Examinando por Autor "Moreno-SanSegundo, José Ángel"

Mostrando 1 - 4 de 4

Deep reinforcement learning for automated search of model parameters: photo-fenton wastewater disinfection case study
(Springer, 2022) Hernández-García, Sergio; Cuesta-Infante, Alfredo; Moreno-SanSegundo, José Ángel; Montemayor, Antonio S.
Numerical optimization solves problems that are analytically intractable at the cost of arriving at a sufficiently good but rarely optimal solution. To maximize the result, optimization algorithms are run with the guidance and supervision of a human, usually an expert in the problem. Recent advances in deep reinforcement learning motivate interest in an artificial agent capable of learning to do the expert’s task. Specifically, we present a proximal policy optimization agent that learns to optimize in a real case study such as the modeling of the photo-fenton disinfection process, which involves a number of parameters that have to be adjusted to minimize the error of the model with respect to the experimental data collected in several trials. The expert spends an average of 4 h to find a suitable set of parameters. On the other hand, the agent we present does not require a human expert to guide or validate the optimization procedure and achieves similar results in 2:5 less time.
Mechanistic modelling of solar disinfection (SODIS) kinetics of Escherichia coli, enhanced with H2O2–part 1: The dark side of peroxide
(Elsevier, 2022-07-01) García-Gil, Ángela; Feng, Ling; Moreno-SanSegundo, José Ángel; Giannakis, Stefanos; Pulgarín, César; Marugán, Javier
The present bi-partite work describes the development and validation of a mechanistic kinetic model of SODIS E. coli inactivation, enhanced with H2O2. In this first part, the mechanism of the baseline dark phenomena is modelled. A mechanistic model involving E. coli cellular respiration, inactivation due to HO· and O2·- radicals, and bacterial thermal inactivation, was developed using a series-event model based on the accumulation of damage and cell recovery corrected with the Arrhenius equation for inclusion of the thermal events. The contribution of external H2O2 was included in the internal H2O2 balance, while the balance of extracellular H2O2 considered the consumption caused by its self-decomposition, interactions with cells’ membrane, and organic matter from dead cells. Specifically, the kinetic parameters of the external H2O2 sinks, the oxidation reaction of intracellular Fenton, and bacterial thermal inactivation were independently estimated by model regression from experimental data of E. coli inactivation and H2O2 consumption at different controlled conditions of temperature and initial H2O2 concentration. We complemented the values of the kinetic constants available in the literature with the unknown kinetic parameters estimated from experiemetnal and literature data. The missing kinetic parameters were successfully validated (bacteria error = 4.5%, H2O2 error = 12.9%). This kinetic model helps to understand the intracellular mechanisms and the contributions of each source of inactivation, with the role of radicals’ damage being most important at temperatures below 40 °C, and the thermal inactivation for temperatures above this value.
Mechanistic modelling of solar disinfection (SODIS) kinetics of Escherichia coli, enhanced with H2O2–Part 2: Shine on you, crazy peroxide
(Elsevier, 2022-07-01) García-Gil, Ángela; Feng, Ling; Moreno-SanSegundo, José Ángel; Giannakis, Stefanos; Pulgarín, César; Marugán, Javier
In this second part of the development of a mechanistic kinetic model of the solar inactivation of E. coli enhanced with hydrogen peroxide, we evaluate the mechanisms based on photonic inactivation and integrate them into the kinetic model of the dark process developed in Part 1. The direct photonic inactivation was modelled using a series-event model based on the accumulation of damage by photons and it was coupled with the model used in Part 1 for modelling the damage caused by radicals using a multiple target – multiple hit model, including recovery constant to define the ability of cells to face the specific photonic damage. Catalase and superoxide dismutase inactivation, the intracellular photo-Fenton reaction, and the overproduction of O2•- in the NADH/NAD+ cycle under solar light were included in the model. Finally, the synergistic effect of the photonic damage with thermal inactivation was included in the kinetic constant of the series-event expression in terms of an Arrhenius equation. The kinetic parameters were obtained by model regression using experimental data at different temperatures, solar radiation, as well as initial cellular and H2O2 concentrations. Our model predictions can accurately describe the experimental data of the SODIS process enhanced with H2O2, thus being very useful to estimate disinfection profiles and inactivation routes at different irradiance conditions, water temperature and H2O2 concentration. Finally, an integrated mechanism of E. coli inactivation under the SODIS/H2O2 process is provided.
Predicting the bactericidal efficacy of solar disinfection (SODIS): from kinetic modeling of in vitro tests towards the in silico forecast of E. coli inactivation
(Elsevier, 2022-01-01) Samoili, Sofia; Farinelli, Giulio; Moreno-SanSegundo, José Ángel; McGuigan, Kevin G.; Marugán, Javier; Pulgarín, César; Giannakis, Stefanos
In this study, the possibility of predicting the efficacy of Solar water disinfection (SODIS) for the removal of bacterial pathogens was assessed by the development of a three-level plan: firstly, systematic E. coli inactivation was performed (in vitro) in Lake Geneva water, under otherwise controlled conditions of water temperature (20–50 °C), sunlight intensity (0–1200 W/m2), presence of natural dissolved organic matter (DOM, 0–6 mg/L) and turbidity (0–50 NTU). As a second step a kinetic evaluation led to the selection of the most relevant parameters to be included in a novel static and dynamic model theoretical formulation. The static and dynamic models reliably described the experimental findings (bacterial inactivation under various climatic conditions) and were considered as equally eligible candidates for disinfection modeling. The final step considered ambient temperature, incident radiation and cloud-cover data to forecast (in silico) SODIS efficacy in Africa as a case study. The simulation results were compared with the experimental data and indicated that most African regions are suitable for SODIS processes, but there are areas of risk correlated with climatological conditions (cloud-cover and temperature). The results of this study could be applied for regional in decision-making strategies for application of SODIS or in the search for viable alternatives to SODIS in cases where it is deemed unsuitable.