Implementace energetického managementu pro senzory napájené termoelektrickým generátorem

Abstract

This thesis deals with energy management methods for sensors powered by a thermoelectric generator. The thesis examines methods applied at both hardware and software levels of the monitoring device in order to optimize energy consumption and extend the lifetime of the device. The proposed energy management in this work involves the implementation of an algorithm from the field of machine learning, namely Q-learning. This algorithm is designed to dynamically control the activity of the device depending on the energy availability. The solution has been tested in terms of energy consumption and device response to decreasing amounts of available energy. The device has demonstrated the ability to operate when powered by a volatile energy source.