Optimization techniques are used as efficient strategies for the operation of Water Supply Systems (WSS). However, with the increase in complexity of the WSS’s network, the computational resources required by optimization methods also increases, potentially compromising optimal operation in real-time. This work presents a Reinforcement Learning (RL) model as an alternative to traditional optimization techniques for solving the Pump Scheduling Problem (PSP). The model is made of a Deep Q-Learning Network (DQN) agent and an environment where EPANET 2.0 simulates the hydraulic behaviour of water networks. This study aims at analysing the RL agent’s capability in arriving to a near optimal solution in the PSP, testing several energy tariffs and consumption demand patterns. While the computational cost of training the agent is significantly higher than that of finding optimal strategies using traditional optimisation algorithms, once trained the model is capable of providing near-optimal solutions almost immediately. A state features analysis is done where different configurations of both state features and rewards are tested. The results of the RL model are satisfactory by giving information of both demand and tariff patterns, while limiting the number of pump starts for the time window. A generalization analysis is made as well. The RL model achieved an average accuracy of 94% when trained with multiple tariff and consumption patterns for 60k seconds, compared to the results obtained using a non-linear programming optimizer. Additionally, the RL model used only 1% of the CPU time after training.

Neste estudo foram investigados parâmetros hemodinâmicos na artéria aorta abdominal, comparando, para condições específicas, casos com e sem estenoses. A estenose é caracterizada pelo estreitamento anormal da artéria que pode levar a complicações graves e afetar o fluxo sanguíneo. Dos casos em estudo, o caso da artéria sem estenoses é real (refere-se a um paciente específico) enquanto que o caso da artéria com estenoses estudado resulta da introdução de estenoses na geometria da artéria aorta abdominal para as mesmas condições de simulação do caso sem estenoses. O estudo assenta no pressuposto de ocorrer um escoamento laminar (do sangue) tendo-se usado o modelo de Quemada para representar o comportamento da viscosidade do sangue. A análise de diferentes parâmetros hemodinâmicos é feita para um batimento cardíaco. Os resultados obtidos revelaram diferenças significativas nos parâmetros hemodinâmicos estudados (RRT OSI, AWSS, AWSSV, pressão e velocidade) entre os dois casos estudados. Na presença das estenoses, observou-se um aumento dos valores dos parâmetros OSI, AWSS, AWSSV, como consequência do aumento da velocidade do fluxo sanguíneo na região das estenoses, não se tendo verificado alterações significativas nos campos da velocidade. Estes resultados podem fornecer informações relevantes sobre o impacto da estenose na hemodinâmica da artéria aorta abdominal. Compreender as alterações no comportamento hemodinâmico devido à presença de estenoses pode ajudar no diagnóstico precoce e no planeamento adequado de intervenções cirúrgicas. De um modo geral resulta deste trabalho que a presença de estenoses na artéria aorta abdominal pode afetar significativamente os parâmetros hemodinâmicos estudados.

ABSTRACT Litopenaeus vannamei shrimp farming is an important source of food and income in the northeast Brazilian countryside. However, the activity involves high risks, especially because of the impacts caused by diseases, such as White spot syndrome (WSS). WSS is the one which causes high losses to the Brazilian shrimp farming sector. This disease impacts the industry of the country since 2016. The infectious agent is a double stranded DNA virus capable of infecting a wide range of hosts and to cause massive mortalities. Thus, the present work aims to identify and to characterize the risk factors associated with the mortalities in L. vannamei culture systems located in the western littoral of the state of Ceará. First, it was sought to characterize the occurrence of WSS outbreaks in the study region through presumptive analysis, histopathological analysis, qPCR and water quality analysis, in order to assess the effects of the disease on mortality rates. Later, a questionnaire covering 32 risk factors associated with shrimp farming was applied to 29 farms from the region to prospect risk factors. The surveyed farms were characterized by low technological level, adopting strategies of coexistence with the WSSV, stocking huge earthen ponds with low shrimp density. Micro producers were 59% of the farmers, followed by 14% of small producers, 21% of medium sized producers and only 7% large scale producers. The qPCR analysis detected the presence of the virus in all 10 monitored farms. The shrimps presented a low infection severity index in the histopathological analysis and a good sanitary state in the presumptive analysis. The correlation and regression analysis showed that the mortality rates in the farms were not affected by the sanitary parameters evaluated. The study points that mortality in the culture system seems to be affected by other factors besides WSS. The risk factor analysis model developed pointed to three risk factors associated with the mortality rates: (1) the use of settling ponds; (2) the average level of the ponds during shrimp stocking; (3) the macroscopic aspect of the cultured shrimp.