Development of an IOT-based measurement system for investi̇gating energy losses due to dust in photovoltaic systems

Solar power plants, known for their environmentally friendly structures, significantly contribute to reducing the carbon footprint and are emerging as a renewable alternative to conventional energy sources worldwide. The increasing use of these plants offers both environmental and economic advantages. However, one of the strengths of solar power plants, the traceability of production data, is often limited due to the high cost of sensors used. Additionally, the lack of monitoring metrics for dust accumulation and the uncertainty of dust-related losses can lead to power losses, negatively affecting the system's efficiency. This study aims to develop a low-cost, Internet of Things (IoT)-based monitoring system for solar power plants, utilizing fuzzy logic methods for pollution status decisions. The system calculates dust accumulation rates based on power output and uses fuzzy logic to decide on the pollution status of solar panels. At the Bursa Technical University Mimar Sinan Campus, two identical solar panels were monitored and recorded using IoT-based sensors to track values such as current and surface temperature. These measurements were used to calculate solar panel output power, irradiation, and dust accumulation rate, creating real data-based measurement records. One panel was cleaned weekly, while the other was left dirty. Sensors were connected to an open-source programmable circuit board, the ESP32, to create a low-cost IoT-based monitoring system. The thesis also developed a decision-making method using fuzzy logic for determining the pollution status of solar panels. The developed method uses dust accumulation rates based on solar panel power output to decide on the panels' pollution status. The results from the thesis show that an IoT-based monitoring system for solar power plants can be implemented cost-effectively, dust accumulation rates can be calculated based on power output, and pollution status decisions can be made using a fuzzy logic decision-making system. In this regard, the thesis outputs are thought to be applicable in existing solar power generation facilities in practice.