Las tendencias en la tecnología actual permiten la reducción tanto en tamaño como en potencia consumida de los sistemas digitales complejos. Esta disminución en el tamaño y el consumo da lugar al concepto de dispositivos portátiles que se integren en la vida pertenencias personales y cotidianas como ropa, relojes, gafas, etc. La fuente de alimentación es un factor limitante en la movilidad de los dispositivos portátiles que se ve reducida por la duración de la batería. Además, debido a los costos y difícil accesibilidad, la sustitución o recarga de las baterías a menudo no es viable para los dispositivos portátiles integrados en ropa inteligente. Los dispositivos vestibles están distribuidos en las pertenencias personales y, por tanto, la recolección de energía del usuario es una alternativa para su alimentación. Dispositivos vestibles pueden crear, al igual que los sensores de una red de sensores inalámbricos (WSN), una red de área corporal. El principal objetivo de esta tesis es el estudio de generadores piezoeléctricos, inductivos y termoeléctricos que recolectan energía del cuerpo humano de forma pasiva. El principio físico de un transductor es el mismo independientemente de si la fuente proviene del entorno o del cuerpo humano. Sin embargo, las limitaciones relacionadas con la baja tensión, corriente y niveles de frecuencia conllevan nuevos requerimientos que no están presentes en el caso de la utilización de las fuentes que ofrece el entorno y que suponen el principal desafío de esta tesis. El tipo de energía entrada y transductor a utilizar forman un tándem donde la elección de uno impone el otro. Es importante que las mediciones se realicen diferentes partes del cuerpo humano, mientras se realizan diferentes actividades físicas para localizar las posiciones y las actividades que producen más energía. El acoplamiento mecánico entre transductor y cuerpo humano depende de la ubicación del transductor y la actividad que se realiza. Un diseño específico, teniendo esto en cuenta puede aumentar más de un 200% la eficiencia del transductor como se ha demostrado con láminas piezoeléctricas situadas en plantillas de zapatos. Se han realizado mediciones de aceleraciones en diferentes partes del cuerpo y diferentes actividades para cuantificar la cantidad de energía disponible en actividades cotidianas. Se ha realizado una simulación a nivel de sistema, modelando los elementos de un sistema de energía autoalimentado. El transductor se ha modelado usando las ecuaciones físicas que lo describen con el objetivo de incluir la parte mecánica del sistema. Se han utilizado modelos eléctricos y de comportamiento para el resto de los componentes. De esta manera, el proceso de diseño de la aplicación en su conjunto (incluyendo la carga y un elemento de almacenamiento de energía cuando es necesario) se simplifica a la hora de lograr los requisitos planteados. Obviamente, la carga debe ser un dispositivo de bajo consumo como por ejemplo un transmisor RF. En este caso, es preferible alimentar la carga de forma discontinua, sin una batería, como se deduce de los resultados obtenidos mediante simulación. Sin embargo, la evolución de los transmisores RF de baja potencia puede cambiar esta conclusión en función sobre todo de la evolución del consumo de energía en stand-by y el tiempo de configuración para la operación de transmisión. Se ha deducido a partir del análisis de los generadores inductivos que el análisis en el dominio temporal permite calcular algunas magnitudes que no están disponibles en el dominio frecuencial. Por ejemplo, la potencia máxima se puede calcular en el dominio frecuencial, pero para aplicaciones de recolección de energía es más interesante saber el valor de la energía recuperada durante un cierto tiempo o la potencia media ya que la potencia generada por las actividades humanas pueden ser muy discontinua. Se ha demostrado que los transductores recolectores de energía son capaces de suministrar alimentación a dispositivos electrónicos de baja potencia, como quedó demostrado con un transmisor RF alimentado por una termogenerador que emplea el gradiente de temperatura existente entre el cuerpo humano y el entorno (3-5 K) y que es capaz de realizar medidas y transmitirlas una vez cada segundo / The trends in technology allow the decrease in both size and power consumption of complex digital systems. This decrease in size and power gives rise to the concept of wearable devices which are integrated in everyday personal belongings like clothes, watch, glasses, et cetera. Power supply is a limiting factor in the mobility of the wearable device which gets restricted to the lifetime of the battery. Furthermore, due to the costs and inaccessible locations, the replacement or recharging of batteries is often not feasible for wearable devices integrated in smart clothes. Wearable devices are devices distributed in personal belongings and thus, an alternative for powering them is to harvest energy from the user. Therefore, the energy can be harvested, distributed and supplied over the human body. Wearable devices can create, like the sensors of a Wireless Sensor Network (WSN), a Body Area Network. A study of piezoelectric, inductive and thermoelectric generators that harvest passive human power is the main objective of this thesis. The physical principle of an energy harvesting generator is obviously the same no matter whether it is employed with an environmental or human body source. Nevertheless, the limitations related to low voltage, current and frequency levels obtained from human body sources bring new requirements to the energy harvesting topic that were not present in the case of the environment sources. This analysis is the motivation for this thesis. The type of input energy and transducer form a tandem since the election of one imposes the other. It is important that measurements are done in different parts of the human body while doing different physical activities to locate which positions and activities produce more energy. The mechanical coupling between the transducer and the human body depends on the location of the transducer and the activity that is done. A specific design taking this into account can increase more than a 200% the efficiency of the transducer as has been demonstrated with piezoelectric films located in the insoles of shoes. Acceleration measurements have been performed in different body locations and different physical activities, in order to quantify the amount of available energy associated with usual human movements. A system-level simulation has been implemented modeling the elements of an energy self-powered system. Physical equations have been used for the transducer in order to include the mechanical part of the system and electrical and behavioral models for the rest of the components. In this way, the process of the design of the complete application (including the load and an energy storage element when it is necessary) is simplified to achieve the expected requirements. Obviously, the load must be a low power consumption device as for example a RF transmitter. In this case, it is preferable to operate it in a discontinuous way without a battery as it is deduced from simulation results obtained. However, the evolution in low power transmission modules can change this conclusion depending mostly on the evolution of the power consumption in stand-by mode and the configuration time in transmission operation. It has been deduced from the analysis of inductive generators that time-domain analysis allows to calculate some magnitudes that are not available in frequency domain. For example, the maximum power can be calculated in frequency domain, but for energy harvesting applications it is more interesting to know the value of the recovered energy during a certain time, or the average power since the power generated by human activities can be highly discontinuous. It has been demonstrated that energy harvesting transducers are able to supply power to present-day low power electronic devices as was demonstrated with a RF transmitter powered by a thermogenerator that employs the temperature gradient between human body and the environment (3-5 K) and that it is able to sense and transmit data once every second.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / This dissertation presents the development of a system for converting the mechanical energy from vibrations into electrical energy. The conversion is performed in a low-cost piezoelectric transducer, commonly known as buzzer. The main purpose of this system is to charge, or to extend the time between charges, of chargeable batteries up to 2V. In order to control the charging process, an integrated energy processor was also designed. Processor design is presented from its specification, followed by circuit topology definition, electric simulation, layout, extraction of circuit from layout and a final simulation including layout effects. The main contribution of the investigation is to show how much energy could be obtained from vibrations with a low-cost transducer, comparing its performance to full custom generators. The final system implementation is very simple, composed by a generator (a buzzer with a steel ball glued onto its center) and an integrated circuit that controls the charge delivered to the battery, sensing the voltage across its terminals. An efficiency of 55% is expected, being comparable to results published by other researchers. / Esta dissertação apresenta o desenvolvimento de um sistema para converter energia mecânica de vibrações em energia elétrica. A conversão é realizada através de um transdutor de baixo custo comumente chamado de buzzer. O principal objetivo deste sistema é carregar ou estender o tempo entre cargas de baterias recarregáveis de até 2 V. Para o controle do processo de carga, um processador de energia integrado também foi desenvolvido. O projeto do processador de energia é apresentado segundo especificações como definição de topologia, simulação elétrica, layout, extração elétrica do circuito através do layout e a simulação final incluindo os efeitos do layout. A principal contribuição desta dissertação é mostrar como muita energia poderia ser obtida de vibrações com um transdutor de baixo custo, comparando sua performance a outros microgeradores. O sistema final implementado é muito simples, composto por um microgerador (buzzer com uma esfera de aço colada no centro) e um circuito integrado que controla a carga da bateria, através da monitoração da tensão da mesma. Uma eficiência de 55% é esperada, sendo comparável com os resultados obtidos por outros pesquisadores.

Le domaine de l’assistance cardiaque connait actuellement une rupture technologique avec l’apparition du pacemaker sans fil. Grâce à ces nouveaux dispositifs, la prise en charge des patients est simplifiée. En outre, la suppression des sondes devenues obsolètes devrait permettre une réduction drastique des problèmes rencontrés avec les pacemakers traditionnels. Cependant, la question de l’alimentation reste posée. Dans ce travail de thèse, nous tentons d’apporter une solution à base de microgénérateur piézoélectrique inertiel récupérant une portion de l’énergie vibratoire des battements cardiaques. La démarche suivie consiste tout d’abord à définir le besoin et la pertinence d’une solution à base de récupérateur d’énergie. Nous analysons ensuite l’allure de signaux cardiaques qui ont été enregistrés à l’aide d’accéléromètres directement positionnés sur le site de stimulation. On montre ainsi que le gisement vibratoire adressé (i.e. les battements cardiaques) imposent des récupérateurs vibrant aux alentours de 16 Hz. Ces fréquences sont extrêmement faibles en comparaison des microgénérateurs présentés dans la littérature (typ. > 100 Hz). Dans un second temps, et indépendamment de considérations purement technologiques, nous établissons, à l’aide de modèles analytiques et numériques, le dimensionnement optimal permettant de répondre simultanément aux spécifications dimensionnelles et au niveau de puissance récoltée nécessaire. Cette phase d’optimisation montre qu’un compromis entre fréquence de résonance et puissance délivrée doit être fait et, plus particulièrement, que celui-ci conduit à l’expression d’un besoin en termes d’épaisseur de couches piézoélectriques auquel aucune des technologies standards ne permet de répondre. Nous présentons, dans ce manuscrit, les travaux qui ont ainsi été menés pour développer une technique de réalisation de couches épaisses de PZT (typ. 15 à 100 µm) par amincissement de céramiques massives. Ce mode de réalisation est enfin mis en œuvre pour la fabrication d’un démonstrateur à l’échelle, de type poutre encastrée-libre bimorphe vibrant à 16 Hz. Nous montrons finalement que les résultats obtenus à partir de battements cardiaques reproduits en laboratoire (10-15 µW) sont en ligne avec les besoins exprimés pour la mise en œuvre d’une solution d’alimentation pour pacemaker sans fil. Ce travail de thèse a été conduit dans le cadre du projet HBS (Heart Beat Scavenging) notamment en collaboration avec la société LivaNova-Sorin CRM (Cardiac Rythm Management). Il est fortement probable que la décision initiale d’articuler l’ensemble de tâches accomplies autour des besoins de l’utilisateur final soit une des clés de la réussite de ce travail. En effet, les démonstrateurs développés dans ce travail de thèse ont, par la suite, été testés avec succès sur l’animal. Ils ont également donné lieu à un nouveau projet dont un des objectifs est d’adresser les aspects de fiabilité et de vieillissement. Ces nouvelles tâches correspondent ainsi à la poursuite de la montée en TRL (Technology Readiness Level) vers les étapes de pré-industrialisation. / The field of cardiac assistance is currently experiencing a new technological breakthrough with the introduction of the leadless pacemaker. With these new devices, the care of patients is simplified. Furthermore, removal of the leads should allow a drastic reduction of the problems encountered with conventional pacemakers. However, the question of the energy supply remains. In this thesis, we try to provide a solution based on piezoelectric inertial micro-generator in order to harvest a portion of the heartbeat vibrational energy. The approach is to first define the need and relevance of a solution based on energy scavenging. We then analyze the cardiac signals that were recorded using accelerometers positioned directly on the stimulation site. It is shown that the addressed vibration source (i.e. heartbeats) impose the devices to vibrate at around 16 Hz. These frequencies are extremely low compared to microgenerators presented in the literature (typ.> 100 Hz). Secondly, regardless of technological considerations, and using analytical and numerical models, we identify the optimal device dimensions in order to simultaneously meet the specifications in terms of size and required harvested power. This optimization phase shows that a trade-off between resonant frequency and output power must be made and, more particularly, that it leads to the expression of a need in terms of piezoelectric layer thickness to which none of the standard technologies can currently answer. Therefore, we present the work that has been undertaken to develop a technique for producing thick layers of PZT (typ. 15 to 100 µm) by the thinning and the polishing of bulk ceramics. Then, this technique is implemented for the fabrication of our demonstrator: a cantilever of bimorph type vibrating at 16 Hz. Finally, we show that the obtained results (10-15 µW) from heartbeats reproduced in the laboratory are in line with the expressed needs for the implementation of an energy supply solution for leadless pacemakers. This thesis work has been conducted in the frame of the HBS project (Heart Beat Scavenging) especially in collaboration with the company LivaNova - Sorin CRM (Cardiac Rhythm Management). It is highly believed that the original decision to articulate all the tasks that we performed around the end user needs was a key to the success of this work. Indeed, the demonstrators developed in this thesis have subsequently been successfully tested on animals. They also led to a new project whose objectives are to address the reliability and aging of these demonstrators. These new tasks correspond to the continuation of the TRL increase (Technology Readiness Level) to the stages of pre-industrialization.