Dissertação (Mestrado em Engenharia Mecânica) - Centro Universitário FEI, São Bernardo do Campo, 2016.

Biochemical reactions make up most of the activity in a cell. There is inherent stochasticity in the kinetic behavior of biochemical reactions which in turn governs the fate of various cellular processes. In this work, the precision of a method for dimensionality reduction for stochastic modeling of biochemical reactions is evaluated. Further, a method of stochastic simulation of reaction kinetics is implemented in case of a specific biochemical network involved in maintenance of long-term potentiation (LTP), the basic substrate for learning and memory formation. The dimensionality reduction method diverges significantly from a full stochastic model in prediction the variance of the fluctuations. The application of the stochastic simulation method to LTP modeling was used to find qualitative dependence of stochastic fluctuations on reaction volume and model parameters. / text

Solar and wind generation are one of the most rapidly growing renewable energy sources, and is regarded as an appealing alternative to conventional power generated from fossil fuel. This is leading to significant levels of distributed renewable generation being installed on distribution circuits. Although renewable generation brings many advantages, circuit problems are created due to its intermittency, and overcoming these problems is a key challenge to achieving high penetration. It is necessary for utilities to understand the impacts of Photovoltaic (PV) generation on distribution circuits and operations. An impact study is intended to quantify the extent of the issues, discover any problems, and investigate alternative solutions. In this manner, system wide and local impact study are proposed in the dissertation. 1) System wide impact study This study considers system effects due to the addition of Plug-in Hybrid Vehicles (PHEV) and Distributed Energy Resource (DER) generation. The DER and PHEV are considered with energy storage technology applied to the residential distribution system load. Two future year scenarios are considered, 2020 and 2030. The models used are of real distribution circuits located near Detroit, Michigan, and every customer load on the circuit and type of customer are modeled. Monte Carlo simulations are used to randomly select customers that receive PHEV, DER, and/or storage systems. The Monte Carlo simulations provide not only the expected average result, but also its uncertainty. 2) Local impact study Analysis of high PV penetration in distribution circuits using both steady-state and quasi steady-state impact studies are presented. The steady-state analysis evaluates impacts on the distribution circuit by comparing conditions before and after extreme changes in PV generation at three extreme circuit conditions, maximum load, maximum PV generation, and when the difference between the PV generation and the circuit load is a maximum. The quasi steady-state study consists of a series of steady-state impact studies performed at evenly spaced time points for evaluating the spectrum of impacts between the extreme impacts. Results addressing the impacts of cloud cover and various power factor control strategies are presented. PV penetration levels are limited and depend upon PV generation control strategies and the circuit design and loading. There are tradeoffs in PV generation control concerning circuit voltage variations, circuit losses, and the motion of automated utility control devices. The steady state and quasi steady-state impact studies provide information that is helpful in evaluating the effect of PV generation on distribution circuits, including circuit problems that result from the PV generation. In order to fully benefit from wind power, accurate wind power forecasting is an essential tool in addressing this challenge. This has motivated researchers to develop better forecast of the wind resources and the resulting power. As a solution for wind generation, frequency domain approach is proposed to characterize and analyze wind speed patterns in the dissertation. 3) Frequency Domain Approach This study introduces the frequency domain approach to characterize and analyze wind speed patterns. It first presents the technique of and the prerequisite conditions for the frequency domain approach. Three years of wind speed data at 10 different locations have been used. This chapter demonstrates that wind speed patterns during different times and at different locations can be well characterized by using the frequency domain approach with its compact and structured format. We also perform analysis using the characterized dataset. It affirms that the frequency domain approach is a useful indicator for understanding the characteristics of wind speed patterns and can express the information with superior accuracy. Among the various technical challenges under high PV penetration, voltage rise problems caused by reverse power flows are one of the foremost concerns. The voltage rises due to the PV generation. Furthermore, the need to limit the voltage rise problem limits PV generators from injecting more active power into the distribution network. This can be one of the obstacles to high penetration of PVs into circuits. As a solution for solar generation, coordinated control of automated devices and PV is proposed in the dissertation. 4) Coordinated Automated Device and PV Control A coordinating, model-centric control strategy for mitigating voltage rise problems due to PV penetration into power distribution circuits is presented. The coordinating control objective is to maintain an optimum circuit voltage distribution and voltage schedule, where the optimum circuit operation is determined without PV generation on the circuit. In determining the optimum circuit voltage distribution and voltage schedule, the control strategy schedules utility controls, such as switched capacitor banks and voltage regulators, separate from PV inverter controls. Optimization addresses minimizing circuit losses and motion of utility controls. The coordinating control action provides control setpoints to the PV inverters that are a function of the circuit loading or time-of-day and also the location of the PV inverter. Three PV penetration scenarios are considered, 10%, 20%, and 30%. Baselines with and without coordinating controls for circuit performance without PV generation are established, and these baselines are compared against the three PV penetration scenarios with and without coordinating control. Simulation results are compared and differences in voltage variations and circuit losses are considered along with differences in utility control motion. Results show that the coordinating control can solve the voltage rise problem while minimizing circuit losses and reducing utility control motion. The coordinating control will work with existing PV inverter controls that accept control setpoints without having to modify the inverter controls. 5) Coordinated Local and Centralized PV Control Existing distribution systems and their associated controls have been around for decades. Most distribution circuits have capacity to accommodate some level of PV generation, but the question is how much can they handle without creating problems. It proposes a Configurable, Hierarchical, Model-based, Scheduling Control (CHMSC) of automated utility control devices and photovoltaic (PV) generators. In the study here the automated control devices are assumed to be owned by the utility and the PV generators and PV generator controls by another party. The CHMSC, which exists in a hierarchical control architecture that is failure tolerant, strives to maintain the voltage level that existed before introducing the PV into the circuit while minimizing the circuit loss and reducing the motion of the automated control devices. This is accomplished using prioritized objectives. The CHMSC sends control signals to the local controllers of the automated control devices and PV controllers. To evaluate the performance of the CHMSC, increasing PV levels of adoption are analyzed in a model of an actual circuit that has significant existing PV penetration and automated voltage control devices. The CHMSC control performance is compared with that of existing, local control. Simulation results presented demonstrate that the CHMSC algorithm results in better voltage control, lower losses, and reduced automated control device motion, especially as the penetration level of PV increases. / Ph. D.

La micropile à combustible à respiration est développée conjointement à STMicroelectronics Tours et au CEA Liten de Grenoble. De très faible puissance (stack de 1W), elle sera à moyen terme utilisée dans un système de recharge portable pour petites batteries Li-Ion (téléphones portables). Le fonctionnement et la structure de ces micropiles sont tels qu'elles sont très sensibles, entre autres, aux conditions atmosphériques caractérisant leur environnement. Cette sensibilité résulte en un comportement électrique très marqué et complexe. Or, l'aspect nomade de l'application fait que celle-ci devra pouvoir faire face à des atmosphères diverses et variées. Il est donc nécessaire de comprendre les interactions liant le comportement électrique de la micropile et l'environnement. Leur modélisation pourra par la suite apporter des éléments concrets en termes de pilotage d'auxiliaires (micro ventilateurs…) et de design de packaging, visant à contrôler l'environnement immédiat de la micropile de la meilleure façon possible. A cet effet, de nombreuses mesures, réalisées sous atmosphère maîtrisée, et sous plusieurs régimes de fonctionnement électrique, ont été croisées entre elles. Elles nous ont permis de poser les hypothèses d'un modèle quasistatique macroscopique de la micropile, reliant les conditions atmosphériques et opératoires à la réponse électrique de la micropile. Ce modèle a été développé à partir de la théorie de la diffusion en milieu poreux. Ce modèle quasistatique, faisant intervenir une description de la diffusion protonique cathodique, permet de représenter le comportement de la micropile sur une large gamme de conditions atmosphériques, et illustre physiquement autant les situations d'assèchement que de noyage. L’approche a ensuite été élargie au développement d'un modèle petit signal, paramétré grâce à une approche multi spectrale et multi conditions opératoires. Celui-ci permet entre autres de quantifier la dynamique associée au phénomène de diffusion protonique, tout en consolidant sa description quasistatique, ceci faisant intervenir des paramètres cohérents avec ceux du modèle quasistatique. Enfin, à la croisée des approches quasistatique et petit signal, les bases d'un modèle dynamique fort signal sont proposées. Elles font intervenir le modèle fort signal propre au LAPLACE, en y injectant la réponse dynamique à l'environnement et à la sollicitation électrique du bilan hydrique. Ce modèle, paramétré avec les paramètres issus du quasistatique et du petit signal, permet de représenter le comportement non linéaire de la micropile sur une large gamme de fréquences de sollicitations galvanostatiques fort signal. / The micro breathing fuel cell is developed by STMicroelectronics Tours and the CEA Liten of Grenoble. It is very low power (1W stack) and will eventually be used in a portable charging system for small Li-Ion batteries (cell phones). The structure of these micro fuel cells is such that they are very sensitive, among other things, to weather conditions characterizing their environment. This sensitivity results in a very complex electrical behavior. But the portable aspect of the application implies that it will have to cope with various atmospheres. It is therefore necessary to understand the interactions linking the electrical behavior of the micro fuel cell and the atmosphere. A model may then provide some concrete leads in terms of auxiliary control (micro fans ...) and packaging design, to control the immediate environment of the microcell in the best possible way. To this end, a lot of measure were carried out under controlled atmosphere, and in several electrical operating modes, and were crossed with each other. They let us build the assumptions for a macroscopic steady state model of micro fuel cell, linking atmospheric and operating conditions to the electrical response of the micro fuel cell. This model was inspired by the theory of diffusion in porous media. This steady state model, involving a description of a cathodic protonic diffusion, is used to represent the behavior of the micro fuel cell on a wide range of atmospheric conditions, and physically illustrates both drying out situations than drowning. The approach was then extended to develop a small signal model, configured with a multi spectral and multi-operating conditions approach. It allows among other things to quantify the dynamics associated with the phenomenon of proton diffusion, while consolidating its steady state description, this involving parameters consistent with those of the steady state model. Finally, at the intersection of the steady state and small signal approaches, the bases for a large signal dynamic model are proposed. They involve the large signal model which is specific to the LAPLACE, by injecting in it the dynamic response to environmental stress and to water balance. This model, with parameters set from the steady state and small signal models, turns out to be able to represent the nonlinear behavior of the micro fuel cell over a wide range of frequencies of the galvanostatic strong signal solicitation