In the world we live in today, and especially in recent decades, the use of the Internet has become a daily activity for a large part of the population. One of the latest applications that is growing exponentially is the Internet of Things (IoT). The relentless increase in the use of IoT devices has become an issue to be solved. It is necessary to feed all these devices which, despite requiring little power to operate, are a large number and can be found in places that are difficult to access. The project within which this work is framed consists of powering these IoT devices with sustainable energy harvesters installed inside smart buildings. These devices can be used to harvest light, radio frequency, thermal or mechanical energy. In our case, light energy has been studied As we will see later, in the P-V characteristic of a photovoltaic cell it is easily recognizable the maximum power point, that is, the point where we want the cell to work. Therefore, the objective of this work is to design and characterize MPPT or Maximum Power Point Tracker circuits that are able to make the light energy harvesters work as close as possible to their maximum power. One aspect to consider is that photovoltaic cells under artificial light will provide small power levels, and therefore the consumption of the MPPT circuits cannot be elevated. After performing a characterization of commercial photovoltaic cells under different types of artificial light, the most optimal MPPT technique for this study was found to be Fractional Open Circuit Voltage (FOCV). Three commercial MPPT circuits have been connected to photovoltaic cells installed under artificial light. The BQ25570 model, from Texas Instruments, has been found to be the most efficient MPPT over a wide range of illuminances, reaching values higher than 90% efficiency. Analog Devices¿ ADP5092 has shown similar efficiency values at high illuminances but significantly lower efficiency at low illuminances, with values below 60%. Finally, the MPPT SPV1050, from STMicroelectronics, has been clearly less efficient than the others.
estudiantat.upc.edu
