The successful completion of our cooperation with KInIT - the Kempelen Institute of Intelligent Technologies in the field of data analysis and detection of anomalies and anomalous measurements in a group of supply points – opened a new stage of cooperation for us. An important part of the cooperation was completed on June 30^th of this year and focused on quantifying the "abnormality rate" of measurements in the balance group. As a part of this collaboration, several methods were successfully tested. Its result was transformed into a structured form for further use at the technical and commercial levels. The detection algorithm was designed to allow continuous technical and commercial monitoring and evaluation of the "anomaly rate" of the customer’s portfolio.

The fresh step in partnership focuses on increasing the accuracy of power production prediction from photovoltaic power stations. This will be achieved through the use of advanced statistical models, machine learning, and artificial intelligence methods. The research task has a broader scope and to extends to several modern trends in the electric power industry, such as digitization, optimization, and management, or the integration of renewable energy sources into the electricity system. The final algorithm will be able to predict the performance of a single photovoltaic power plant as well as a group of power plants within a virtual power plant.