Winterthur/Zurich – Researchers from the Zurich University of Applied Sciences (ZHAW), in collaboration with Fluence Energy, have developed a software module for optimizing existing solar installations. It uses artificial intelligence to combine specialist knowledge with data.
Researchers from the Institute for Data Analysis and Process Design (IDP) at ZHAW in Winterthur, in collaboration with Fluence Energy, have developed a software module that allows for the optimal maintenance of existing solar installations. The program is capable of detecting and categorizing energy losses generated by failures, tracker errors, or pollution, as explained by ZHAW in a statement. With the help of artificial intelligence (AI), the module can also indicate whether the costs resulting from the energy loss are greater than the costs of any repairs carried out.
"Our challenge was to develop a software module that generates its estimates using data-driven AI but also incorporates the extensive expertise about the facility to make decisions that are comprehensible even for the engineers operating the plant," said project leader Lilach Goren Huber from IDP in the statement. To achieve this, the researchers constructed a neural network that combines data obtained from the operation of the facility with the expertise of its operators. In this process, the required amounts of data on disruptions were generated using expertise as artificial data with realistic disruption patterns. "With the combination of valuable expertise and a data-driven deep learning model, the module can detect faults in the facility earlier and more accurately than previous approaches," explained the ZHAW researcher. ce/hs