St. Experts of St. Petersburg State Electrotechnical University “LETI” have developed a neural network that detects defects in the production of elements for solar cells. LETI told socialbites.ca about this.
Solar energy is one of the most promising alternative energy types today. Solar elements are manufactured from a relatively cheap and widely available semiconductor material (silicon). Despite the fact that modern silicon cell production is almost completely automated, errors occur during the transportation of silicon wafers between various technological processes.
Detection of defects in silicon wafers is carried out using the effect of electroluminescence. Its essence is that when voltage is applied to a silicon solar cell, its surface glows (electroluminescence effect), which is clearly defined using an infrared camera (IR). If there is no defect on the surface, the solar cell will shine evenly. The presence of defects will be indicated by areas of different brightness levels.
Scientists collected and processed a database containing IR images of more than 100 thousand silicon solar cells from the production line of Hevel LLC in Novocheboksarsk. Based on this information, experts taught the neural network to highly accurately identify and classify various defects that arise at different stages of solar cell production.
“Although the number of defects during production is 0.2-0.3%, for solar energy this is a significant amount that affects the efficiency of the solar cell. Now manufacturers of high-efficiency silicon solar cells are fighting for a tenth of a percent. Using our analytics tool, you can not only detect defects, but also understand where they occur in the production line and then eliminate them. Thus, the method we developed made it possible to increase the production efficiency of the Hevel LLC enterprise,” Evgeniy Terukov, St. Professor of the Photonics Department of St. Petersburg Electrotechnical University “LETI” and Deputy Director General for Scientific and Technical Research Center for Thin Film Technologies in Energy told socialbites.ca.
Previously on LETI import changed Microwave antennas that can be quickly configured to control drones.
Source: Gazeta
Jackson Ruhl is a tech and sci-fi expert, who writes for “Social Bites”. He brings his readers the latest news and developments from the world of technology and science fiction.