A Method for Searching for Defective Solar Panels in Telemetry Data of a Power Plant Based on the Results of a Digital Twin

Abstract

Searching for faulty, and therefore operating in abnormal mode, solar panels at a power plant is an urgent task in the context of the development and growth of the share of solar energy in electricity generation. The research is aimed at developing and evaluating the effectiveness of a new methodology and software algorithm for searching for anomalies in the operation of solar panels based on the results of a digital twin created and trained using telemetry data from a solar power plant. The methodology is based on studies of deviations in power values at the point of maximum efficient operation of the solar panel, calculated by the digital twin, from the average statistical values for the power plant. Using the proposed methodology, over six months of direct observations, 16 anomalies in the operation of the solar panels of the power plant were discovered and confirmed. It has been established that when analyzing deviations of normalized power values at the maximum power point PN, it is possible to detect solar panels that have defects or operate with loss of efficiency.