Search

Large-scale systems are characterised by having a large number of components working in coordination. These systems can be composed of geographically distributed elements with resource limitations. In this way, control strategies for large-scale systems have challenges related to information flow, processing time and capacity, controller design, and energyresource optimisation. One particular large-scale system is the next-generation mobile communications cellular network, which will comprise hundreds of base stations interacting with thousands of users in milliseconds. One of the main challenges with next-generation cellular networks (NGCNs) is the higher energy consumption caused by the expected number of connected devices. Here, renewable energies are a good option to face the growing demand of energy consumption. However, there are still major challenges related to the appropriate control schemes to minimise ongrid consumption and optimise energy management in cellular networks with hybrid energy sources (grid and renewable energy). In this thesis, different control strategies for large-scale networks are proposed. These control strategies are assessed over an NGCN powered by hybrid energy sources to reduce grid consumption. The energy-efficiency problem is studied from the viewpoint of the energetic processes – i.e., on-place renewable energy is available, and mechanisms to reduce the gridenergy consumption should be developed. The proposed mechanisms are based on previous research that shows the relationship between the number of users connected to a cellular network base station (BS) and its energy consumption. For this reason, the study of optimal control mechanisms that balance the load of users over the available BSs according to the renewable energy available is a key element in the field of energy efficiency in cellular networks. These schemes are assessed through simulations and then compared with the scheme actually used to manage the user–BS association in cellular networks. The results show that the proposed control schemes improve grid-electricity consumption compared with the traditional association mechanism while still maintaining adequate quality-of-service (QoS) levels. The control schemes for the energy-efficiency problem were studied in two timescales. The short timescale (of the order of seconds) was used to analyse the user–BS association problem in a network configuration with hybrid power sources without an energy storage system. The long timescale (of the order of hours) was used to study load balancing of aggregated traffic in each BS with hybrid power sources and an energy storage system. Finally, the proposed controllers are of different types: (i) centralised, (ii) distributed at the base-station level, and (iii) distributed at the user level Resumen: Los sistemas de gran escala se caracterizan por tener un gran n´umero de componentes trabajando de forma coordinada. Estos sistemas est´an conformados por elementos que pueden estar distribuidos a lo largo de una extensa ´area geogr´afica y poseer restricciones en cuanto a la disponibilidad de recursos necesarios para su funcionamiento. Teniendo en cuenta estas caracter´ısticas, las estrategias de control para sistemas de gran escala presentan retos relacionados con el flujo de la informaci´on, la capacidad y tiempo de procesamiento, el dise˜no de los controladores y la optimizaci´on de los recursos disponibles. Un tipo particular de sistema de gran escala son las redes de comunicaci´on m´ovil celular de pr´oxima generaci´on, que se encontrar´an conformadas por cientos de estaciones base que interactuar´an con miles de usuarios en instantes de tiempo del orden de los milisegundos. Uno de los principales desaf´ıos en las redes celulares de pr´oxima generaci´on (RCPG) es el incremento en el consumo energ´etico causado por el crecimiento esperado de dispositivos conectados. En este contexto, las energ´ıas renovables son una buena alternativa para afrontar la creciente demanda de consumo energ´etico. Sin embargo, existen importantes desaf´ıos relacionados con los esquemas de control adecuados para minimizar el consumo de energ´ıa proveniente de la red el´ectrica convencional (grid) y optimizar la gesti´on energ´etica en redes celulares con fuentes de alimentaci´on h´ıbrida (grid y renovable). En esta tesis, se proponen y eval´uan diferentes estrategias de control para redes de gran escala, utilizando como caso de estudio las RCPG alimentadas con fuentes h´ıbridas y su objetivo de reducir el consumo grid. El problema de la eficiencia energ´etica es estudiado desde el punto de vista de los procesos energ´eticos, es decir, de la disponibilidad de energ´ıa renovable en el emplazamiento del sistema y los mecanismos para reducir el consumo energ´etico. Los mecanismos propuestos se basan en investigaciones previas que demostraron la relaci´on existente entre el n´umero de usuarios conectados a las estaciones base (EB) de la red y su consumo energ´etico. Por esta raz´on, el estudio de mecanismos de control ´optimo que balanceen la carga de tr´afico sobre las EB de acuerdo con la energ´ıa renovable disponible es un elemento clave en el campo de la eficiencia energ´etica en redes celulares. Estos esquemas son evaluados a trav´es de simulaciones y comparados con el mecanismo usado actualmente por las redes celulares para gestionar la asociaci´on de los usuarios a las EB. Los resultados de la tesis muestran que los esquemas de control propuestos mejoran el consumo grid comparado con el mecanismo de asociaci´on tradicional a la vez que mantienen adecuados niveles de calidad del servicio. Los esquemas de control para el problema de la eficiencia energ´etica fueron estudiados en dos escalas de tiempo. La corta escala de tiempo (del orden de los segundos) fue usada para analizar el problema de la asociaci´on de los usuarios a las EB en una configuraci´on de red con fuentes de potencia h´ıbridas y sin sistema de almacenamiento energ´etico. La larga escala de tiempo (del orden de horas) fue utilizada para estudiar el balanceo de carga de tr´afico agregado en cada EB, con fuentes h´ıbridas de potencia y con sistema de almacenamiento energ´etico. Finalmente, los controladores desarrollados son de diferentes tipos: i) esquema centralizado, ii) esquemas distribuidos a nivel de usuario y iii) esquemas distribuidos a nivel de estaciones base. Doctorado

International audience; This paper treats supercapacitors matrix characterization. In order to cut off transient power peaks and to compensate for the intrinsic limitations in embedded sources, the use of supercapacitors as a storage system is quite suitable, because of their appropriate electrical characteristics (huge capacitance, small series resistance, high specific energy, high specific power), direct storage (energy ready for use), and easy control by power electronic conversion. This use requires supercapacitors modules where several cells connected in serial and/or in parallel, thus a bypass system to balance the charging or the discharging of supercapacitors is required. In the matrix of supercapacitors, six elements of three parallel BCAP0350 supercapacitors in serial connections have been considered. This topology permits to reduce the number of the bypass circuits and it can work in degraded mode. Actually, it allows the system to have more reliability by providing power continually to the load even when there are one or more cells failed. Simulation and experimental results are presented and discussed.

International audience; This study presents a study of the reduction in battery stresses by using supercapacitors (SCs) in a 500-kVA rated UPS. We aim at investigating the optimal supercapacitors-battery combination versus the SCs cost. This investigation is threefold; first, supercapacitors and battery models developed using MATLAB/Simulink are presented and validated. Second, the architecture and the simulation of the designed system that combines the SCs and the battery are shown. The supercapacitors are used as high-power storage devices to smooth the peak power applied to the battery during backup time and to deliver full power during short grid outages. By charging the SCs through the battery at a suitable rate, all impulse power demands would be satisfied by the supercapacitors. Third, extensive simulations are carried out to determine the gain in battery RMS current, the gain in energy losses, the energy efficiency and the elimination rate of surge load power. These four performance parameters are determined by simulation and then analyzed. The influence of the SCs recharge on the performance indicators is highlighted. A thorough analysis involving optimal study proposes to draw the optimal SCs number and filter constant from the variation of the aforementioned parameters versus the cost of the SCs.

This paper treats supercapacitors matrix characterization. In order to cut off transient power peaks and to compensate for the intrinsic limitations in embedded sources, the use of supercapacitors as a storage system is quite suitable, because of their appropriate electrical characteristics (huge capacitance, small series resistance, high specific energy, high specific power), direct storage (energy ready for use), and easy control by power electronic conversion. This use requires supercapacitors modules where several cells connected in serial and/or in parallel, thus a bypass system to balance the charging or the discharging of supercapacitors is required. In the matrix of supercapacitors, six elements of three parallel BCAP0350 supercapacitors in serial connections have been considered. This topology permits to reduce the number of the bypass circuits and it can work in degraded mode. Actually, it allows the system to have more reliability by providing power continually to the load even when there are one or more cells failed. Simulation and experimental results are presented and discussed.

International audience