methods from the literature while reducing the amount of policy samples required from the expert. We also discuss the use of our algorithm in higher dimensional problems, using both Monte Carlo and gradient methods. We present illustrative results of our algorithm in several simulated examples of different complexities.