Master theses
Development of Analytical Tools for Diagnosing Photovoltaic Power Plants Using Machine Learning
Author
Samuel Händl
Year
2026
Type
Master thesis
Supervisor
Ing. Vojtěch Zavřel, Ph.D.
Reviewers
Ing. Ivo Petr, Ph.D.
Department
Summary
The thesis focuses on the development of analytical tools for diagnosing photovoltaic power plants using machine learning.
The main objective is to design and implement a method for estimating the tilt and azimuth of photovoltaic panels based on power output data and PV system location. The study applies nonlinear least square method that minimizes the difference between POA irradiance and power output data.
The implementation is carried out in Python, emphasizing code modularity and reusability. The methodology is verified using data from real installations, and its accuracy is evaluated using mean absolute error.